&EPA
United States
Environmental Protection
Agency
Robert S. Kerr
Environmental Research
Laboratory
Ada OK 74820
EPA-600/2-78-058
March 1978
Research and Development
Proceedings of the
Second Open Forum on
Management of
Petroleum Refinery
Waste water
Environmental Protection
Technology Series
-------
EPA-600/2-78-058
March 1978
PROCEEDINGS OF THE SECOND OPEN FORUM
ON
MANAGEMENT OF PETROLEUM REFINERY WASTEWATER
Presented by
The U.S. Environmental Protection Agency
The American Petroleum Institute
The National Petroleum Refiners Association
The University of Tulsa
Project Officer
Fred M. Pfeffer
Source Management Branch
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma
Editor and Project Director
Francis S. Manning
University of Tulsa
Tulsa, Oklahoma
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ADA, OKLAHOMA 74820
-------
DISCLAIMER
This report has been reviewed by the Robert S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
CONTENTS
Disclaimer ii
Acknowledgements vi
Welcome Address 1
Keynote Address 3
SESSION I Regulatory and Research Direction in U.S. EPA 12
Lamar Miller "Toxic Strategy Overview" 13
Marvin L. Wood "EPA's Research in the Refining Category" 20
SESSION II Panel Discussion: Research Supported in the Private Sector 35
Robert T. Denbo "Discussion of Research Related to Petroleum 36
Refinery Wastewater Sponsored by the API
Committee on Refinery Environmental Control"
Kent G. Drummond "Water Quality Committee of API" 43
Francis S. Manning "Overview of Research on Petroleum Refinery 48
Wastewaters at U.S. Universities"
David C. Bomberger "Overview of Research on Petroleum Refining 53
Wastewaters at Independent Contract Research
Organizations"
Judith G. Thatcher "Current API Studies of Residuals in Refinery 56
Effluents"
SESSION III Open Question and Answer Session - Individual Problems in 64
Management of Refinery Wastewater
Banquet Proceedings and Address 80
SESSION IV Origin and Interpretation 85
Ridgway M. Hall, Jr. "Regulation of Problem Pollutants Under the 87
Federal Water Pollution Control Decree"
W. M. Shackelford "Evolution of the Priority Pollutant List From 103
the Consent Decree"
R. W. Dellinger "Incorporation of the Priority Pollutants into 112
Petroleum Refining"
Leon Myers "Generating Problem Pollutants Data for the EGD 120
Document: Refining"
m
-------
A. Korim Ahmed "Considerations for Defining Substances
Hazardous to the Environment"
SESSION V Problems 155
Dwight Bellinger "EPA's Analytical Development Program for 157
Problem Pollutants"
Fred T. Weiss "Fates, Effects and Transport Mechanism of 1°6
Pollutants in the Aquatic Environment"
Donald I. Mount "Measuring Aquatic Impact of Toxic Con- 185
taminants"
Davis L. Ford "Overview of Advanced Treatment Systems" 193
SESSION VI Future Considerations in Biotreatment 215
J. F. Grutsch "Design and Operations: Bases for an Activated 217
Sludge Route to BAT (1983) Water Quality Goals"
Robert W. Griffin "Consideration in Reuse of Refinery Waste- 281
water"
Milton R. Beychok "State-of-the-Art in Sour Water Stripping" 293
Irv Kornfeld "Refinery Discharges to a Large Municipal 304
Sewerage System"
SESSION VII Powdered Activated Carbon 323
Francis L. Robertaccio "Combined Powdered Activated Carbon 325
Treatment-Biological Treatment: Theory
and Results"
Colin G. Grieves "Powdered Activated Carbon Enhancement of 344
Activated Sludge for BAT Refinery Wastewater
Treatment"
Paschal B. Dejohn "Case Histories: Application of PAC in 369
Treating Petroleum Refinery Wastes"
James F. Dehnert "Case History: Use of PAC With a Biodisc- 389
Filtration Process"
SESSION VIII Add-On Granular Activated Carbon 401
Fred M. Pfeffer "Pilot-Scale Effect on Specific Organics 403
Reduction and Common Wastewater Parameters"
R. H. Zanitsch "Granular Carbon Reactivation-State of the Art" 419
L. W. Crame "Activated Sludge Enhancement: A Viable 440
Alternative to Tertiary Carbon Adsorption?"
SESSION IX Costs/Benefits 470
W. Wesley Eckenfelder "Overview of Costs/Benefits" 472
Lial Tischler "Treatment Cost-Effectiveness as a Function of 481
Effluent Quality"
IV
-------
Carl E. Adams, Jr. "The Economics of Managing Refinery 503
Sludges"
Melville Gray "Compliance Monitoring Costs for the 526
Priority Pollutants"
Leo J. Duffy "Analytical Costs in the Problem Pollutants" 531
List of Participants 542
-------
ACKNOWLEDGMENTS
The Second Open Forum on Management of Petroleum Refinery Wastewater was
held at the Sheraton Inn-Skyline East in Tulsa, Oklahoma. This four-day (June 6-9, 1977)
symposium was sponsored by the U.S. EPA in the form of Grant No. R804968-01-0
to the University of Tulsa. Fred M. Pfeffer and Francis S. Manning served as Project
Officer and Project Director respectively. In addition, the American Petroleum
Institute, the National Petroleum Refiners Association and the University of Tulsa con-
tributed matching funds.
The cooperation of the API, the NPRA, and many petroleum industries in suggesting
and engaging speakers and publicizing the Second Open Forum is gratefully acknowledged.
Success of any project is aided immeasurably by support "from the top. " Such
support was provided enthusiastically by Arne E. Gubrud, API; HerbBruch, NPRA; Bill
Galegar, EPA, Ada, Oklahoma; 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 to Katie Whisenhunt, Shirely Clymer, Cathy Whisenhunt,
and Cathy Clymer for managing the registration and typing the proceedings; Ed and Dan
Andrews for tape recording the proceedings; Nelda Whipple for transcribing the tapes and
typing the proceedings; and John Byeseda for photographing the speakers and session
chairman.
vi
-------
WELCOMING REMARKS
William C. Galegar
Director, R.S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
Honored Guests, Ladies and Gentlemen...
As we assemble here for the opening of the Second Open Forum on "Manage-
ment of Petroleum Refinery Wastewaters," I am reminded of a saying attributed to the
French author, Alexandre Dumas. It is ... "Nothing Succeeds like Success!"
In a sense, we will be challenging the truth of that statement during the
next few days. As those who participated in the first Open Forum just over a year
ago will recall, it was a very successful gathering. It clearly established the advantages
of the exchange of technical information among the industry, academic, all levels of
government agencies and the interested public.
The question is, what do we do for an encore? In welcoming you here, I
want to sketch for you some of the efforts made by the hard-working committees putting
the conference together to improve on the earlier product.
We have maintained the firm base on which the first meeting was built —
the co-sponsorship of the Environmental Protection Agency, the American Petroleum
Institute, the National Petroleum Refiners Association and the University of Tulsa.
We have again brought together a group of outstanding speakers, each with
specialized knowledge,.who will maintain the high level of technical excellence and
the presentations last year.
We; have directed our emphasis to a more specific objective, that of
"Environmental Conservation in Petroleum Refinery Wastewater Treatment."
Finally, we have solicited the views of the Natural Resources Defense
Council and the Environmental Defense Fund on the scientific rationale behind the
differences which led to the courts and the consent decree. This, we believe, adds
a new dimension to our forum.
Along with expressing the formal welcome to you.. .many attending for the
second time.. .allow me to include the hope that through your active participation
1
-------
in the sessions we will prove beyond a doubt that Mr. Dumas was correct...
that we will succeed in taking another big step toward our mutual goal of
improving enviornmental quality.
BIOGRAPHY
William C. Galegar
William C. Galegar holds a B.S. in
Chemical Engineering from Oklahoma State
University and an M.S. in Chemical Engine-
ering from the University of Oklahoma.
His career has included 11 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
"WHAT DOES JULY 1, 1977, MEAN NOW?"
Joe G. Moore, Jr.
Head, Graduate Program in Environmental Sciences
The University of Texas at Dallas
It is good to be with you. I spoke to you in your prior conference as
the Program Director of the National Commission on Water Quality, attempting
to predict whether or not the Commission would recommend changes in Public
Law 92-500. The Commission did make such recommendations. A year has
elapsed since those recommendations were delivered to the Congress, and we
now approach one of the critical deadlines in that statute. So that I have
some knowledge of my audience, how many of you are engineers and consultants
to oil companies? I see, mostly that. Alright, how many of you are lawyers?
Any lawyers? Interesting. How many of you are oil company employees as
opposed to consultants involved in water pollution control? Okay. How many
of you will claim to be representatives of the Environmental Protection
Agency besides Bill?
I would like to focus primarily upon the question, "What Does July 1,
1977 Mean Now?", now that it is the first week in June. For some reason,
there is a general feeling that that date is no longer significant since the
deadline is almost gone. How many effluent limitations for best practical
control technology currently available are final four and one-half years
after the passage of Public Law 92-500? How many permits required to have
been issued under P.L. 92-500 are now coming to the end of their five year
terms? What has been the effect of the extensive litigation with regard to
the concepts involved in P.L. 92-500, particularly as related to July 1,
1977, as the answer to that question might bear upon the requirements of
July 1, 1983? You see, this statute is reaching the mid-point in its first
ten year plan, and the question is, "Where are we, and where are we going?"
I would like to review first of all the fundamental concepts that
92-500 requires effluent limitations as the basis for permits under the
NPDES (the National Pollution Discharge Elimination System). Effluent limi-
tations are the basis for permits for every single point source discharge in
the nation. The backbone of the Act is really in sections 301 and 304 which
prescribe the promulgation by the Environmental Protection Agency of "best
practicable control technology currently available" and "best available
technology economically achievable," the effluent limitations upon which
permits are supposed to be based. The first U.S. Supreme Court decision,
du Pont I vs. Train,has been decided with the opinion that the approach of
the Environmental Protection Agency in developing single-number effluent
-------
limitations is valid. The court also bridged the gap between the different
appeals procedure for sections 301 and 304 in such a way as to" validate what
EPA had thus far done with their effluent limitations approach. In addition,
the court held that it is appropriate to issue single-number effluent limi-
•tations as opposed to a range of numbers as had been argued is actually re-
quired by section 304. The effect of this ruling is to validate the ap-
proach used by the EPA in the promulgation of effluent limitations. It was
also held appropriate that the agency issue not only the guidelines under
section 304 but the effluent limitations under section 301 simultaneously.
In the final analysis, EPA has won one of the most crucial issues to the
application of P.L. 92-500 even though their record of litigation in most
cases has not been that outstanding. The other sections that do require
effluent limitations or provide some basis for permitting conditions are
section 306 for new source performance standards, section 307 for toxic and
pretreatment standards, and a section that has been often overlooked, sec-
tion 302 which requires that the Administrator issue effluent limitations
for water quality standards achievement for those waters which have been
characterized by the EPA as water quality limited. To my knowledge, the
Administrator of EPA has issued no effluent limitations for the achievement
of water quality standards described in section 303 with the effluent limi-
tations detailed in section 302. The reason I mention that particular pro-
cedure is because, in my view, that may come to be one of the more critical
provisions of P.L. 92-500 as we move past July 1, 1977.
Let me digress for just a moment, to mention the Congressional issues,
or the legislative issues, with which the Congress has contended since the
National Commission on Water Quality made its recommendations. I shall men-
tion the ones that I regard to be crucial in judging what Congress may do.
First of all, the construction grants program is running out of money; that
is, the construction grants program for municipal wastewater treatment
plants. There are several states which will have, before this calendar year
is ended, committed or obligated all of their shares of the $18 billion
originally authorized in P.L. 92-500. That means that some action either
must be taken by the Congress to provide additional funding for construction
grants under P.L. 92-500, or the publically owned treatment works construc-
tion grant program will grind to a halt in some states. One of the crucial
states in which the construction grant funds will be exhausted is Maine.
Section 404, which allows the Corps of Engineers to issue permits for dis-
posal of dredged materials into navigable waters and adjacent wetlands con-
tinues to be controversial in the Congress, with a wide divergence between
the viewpoints of the Senate Subcommittee of the Environment and Public
Works Committee and the House Public Works Committee; that issue remains
unresolved. On the House side, the Members have expressed continuing inter-
est in decentralization of the program, or what has come to be known as
certification to the states for various activities required under the Act.
I think the House will continue to be interested in that particular issue.
There was an effort to extend the July 1, 1977, deadline so that is an issue
or was an issue in the legislation that has been considered twice by the
Congress in the last 15 months. There is also the continuing question about
what to do with the 1983 deadline and whether or not that deadline should be
extended. These represent what to me are the major issues; probably the two
critical ones are the question of any extension of deadlines and the money
-------
for construction grants. I regard section 404 as a peripheral issue to the
general thrust of the water pollution control program; nevertheless, it does
provide a basis for disagreement between the Houses.
When Congress adjourned in the fall of 1976 there was pending legisla-
tion to amend the pollution control act. It was in conference committee
between the two Houses, and there was rather a stormy session at which there
was substantial disagreement between Congressman Jim Wright, then the ranking
majority member on House Public Works, and Senator Muskie, Chairman of the
Subcommittee on Environmental pollution. At that time, apparently the con-
ferees adjourned and decided not to meet again, or at least did not have an
opportunity or occasion to meet again. The bill was stalled, in any event,
by the filibuster then occurring on the Senate floor over the amendments to
the Clean Air Act of 1970. Thus the congressional session ended with major
disagreements over the general thrust of both air and water pollution control
legislation.
With the change in Administrations and an emphasis on the need to stim-
ulate the economy, the Senate Public Works Committee again, through a device
designed to provide money just for construction grants, precipitated a par-
limentary issue in such a way as to frustrate the legislation amending
92-500. The sequence of events is complicated because the water pollution
control program became intermingled with the question of economic recovery
in such a way that it was sometimes difficult to determine just exactly what
was happening. Senator Muskie succeeded in having added to the Administra-
tion's jobs bill an appropriation to continue the construction grants pro-
gram for a year. Since this issue represents one of the major questions in
the amendment to P.L. 92-500, had that appropriation remained in the jobs
bill, then some of the leverage which the House felt was essential in the
consideration of other amendments would have been lost to the House. When
the jobs bill was returned to the House where it had originated, the appoint-
ment of conferees from the House was delayed. In other words, the jobs bill
which had originated in the House had gone to the Senate where this amend-
ment had been added and then the bill had been returned to the House for the
appointment of conferees. Had the House gone immediately to conference,
they would have been confronted with the situation in which all they could
attempt to do would be to eliminate the $4 billion authorized and appro-
priated for construction grants for publicly owned treatment works. Faced
with that alternative, the House decided that they would proceed in such a
way as to precipitate amendments to 92-500 into the conference on the jobs
bill; thus, the appointment of conferees was delayed until the House passed
a bill amending 92-500 and covering some of the points I mentioned a moment
ago, as well as others, that are in controversy between the two Houses. The
House then appointed conferees with both the jobs bill and the House amend-
ment of 92-500 in the same conference. So the conference began. Initially,
the House proposed that the question of construction grants authorization be
separated from the jobs bill and the jobs bill be agreed upon and sent back
to both Houses for passage and thence to the President. The Senate declined.
After several meetings Senator Muskie, apparently concluding that there was
no chance of the construction grants appropriation surviving in the jobs
bill, finally agreed to the original House offer. The House dropped consid-
eration of any amendments to the jobs bill that would relate to 92-500, and
-------
the Senate accepted the deletion of the appropriation for construction
grants. Where that leaves the water pollution issues is that the House-
passed bill is now theoretically pending in the Senate, but there is a
question as to whether the Senate will actually consider that piece of
legislation.
Senator Muskie called hearings beginning this week in several cities
across the U.S. and extending them throughout the month of June, culminating
in a series of hearings in Washington late in June on the provisions of 92-
500. Unfortunately, the hearing cities do not afford real opportunity for
participation. For example, one in Portland, Maine; one in Greeley, Colo-
rado; one in Duluth and another in Alexandria, Minnesota; one in New Orleans,
and one that was initially scheduled in Berkeley but has now been resched-
uled, I understand, for San Francisco, with one other hearing at a place
listed in the ammouncement as "undetermined," Iowa. Whether these hearings
will produce a great deal of testimony remains to be seen. There will then
be several days of hearings in Washington.
I have reviewed that legislative history, because I want to return to
the fundamental question, "where are we with July 1, 1977?" For soroe^reason
everyone seems to think that we're so near the deadline that it doesn't make
any difference. One of the reasons I asked whether there were lawyers here
is because I think the lawyers are going to be very busy in the month of
July. And if your clients or your employers are not in compliance with per-
mit conditions on July 1, 1977, I would suggest that you call your lawyer's
attention to the fact that they should be prepared to respond to litigation
shortly after July first. I'm in a minority in what I think is likely to
happen, I know. It is very difficult to predict; quite frankly the closer
we get to this date, the more difficult it may be to predict, exactly what
will happen.
Just let me mention some choices that are available to the Adminis-
trator of the EPA. When I talked with Bill Galegar about this appear-
ance, I told him that the title of my remarks was going to be, "What Now,
Brown Cow?" Because the question is, "What now, Mr. Administrator of EPA?"
Of course, one alternative is to do nothing and just let things proceed as
they have been proceeding, with no change. Another possibility is to
selectively institute enforcement proceedings in the courts against indus-
trial dischargers. Pick out the horrible examples and the chances are,
if this is the course that EPA has under consideration, undoubtedly U.S.
Steel will be at the top of the list. There may be some oil companies close
behind. What I'm saying is that the Administrator could selectively choose
against whom the law would be enforced and, in this alternative, institute
no actions against publicly owned treatment works, that is, municipalities.
Everyone sort of shrugs and says, well, what can you do about the cities?
Would it help to put Abraham Beame in jail? It might help Bella Abzug,
but I don't know that would help anyone else. Another alternative would be
to selectively enforce, that is, pick some flagrant Industrial violators
or recalcitrants and institute suit against them; do the same for some
selected municipalities, hopefully municipalities with Republican mayors or
Republican administrations, so that you could get even with those actions
which were instituted under the Republican Administration against some of
-------
the larger cities in the country. Here again, this is a selective alter-
native; the Administrator picks those against whom he would enforce. An-
other alternative would be to continue the procedure of ECSL letters. If
you haven't seen an ECSL letter, or don't know what this means, it is, in
effect, an administrative device for selective extension of the July 1, 1977
deadline - - an administrative device, you will notice. This procedure was
roundly criticized publicly by Senator Muskie and staff of the Senate envi-
ronmental Pollution Subcommittee as being illegal under 92-500 and might now
be characterized as a Republican dodge of the intended effect of the statute.
Quite frankly, I think it's illegal. I do not believe that it has support
in the statute, and, therefore, I think those who go this route may find
difficulty if the courts, as one court already has, hold it an invalid appli-
cation of 92-500. By the way, EPA may still be thinking that the ECSL letter
offers a means of doing something about the July 1, 1977 deadline.
Of course the real alternative that, in my view, is mandated by the
statute is the filing of suits against every point source discharger in the
nation that is now, or will be on July 1st, in violation of the law. Regard-
less of whether it is a practical solution or whether it offers any real
chance in terms of being seriously considered administratively, in my view,
this is what the law requires. The objective would be to secure as rapidly
as possible consent decrees in the courts spelling out the conditions under
which dischargers would comply with the requirements for July 1, 1977.
You should be aware that the EPA is, however, approaching industrial
dischargers with an unusual question as a basis for determining penalties,
"how much money have you saved by not installing 'best practicable control
technology currently available'," or, to put it differently, "what would you
have had to spend had you timely complied with the requirements by July 1,
1977?" This concept of the appropriate penalty is one that was embodied in
the proposed revisions to the Clean Air Act of last year and will undoubt-
edly be, in my view, proposed by the Administration in any recommended
changes to 92-500. The argument is that collecting a penalty equivalent to
what the discharger would have had to spend will somehow restore equity be-
tween those who have complied and those who failed to comply. If you are
now involved in litigation, you may as well tell your lawyers to be pre-
pared - - and you might as well begin to try to figure what the cost would
have been if you had complied with the statute - - because I think this will
be an issue, at least in the discussions of settlement of pending litigation.
The last alternative that I have identified is the Administrator could
ask for Congressional relief. Frankly, this appears to be very unlikely at
this point in time. The controversial issues in 92-500 are interlocked in
such a way that it would be very difficult to solve any single one of them,
as the Senate learned by trying to put construction grants money in the jobs
bill; it is very difficult to take any single one of them and isolate it for
statutory amendment.
?'-.• ; :
How would yoii -like to be faced with the alternatives I've suggested?
For some reason, thei impression has grown that if we can just get past July 1,
1977, it will go away. I don't think so.
-------
There are a series of statutory provisions that are interlocked in such
a way that you cannot plan with very much certainty what may happen in an
individual case. I'm just going to read a sentence or two from the series
of these. Beginning with section 301 - - and sometimes we overlook these
kinds of provisions because they're stated right in the front of the section
and we don't get around to them because we focus upon the substance of a
provision section 301 starts out by saying, "Except as in compliance with
this section and sections 302, 306, 307, 318, 402, and 404 of this Act, the
discharge of any pollutant by any person shall be unlawful." Now, the sec-
tions cited are section 301, the one under which the Administrator prescribes
general effluent limitations for 1977 and 1983; 302 is the section that re-
quires effluent limitations for water quality limited waters; 306 is the
section that prescribes new source performance standards; 307 requires toxic
and pretreatment standards; 318 contains a special provision on aquiculture;
402 is the permit section; and 404 governs the disposal of dredged spoil to
which I referred earlier. I'm going to read the first sentence of 402,
"Except as provided in sections 318 and 404 of this Act, the Administrator
may, after opportunity for public hearing, issue a permit for the discharge
of any pollutant, or combination of pollutants, notwithstanding section 301
(a), upon condition that such discharge will meet either all applicable
requirements under sections 301, 302, 306, 307, 308, and 403 of this Act, or
prior to the taking of necessary implementing actions relating to all such
requirements, such conditions as the Administrator determines are necessary^
to carry out the provisions of this Act." (Emphasis added.) What that pro-
vision says is that if the Administrator for some reason has not done what
he's required to do under those sections 301, 302, 306, 307 and so on
he may require "such conditions as the Administrator determines are necessary
to carry out the provisions of this Act." Since the Administrator has not
prescribed effluent limitations under section 302, and even though the ef-
fluent limitations under section 301 and 304 have not become final - - nor
have those for new source performance standards or toxic pretreatment stand-
ards - - the Administrator may apply such conditions as he deems appropriate
to achieve the objectives of this statute. At the end of that same section
402, there is a provision, subsection (k), which says, "compliance with a
permit issued pursuant to this section shall be deemed compliance, for the
purposes of sections 309 and 505, with sections 301, 302, 306, 307, and 403,
except any standard imposed under section 307 for a toxic pollutant inju-
rious to human health." Thus, if you have a permit that has been issued by
the Administrator or the state acting under delegation of the NPDES author-
ity, then that does comply with the law. The permit contains the bottom-
line conditions. It is true that the permits may expire as late as 1979, so
that your client or company may not be confronted with a permit to be re-
issued this year or during the next year or so. Remember, however, that a
violation of the permit is a violation of the statute, so, if there is any
evidence that a permit is being violated, either as to the completion of the
waste treatment facilities or the actual parameters contained in the permit,
the discharger is subject to enforcement action under the statute.
One of the "hookers" about which no one is certain is the application
of section 505, the first provision of which reads, "Except as provided in
subsection (b) of this section, any citizen may commence a civil action on
8
-------
his own behalf, (1) against any person (including (i) the United States, and
(ii) any other governmental instrumentality or agency to the extent per-
mitted 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 a State with respect to such
standard or limitation, or (2) 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." The effect of that
provision is to say that even if the Administrator of the EPA determines that
he will not, on behalf of the U.S., institute enforcement action against a
violating discharger, any citizen may institute action not only against the
violator but against the Administrator for failure to execute the provisions
of the statute in those cases where he does not have discretion.
I haven't read the statutory provision describing enforcement, but that
provision very clearly says that, when the Administrator becomes aware of a
violation, he shall not, he may but he shall institute proceedings.
At the risk of being proved wrong in less than a month, I say again, you had
better advise your lawyers to be prepared for litigation, if you are in
violation of any part of an existing permit.
I will outline what could be, I think, a realistic chain of events.
First of all, any Administrator subject to confirmation has to commit himself
to enforcing the law. Can you imagine a question in a confirmation hearing
where a Senator asks a prospective administrator, "what do you intend to do
about the July 1, 1977 deadline?" If the Administrator wishes to "pass
muster," he almost has to say, "I will enforce the law." Of course there is
a difference between a confirmation hearing and July 1, 1977, but the issue
will remain on July 1, 1977. The Administrator has taken an oath to enforce
the statute. If the EPA decides that they will selectively enforce, either
against industrial dischargers or against a selected group of industrial and
municipal dischargers, in my view, it will merely be a question of time until
they are compelled to file suit against every discharger in violation of per-
mit conditions. Which is to say, they may begin with the intent that they
will selectively enforce the statute, but it will merely be a question of
time until events will lead them to be compelled to file litigation against
every person they know to be in violation of 92-500. No one could publicly
argue that somehow o-r another it is alright to enforce the law against one
discharger that is in violation and not against another discharger that is
in violation; there would likely be a public outcry. Remember also that it
merely takes one person - - wherever the violating discharges are occurring,
it only takes one person to decide that they will litigate against the dis-
charger and the Administrator simultaneously and, presumably in the same
litigation, if they wish to bring a violating discharger to account. Thus,
you are confronted with that possibility from the outset.
I've concentrated primarily upon what I regard to be the major regula-
tory issues that not only the petroleum industry but also all industries
will face on July 2nd, that is, what do you do about your permit conditions
if you are in violation? Now of course, you can solve this problem by say-
ing we are 100% in compliance and then you don't have to worry about the
possibility of litigation.
-------
I'm supposed to say something about research, so let me just add one or
two words about that topic. First, nobody yet knows how much sludge we are
going to generate or how many residuals we're going to have all across the
U.S. from all these water pollution control efforts. If you're involved in
a research area, sludge is one of the areas on which you should concentrate.
The next major issue, in my view, will be the identification and control of
non point sources. Another is, if we can reduce the volume of water being
used in such a way ss to close the cycle, I think more and more in the future
the emphasis will be on closed cycle systems. Water reuse will continue to
be critical and is highlighted by such episodes as are now being experienced
in California. Probably, the most critical approach to research, however,
is to design production processes to minimize the generation of pollutants.
If you really want to be on the leading edge of the future in water pollution
control, concentrate on the production process rather than the end-of-the-
pipe, or minor modification at the end-of-the-pipe, processes for waste con-
trol. Those firms which manage to design processes that are specifically
intended to minimize pollutants will be the ones that have the business in
the future.
DISCUSSION
J. Dewell - Phillips Petr. Co. - In those states where they have require-
ments more restrictive than EPA where those states do not in themselves have
delegated authority but those restrictive requirements have been written into
the EPA permit, to what extent will EPA act to enforce those more restrictive
state requirements?
Moore - Let me see if I can restate the question as I understand it. In
those states in which there are water quality standards which require strict-
er levels of treatment than the effluent limitations prescribed by EPA and
the conditions are in EPA permits, to what extent is EPA likely to move to
enforce those permit conditions? I'll hazard a guess. I don't think EPA
has yet seriously considered the problems of enforcement generally so I
don't think that they fully appreciate the question yet. My guess would be
that EPA will not move initially to enforce to achieve water quality stand-
ards fixed by the states so much as they're likely to move to enforce a-
gainst those who have violated permit conditions under the effluent limita-
tions of 301 and 304, which is to say that I'm not sure that EPA will assign
very high priority to enforcing water quality standards. You may recall
that in the summer of 1976 the then EPA Administrator went to the Congress
with a recommendation that federal assistance for publicly owned treatment
works for levels of treatment beyond those required for secondary treatment
in 92-500 be withdrawn. That is to say, they were suggesting that the
Federal government's interest should be restricted to a minimum level as
prescribed at the federal level rather than that prescribed by the states.
So, I don't think, offhand, that they'll move to enforce in those areas
where they have water quality standards. Does that answer your question?
10
-------
BIOGRAPHY
Until May of 1976, Joe G. Moore, Jr. was Program Director of the
National Commission on Water Quality, the agency created in the Federal
Water Pollution Control Act Amendments of 1972 to analyze the technological,
economic, social, environmental and institutional imparts of PL 92-500. He
is presently Head, Graduate Program in Environmental Sciences, The University
of Texas at Dallas. Mr. Moore has had some 20 years of governmental service
including Chairman, Texas Water Quality Board, Executive Director of the
Texas Water Development Board, Administrative Assistant to Texas Governors
Connally and Daniel, and Commissioner of the Federal Water Pollution Con-
trol Administration, Department of Interior, directing the national water
pollution control program before the creation of the Environmental Protec-
tion Agency. He is a graduate of The University of Texas at Austin with
bachelor's and master's degrees in government.
11
-------
SESSION I
REGULATORY AND RESEARCH DIRECTION IN U.S. EPA
Chairman
William C. Galegar
Director, R.S. Kerr Environmental Research Laboratory
U.S. EPA Ada, Oklahoma
Speakers
Lamar Miller
"Toxic Strategy Overview "
Marvin L. Wood
"EPA's Research in the Refining Category"
12
-------
"TOXIC STRATEGY OVERVIEW"
W. Lamar Miller
Effluent Guidelines Division
Environmental Protection Agency
Good morning, it is very nice to be here today. I am not sure whether I am a
designated hitter or a pinchhitter. I guess the effect is really the same. Mr. Schaffer
does indeed send his apologies for being unable to be here. He is participating in a
command performance this morning and tomorrow on some of the very topics which Joe
Moore has just been discussing with you. Most specifically, the agency's suggested input
for midcourse corrections which the agency would like to recommend to the Congress in
the very near future, as they relate to P.L. 92-500. I think that may be a sufficiently
good excuse for his inability to be here today. He did want to be here. I'll try to do my
best this morning to try to tell you the same story that he would tell you if he were here.
I'll exempt most of the humorous remarks from that qualification. I don't think that Bob
would use the same stories that I would tell you, but the serious part of the speech will be
close.
I'd make one observation on a remark that the keynote speaker just made. It
somewhat puts me in mind, Bill, of a cartoon that I keep pinned to my wall to remind me
of our actions. Joe Moore's remarks about the approach used to evaluate the cost for
implementation of BPT. The cartoon on my wall is two cavemen talking to one another.
One says to the other in a very philosophical note, "Our future is shaped by our past, so
be careful what you do in your past."
Be mindful that the estimated cost for the installation may also become the
amount of the fine if the installation is not made. Very soon after the Supreme Court
decision was announced I called up a friend in industry and said, "congratulations Bob,
your corporation's name will go down in history as the first major loss in the water fight.
That case will be known forever as the du Pont case. And, by the way, what did industry
win on the appeal? Hadn't you really rather that you had lost some of them and there-
fore kept the old rules instead of the directions that you hope I'll tell you about today?"
I think we ought to keep some of these thoughts in mind as reminders and go back and
look at the 1976 meeting out here at Tulsa. I looked at some of the remarks which were
made by the keynote speaker last year, as a matter of fact, Bill Galegar said that
participation and communication made for a successful meeting; and, he asked this
morning, how are we going to improve on that success. The keynote speaker last year
said that effective communication is a two-way street and cited Joe Moore on effective
communication in saying, "we must not only listen we must hear what is said." And then
a number of speakers proceeded to tell everyone at the meeting that comments should be
solicited from industry. Industry should be heard. EPA should listen to the industry.
EPA should know industry's viewpoint, then reminded the audience of what the courts had
told EPA about the errors of their direction at that time.
13
-------
Folks, at this meeting, this morning, I would like to point out to you that Mr.
Moore was right about communication. We must not only listen but we must hear what is
being said. Mr. Beychok was also right; that communication is a two-way street; but it
needs to be pointed out again; it really is a two-way street. We have to listen to you.
We want to listen to you. We're trying to listen to you. And we're scheduling a lot of
time to sit around and talk to you and listen to you. But, it is a two-way street. While
we're listening, please hear what we're saying to you. It works both ways. Nobody in
this audience is any more tired, and I'm glad Joe asked how many lawyers were here. I
didn't see but one hand go up, I might have been wrong, maybe there are more, but we
are all technical people, let's think that way for a minute. None of the technical
people in this room are any more tired of the adjudicatory processes than are the technical
people at EPA. And the people in the effluent guidelines division are the people who are
catching the bulk of that work effort. We don't encourage the adjudicatory system; we
don't want it; and, the directions are sufficiently different at this time that I don't think.
it is going to be as necessary. Communication, and every time I think about the topic I
am reminded of a good Polish joke, a good Polish joke is a joke told to me by Polish^
people, I'm not one of the government employees who wants to get hung by the ethnic
joke problem. I've no interest in telling that type of joke. But, it does reflect both on
politics and is a reasonable observation of the problems of communication. It seems that
in that great hall up in the sky, once upon a time, Lenin, the Czar of Russia and
Napoleon were reviewing history as they saw the problems that they had in their political
life and the problems in Poland today. Poland's problems were very much in a political
discussion last fall if you remember, Lenin observed that if he had had the ability to
control communication and what people understood as a result of mass communication the
same as the First Secretary of the Party in power during the 1960's uprisings in Poland,
that he would have won the revolution in Russia in 1908, he wouldn't have had to wait
until 1918. The Czar responded that if he had had the ability to control communications
to the people the same as he had observed had been happening today in parts of the
world, that no one would ever have heard of Lenin, and that he would still be the Czar.
Napoleon responded that if he had the ability to control communications and what people
heard and what people understood as was done by the Polish First Secretary, that he would
be convinced today, that after all, he had, in fact, won at Waterloo. I sometimes think
that most of us hear and most of us listen to just about what we want to hear, and that
means both EPA and industry. We're really trying to change that in the effluent guide-
lines division. We're really trying to change it at EPA. Remember that if we both go
that extra mile it will make it a lot easier to make it a two-way street. We've got to
both make the extra effort to listen and to talk. I believe that this group, represented by
refiners and people associated with them, can really make that possible. I don't believe
that it is necessary for either of us to prove that we are willing to go to court, we've
already done that. I hope that it isn't necessary for us to prove to each other that we are
failing by having to go to court. We've both done it enough times to know that we're
willing to go back if necessary but I hope that we can avoid it by learning that it is
unnecessary. Technically we can resolve our next series of issues. We're both going to
have to learn to listen to both sides and we're going to have to both learn to go both ways
on a two-way street. Let's not spend all of our time worrying any more about categori-
zation, let's not spend it worrying about the semanic differences between limitations and
guidelines. It was very interesting to me to hear Joe tell you this morning, that's settled.
As I understand it, it is settled. The problem is, and our direction is, (this is very
important to us and this is literally the most important thing in terms of the costs and
economics of the effluent guidelines division right now), is to determine how best to do
the regulations. By spending our money, and yours, in such a manner as to derive the
greatest benefit for the environment for every dollar that is spent by us or by you. But
14
-------
we're not interested In spending money for the sake of spending money and we're not
interested in deriving regulations for the sake of deriving regulations. We are interested
in implementing the law as we understand it. It was once said by a famous English
politician, Gladstone, the only thing "worse than a politician who thinks he has the
truth up his sleeve is one who thinks God put it there." I want you to know that that
particular observation is to me somewhat like the best advice that my father gave me
(since Bill digressed into the good old boy storeis I think I will too). Dad looked at me
once when I was just ready to finish college and he said, "son never argue with a fool
in public, passersby can't tell the difference." I didn't recognize for twenty years that
my father's remarks were really in fact a paraphrase of Justice Oliver Wendell Holmes'
remarks published as an essay which he entitled, (I think fittingly for this particular
group), "The Hydrostatic Paradox of Controversy", in which Mr. Holmes starts off by
reminding us that if we had two pools of water, one the size of a soup bowl and one the
size of an ocean, connected by a pipe stem, the levels of the water in each would be
the same. This is very similar to the status of both wise men and fools engaged in contro-
versy. The major difference being, the fools know it. The same is true of most of us;
when we all get reduced to a common level of thinking that we have a noble exercise to
go about, and no one can challenge or have any interest in adding to what we do. Now,
you can put the last remark in perspective. I'll tell one more joke, current from the
Washington scene about five men on an airplane. The pilot comes back and he says,
gentlemen I have put this plane on automatic, I've come back to tell you personally that
we have a real problem. This plane is not going to make it to the airport, and we're too
far from any other place to satisfactorily land it. I know nothing else to do but tell you
that we have four parachutes on board, and, unfortunately, there are five of us. You
might think, as a general rule, since I am the captain of this ship, that I would give you
each a parachute and allow you to jump while I stay behind with the plane. That, hoW-
ever, is not to be. I think that you should know that I work for the CIA and I have some
very important information up in my mind that must get to the president tonight. So with
that I am going to take my chute and jump out the door. I'll leave the problem to you.
Another guy stood up and said; "gentlemen you may not think it's very important, I'm
only an ex-president but it's very important that I survive because I still give continuous
advice to the new administration, and I think that none of you would deny that I ought
to have one of those chutes." With that he straps on a chute and jumps out the door.
Another fellow stands up and he says; "I have been the Secretary of State and undoubtedly
have been recognized overwhelmingly as the outstanding intellect of the world. I must
be rescued for the benefit of all the world. My very being is very important; and with
that I will leave"; and he jumped out of the airplane. There was left a hippie student
and a minister. The student looks at the minister and he says; "father, I think that you
should take the remaining chute." The minister looks at him and says, "no, son, you
take it, you take it." The student looks around and he stirs around a little bit and says,
"well I don't really think that we need to engage in this conversation." The minister
looks at him and says, no, this is my life's work, to serve my fellowman, this is all I
have to do, to make other people happy, you should take it and leave and I must make
the sacrifice. I've had a long and healthy life and I've enjoyed it. You go and you
save yourself." And then the boy looked back at him and he says; "but father, it isn't
necessary for either one of us to make this decision; you see, the smartest man in the
world just jumped out the door with my knapsack." I want you to know that the effluent
guidelines division of EPA does not take the attitude that we are necessarily the smartest
people in the world. We are neither going to jump out of that door with a knapsack,
nor are we going to drive it straight into a wall. Joe pointed out, and I think well,
that our litigation record was not too good. He's right. We intend that it's going to
get a whole lot better. The reason for that is that we don't intend to issue regulations
15
-------
that we can't sustain. And If we can sustain them, I think you will probably be reason-
ably ready to agree with them. We may not have to have too much litigation^in the
effluent guidelines area. Now, to give you a reasonable overview of our toxics
strategy, which is presumably what some of you wanted to hear from me.
On June 7, 1976, we entered into a settlement agreement which settled
four cases brought against the agency by environmental groups. These cases were
brought against the agency to force us to expand the list of toxic substances under
section 307, to promulgate final standards for toxic substances already designated under
section 307, and to finalize 8 pretreatment standards. I'm not going to discuss at length
the details of the settlement agreement in as much as at least two other times during this
conference it will be discussed by people who are more competent than me to discuss it.
The materials that are dealt with under section 307a for which the final standards must
be promulgated are pretty well known to most of you; aldrin, dieldrin, benzidene,
endrin, DDT, DDE, DDD. The other materials which are identified in the settlement
agreement have come to be known as the list of 65. Sometimes also referred to as the
list of 129. Sometimes also known as the list of 543. Sometimes referred to as the list
of 43,000. We will specifically delineate later in the meeting for you by another
speaker from our division, exactly what the list really means to this industry. IHs our
attempt to try to focus your interest on specific materials. It's not going to continually
change. We have identified 129 today. We hope that they will be satisfactory to do
the job. The effluent limitations will be established on highly specific, technology
based, limitations for 21 industrial categories identified in the agreement. At the risk
of confusing this particular issue, also the 21 are now in fact 28, because miscellaneous
chemicals which is one of the 21 is made up of 8 different categories. It's no longer 28,
because one of them as of last week is settled and will be no discharge. The carbon
black industry. That puts an end to one of the 28. The effluent limitations will be
promulgated in the form of best available treatment economically achieveable for
July 1, 1983. This particular work activity has become known as the BAT revision in
the terminology of those developing effluent guidelines. We also sometimes refer to it
as BATox, or BAT combined with toxic controls or priority pollutants. You may hear
that term in the future. What that really means is best available treatment with priority
pollutant control. These standards will also include pretreatment for existing sources,
pretreatment for new sources and new source performance standards. The settlement
agreement reflects the agency's strategy for handling toxic materials and other problem
pollutants discharged into the nation's waterways. A number of control options are
available to the agency and standards and regulations will be developed taking many of
these options into account. The first step will be to establish technology based
effluent limitations in a similar fashion as we did for previous effluent limitations, we
don't expect to put any out that are not technically sustainable. If it is determined that
these technology based effluent limitations are not sufficient to control the problem
pollutants, the agency has the option of designating a particular material for coverage
under section 307 and will not hesitate to do so. Under this section more stringent
standards may be established without the detailed economic considerations as required
in establishing effluent guidelines under sections 301, 304, and so forth. There is also
the option of developing and upgrading existing water quality standards to resolve
potential problems with specific materials of concern. You heard Joe refer to a little
used section of the Act this morning. That particular sentence refers to the same little
used section. It is an option available to us for the control of toxic materials.
Specifically, the use of water quality standards to resolve specific problem areas. Our
studies will proceed along slightly different lines than they have proceeded in the past.
In fact the studies will proceed along lines which reflect the organization of the Office
16
-------
of Water Planning and Standards. We have four divisions and we have four study areas
ongoing at the moment. Technical contractors will establish alternate levels of
technology; and that is an important, perhaps new, consideration. -We expect to have
more than one level of possible treatment technology designated in thfe development of
the regulations. These studies will look at the specific pollutants, i.e., the 129 priority
pollutants and the traditional pollutants. Secondly, economic impact analysis as was
done in the past will be a major factor and a major input to the overall study. Thirdly,
studies to provide facts with regard to the impact of the designated priority pollutants on
the aquatic environment will be conducted. Fourth, a study of the environmental mass
balance of each of the materials will be conducted. This effort will look at the manu-
facture of the material, the movement of the material into various products and ultimately
into the aquatic environment, and the source of the materials which do move into the
aquatic environment. Each of the above factors will be weighed in determining the
ultimate choice of technology and/or control options for preparation of regulations, not
just technology and not just economics, but a mass balance and the effect on the aquatic
environment. It is our intent, because of the complexities of all of these studies, to
begin by screening specific industrial discharges and identified subcategories to deter-
mine the presence or absence of the designated priority pollutants. I made a short
presentation to one committee of API, not very long ago, and within two days the Deputy
Assistant Administrator got a telephone call that said; that I said at the meeting, we're
going to have 129 regulations for every refinery. I don't know what I said that led the
man to believe that, but I want to make sure that you don't go away believing that. The
first thing that we are going in our technology based studies is to determine how many of
the 129 are in fact a problem, and how many are not present in significant quantities,
i.e., how many can we eliminate from being a concern from a particular industry or a
particular category. This, hopefully, will allow us to narrow down the number of
materials with which we have to be concerned in developing treatment technology. Once
this has been done, we will proceed to quantify on a more extensive basis the significance
of those materials that we found to be present, that is, how well does existing treatment
handle those that are present and how many of those are no longer a concern. There will
be some cases, for some categories, for some processes, where BAT is equal to BPT for
one of several reasons. Either priority pollutants are not present, or the existing treat-
ment renders them dischargeable in insignificant quantities, or they do not constitute a
major source to the total aquatic environment, or mass balance studies indicates that
the point sources under study do not constitute a major portion of the environmental
impact of the pollutants. There are many reasons why any of these may not cause you a
problem. Once we have quantified and made the determination that a material is
present, and that it is present in an environmental significant quantity, then and only
then will our efforts be concentrated on the selection of appropriate control strategies.
As we proceed to develop these standards we will be opening up the process and will be
inviting more active participation of the industrial community. We will again be
soliciting data, and their assistance in areas of technology and its application to the
problem. As a matter of fact, we have already begun soliciting assistance in technology
application. We intend to let everyone know as best we can what we're doing and what
levels of technology are under consideration at all times. We're going to play the entire
scenario right on top of the table in the public view for all interested parties. We intend
to use fully the authority that we have been granted under section 308 of the Act to gain
information with regard to the various industrial categories on the specific materials of
concern. We're also working very closely with industrial associations, not only this one,
but others, in developing working groups, particularly in the area of analytical methods
development and technology assessment. Of paramount importance to us is orderly and
timely progress. In reading the settlement agreement one will notice that there are
17
-------
specific dates by which time certain portions of the job must be completed. The
schedules are very tight. We will not have the time to provide leisurely reviews. Since
we are opening up the process, soliciting cooperation and allowing everyone to know
what we're thinking as we proceed, we feel that we can eliminate long review periods
as well as a lot of litigation. Everyone is concerned with the discharge of problem or
toxic pollutants and as good citizens everyone wants to control them to the maximum
degree possible. We believe that as long as we proceed in a rational, cost effective
manner, we will make progress together. We believe this to be the proper strategy for
the control of the problem.
DISCUSSIONS
Pete Foley, Mobil Oil Co. - Can you give us an example of how economics will play a
role in choosing the appropriate technology?
Miller - Yes, I can give you an example. Let me just try to describe a very specific
case. Our Analysis and Evaluation Division will take the cost data associated with each
technology level and evaluate it. There may be for instance a number of technology
levels and an associated cost with each one. Obviously one level is no treatment.
Another level may be, for instance, just an API separator. Another level may be an
API separator followed by an extended aeration activated sludge system. Another level
may be a bioplant combined with a filter, or combined with a carbon column within the
process, or a carbon column at the end of the bioprocess. Each of those levels will be
costed. Then the impact on the industry and the economy as a whole will be determined
prior to determining the trade off or the benefit which we think is most cost effective.
We will do this for a number of treatment levels for each regulatory action.
Milt Beychok, Consultant - I was interested in hearing you say that EPA would again use
contractors to gather a great deal of the background information. They do that, and
whether or not the EPA later, as they become more knowledgeable in the field, depends
to a great extent on those contract reports I'm not sure. But they]re at least preceived
by those of us in the field as being a data base that must be contested. Industry often
ends up hiring their own contractor to prepare a rebuttal to the first contractor or to
prepare cost estimates to rebutt your cost estimates. Has EPA ever considered opening
up the process at the front end and gaining the review and approval of all interested
parties in the selection of that first contract?
Miller - I'm not sure I really understand the question. I think that the answer is, no. I
don't think that our administrative procedure allows the government to allow someone
else to determine how they will contract for their business. There are some pretty
stringent rules on what you go through to select contractors. Let me make one remark in
regard to what you said about industry's general feeling or your general feeling about
contractors. Let it be clear, at least from this day forward, no contractor writes
regulations. I've seen that suggested in the past. The policy is pretty specific. All
regulatory actions will be the specific responsibility of the appropriate project office
and his supervision. The first level of regulatory recommendation we also changed as
you will hear in a subsequent presentation by Bobby Del linger. The system by which
we review various levels of technologies, the costs, the associated economic review,
the mass balance in the environment, and the rest of it has been put into what we call
a working group. This group has some highly specific responsibilities. Bobby will dis-
cuss those with you in more detail, but let me just tell you that it's not going to be the
work of a contractor that puts out a draft regulation for you to have to worry about.
18
-------
It's going to be very easy for you to point a finger and say, this man is responsible for
this exercise. We subscribe to that policy and it is the state policy of the new adminis-
tration. You ought to be able to identify who it is and we're going to help you know
who it is. It will not be a hidden contractor as it has been sometimes alleged to be in the
past.
Milt Beychok, Consultant - I appreciate that, and I wasn't suggesting hidden contractors.
I also appreciate that no one can participate in your selection of the final contractor in
terms of the contract terms of price or scheduling. It might be well for the EPA to
solicit from industry whom they at least consider competent enough to be on the original
bidding list and who really does have experience in that particular industrial category.
That was really all that I was talking of.
Miller - I have been in that position before; most specifically in the inorganic chemicals
and the organic chemicals industries* Where industrial representatives said, "why don't
you ask us who you should hire?" Which is effectively what you have said. I looked at
the particular company that made the remark and asked, "is your engineering division
available?" "We know they are competent, we know they know how to design these
processes, will you bid?" And the answer was, "no." It's not always that we don't get
the people you want for us to have because we don't try. Mostly, it's because they don't
want to get into government audit problems. The most recent case that comes to my mind
is an extensive carbon column evaluation study where I asked a lot of people, everyone I
know, who was qualified, please bid this time, please quit criticising us for the quality
of our contractors when you won't even bid. And the bid list was not very long, and it
was for quite a number of dollars. Yes, we have trouble getting good contractors; but
the contractors we have, if we don't hold them to the line, don't blame it on the
contractor. It's our fault. We have a job to supervise them and if we don't do it, it's
our division's responsibility to come up and say, it's our ball. And we will this time. I
think the one we have on this project is doing a very good job right now.
BIOGRAPHY W. Lamar Miller
W. Lamar Miller is the Chief of the
Organic Chemicals Branch of the Effluent Guide-
lines Division. He has a B.S. degree from
Mississippi College and an M.S. and Ph.D. in
Chemistry with minors in Engineering from the
University of Florida. He spent twelve years in
teaching and industrial work prior to joining the
Environmental Protection Agency in 1972.
19
-------
EPA'S RESEARCH IN THE REFINING CATEGORY
Marvin L. Wood, Chief
Source Management Branch
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820
Actually, this is going to be more of an historical reminiscence than a
current commentary on research the EPA is doing on refinery wastewater. For
one thing, we haven't done any research since November; we've been working
for Effluent Guidelines on a technical assistance project. Secondly, all
the projects that we have any current interest in are listed later on in the
program; thirdly, rather than chance saying something that would be corrected
by a person who actually knows what they are talking about, I'm just going
back in history and bring you up-to-date, partly from a philosophical angle,
about what went on in the formulation of the research program and some of the
results of our earlier projects. Actually, the reminiscing may be apropos
since a Harvardian, a fellow by the name of Santayana, once observed that
those who do not learn from the lessons of history are condemned to repeat
them. Do we have the time or money to afford to repeat some of the lessons
of history that we should have learned already, particularly in wastewater
treatment and control in this specific industrial category?
For those who expected a more western philosophy here in the oil capital
of, at least, the United States: there was once an Okie (not from Muskogee,
but a little closer to here—the town of Claremore) who observed that all he
knew was what he read in the newspapers, and he became rather famous for his
comments based on that observation. Wouldn't Will Rogers have had a field day
reading some of today's headlines in the newspapers? Especially with regard
to aspects to the energy "crunch," "crisis," or whatever tag you want on it.
Depending on which expert is being quoted, we may expect within the next one
or two years to buy our gasoline in small goatskins, or at the other extreme,
we might still be driving our gas guzzling V-12's by the year 2077 at a modest
increase of X dollars per gallon for fuel. The truth, as usual, probably lies
somewhere in between.
Where there are problems, there are ways to solve them. A technology as
advanced as ours in a country that lives, moves, clothes itself, and produces
its food from liquid hydrocarbon products can afford to give up this form of
energy. We are not then in the position of holding a wake rather than forum
or pondering over, for example, the wastewater treatment problems associated
with buggywhip factories. Regardless of whatever source these liquid hydro-
carbons are to come from in the future, they're going to have to be refined
and their changing characteristics will cause new wastewater treatment prob-
lems. I hope to cite herein some of the lessons of history so that we may
avoid being condemned to repeat them in our future research and our future
20
-------
wastewater treatment and control efforts. With respect to the soothsayers of
doom or Utopia, some of the fairly recent economic projections were that there
would be no shortage of crude because shale oil would be available in large
quantities when the price of crude reached five dollars a barrel. Crude is
not cheap, not plentiful, and not coming from shale. So much for soothsayers
and oracles—one would almost be tempted to put them into a mythological
category with demons, ogres, virgins, and dragons.
What is known at present, for example, is that there is more oil in known
locations than we have consumed since Colonel Drake drilled his first well in
Pennsylvania. Admittedly, there are problems in recovering this oil, since
it has been subjected to primary production methods, and in some cases,
secondary production methods. It is difficult to believe that a nation can
afford to mine rocks on the moon but cannot develop technology to extract a
little more of the crude that we know is there (and to a goodly degree, how
much is there). Hopefully, we should be able to extract enough to give us
some latitude until adequate stocks are available from exploration and new
production, coal conversion or shale oil or even such "way out" things as
hydrocarbon production from sunlight and cellulosic materials to make a new
(and continuously renewable) source available. Admittedly, everything is
finite and so are these potential sources of the "new crude." Nevertheless,
until the little handy-dandy portable nuclear fusion energy generator comes
along, we're going to have to learn to utilize them, and utilize them in a
more wise fashion than we have in the past. America does live on liquid
hydrocarbons; we have for seven decades, since the automobile and internal
combustion engines were developed; and we are going to continue to, because
we cannot replace overnight the transportation network that moves our fuel,
our food, our fiber, our raw materials, and our finished products.
The situation does suggest that any new treatment options be regarded
carefully since changing feedstocks are well known to have changing effects
on wastewater effluent parameters (physical as well as chemical and biological)
and the treatability there of. The answers to today's problems, as we see
them now, are certainly not expected to be absolute. Until Woehler's syn-
thesis of urea, chemists regarded organic chemicals as incapable of production
by man—they were produced only by nature. Until 1903, the oil industry
"knew" that there wasn't any market for aviation gas or jet fuel—there were
no airplanes—how could there be a market? Until relatively recently, some of
us who call ourselves sanitary engineers—and before I offend more people in
the audience, I'm registered as one—anyway, some sanitary engineers just
"knew" that the Imhoff tank or Emscher tank which removed the big pieces was
going to solve water pollution problems. Then when biological treatment was
developed, we "knew" that it would solve all the effluent treatment problems.
We have been remiss in being so proud of the percentage removal figures; it's
not what we took out that counts, it's what we left going back into the
environment that has hurt us in the past.
If I may leave general philosophy, we can, with one dirty word, start
concentrating on the background of refining wastewater research done at the
Robert S. Kerr Environmental Research Laboratory (RSKERL). The dirty word
is "matrix"; that word was "out" not too long ago, maybe it will be "in"
again by tomorrow; but it fits this situation. If you will consider the
21
-------
number of control and treatment options that are available for effluents in
general, you will find that it presents to a rather complicated listing.
For example, select a number of simple headings for a series of columns—such
things as chemical, physical, biological, and reuse/recycle. Now picture any-
one of these headings and list the number of past and present treatment options
that are available. Let's pick biological. Most of us are principally con-
cerned with nominally aerobic processes, but there are also anaerobic pro-
cesses, and in addition, there are some facultative treatment schemes which
in nature work both aerobically and anaerobically. How many nominally aerobic
processes can we list under one of the three subcategories? How about slow
sand filters, high-rate trickling filters, low-rate trickling filters, contact
beds, Hayes process, activated sludge, extended aeration, Passvier ditches, and
a few more? The matrix, for that is what we have, depending on how closely
one defines each simple single unit process or unit operation, can extend
to nearly 100 possibilities. The combinations and permutations possible for
a treatment plant are somewhat horrific to contemplate, especially if operat-
ing from a zero based budget. Incidentally, a few people have observed that
insofar as research on treatment of refinery wastewaters, we have been operat-
ing from a zero base budget for the last eight years, so anything would be an
improvement—I didn't say that; I'm just repeating it.
One of the most apparent, logical, rational approaches in winnowing out
the more promising treatment options for any given effluent is the good old
state-of-the-art approach. Investigate what has been done; compare the results
to what is being done; what has been dropped; what appears capable of refine-
ment; and concentrate on the most promising. Admittedly, a state-of-the-art
project is highly attractive to many of us bureacrats; after all, it costs
relatively little money; it seldom fails to yield a nice thick pile of paper
as a report, and it cannot be attacked technically except on the grounds of
illiteracy, if the author has failed to research thoroughly all previous
references and include them in his document. Rest assured that the state-of-
the-art, which initiated research in refinery wastewaters at the Kerr Labora-
tory was not predicated on this "safety first basis," but the fact that we
had very little money, a small staff, and as it appeared from the recommenda-
tions of the finished document, a lot to do.
It is generally conceded that it is wise to be aware of the physical,
chemical, and biological characteristics of any wastewater which you wish to
treat. Since treatment plants involve a sizeable investment of time, money,
construction materials, and possible untoward affects downstream (should they
fail to perform as designed), a companion report to the state-of-the-art was
prepared to delineate wastewater characteristics of the various unit processes
and unit operations which existed in a modern refinery, and which, together,
represented the composite wastewater stream. Actually, it was not until
several years after these two reports were published that we obtained a good
statistical basis with respect to only the most common physical and chemical
parameters for the composite stream. This basis was developed during the
EPA/API Wastewater Characterization Survey which happened to be another
Effluent Guidelines Division technical assistance project. The present EPA/
API study for the Effluent Guidelines Division will develop similar information
with respect to many, but not all, of the organic chemicals and their charac-
teristics in refinery wastewaters. No such study, however, is currently
22
-------
planned with respect to the characteristics of individual unit process or
unit operation effluents. Yet, we know that they vary widely, on the basis
of singular or limited observations, from one process to another, from one
unit operation to another. Since it's generally easier to find the needle in
a bale of hay rather than the entire stack, I would commend to your reflection
that this is a fertile area for some future research.
From these two cited documents, plus some black magic or legerdemain
known as a planning system, we developed an in-house document which laid out
on a PERT-type chart our research program in terms of where we had to go and
how to get there—a plan for obtaining the most widely usable results for our
limited research dollars. Although barred from including air pollution con-
trol and solid waste control research, the program document itself did recognize
the obvious inter-relationships between the three media and did attempt at
all times to prevent creating a new problem in the course of finding a solution
to an existing problem. We did this without the usual standing jokes about,
"what the hell, let's airstrip it and let the air polluters worry about it."
Built into the research program, and continuously upgraded as technology and
techniques allowed, was the concept of quality control from initial research
project design through the sampling procedures through the analytical proce-
dures to the data evaluation procedures. Those of you who have been assoc-
iated with projects which EPA has funded are aware of EPA's analytical and
quality control manuals—the "CuSum" statistical quality control procedure in
the latter manual was developed originally at the Kerr Laboratory.
It is germaine to our forum to review the findings of the original state-
of-the-art report and compare them to the projects which have been developed
since the start of the program in 1968, nearly one decade ago. Particularly
of interest, are those recommendations which were not acted upon through lack
of applicant, lack of funding, or lack of analytical and engineering technology.
Some are specifically pertinent to the potentially painful problems that we
are facing now regarding the "priority pollutants." With your permission I
will read these 20 recommendations, instead of depending on memory, because
some of them are so beautiful in their simplicity and yet farsightedness:
"1. Develop sampler to obtain representative samples of floating oils,
dissolved oils, emulsfied oils, and oily sludges."
Has anyone yet developed a sampler that would do all that? I think not.
There has been progress but a representative sample still remains one of the
most difficult and probably one of the impossible things to obtain from any
wastewater stream.
"2. Conduct internal refinery studies to reduce waste volumes and
strengths for old and new refineries."
Back to unit process and unit operation waste streams. Has anyone done
any definitive study of these? In a few instances, that I am aware of, yes.
We are beginning to approach the area of this recommendation, but not exten-
sively.
23
-------
"3. Extend biosystem studies to optimize treatment efficiencies and
handle shock loading."
Admittedly, there has been a great deal of work on optimizing biosystems
firom an operational standpoint. A project in an allied field of research may
have developed a monitor that, while not eliminating shock loading, will at
least give adequate forewarning to divert and hold spills or shock loadings
and toxicants that might upset a biological system until they can be handled
separately, diluted, or bled back through the system at an acceptable concen-
tration.
"4. Devise a continuous monitor for hydrocarbon detection in wastewaters
using common refinery laboratory equipment."
One of our laboratory scientists has developed a procedure using a gas
chromatograph, which is now common equipment in most laboratories, to determine
hydrocarbons in wastewater. As opposed to the freon, or petroleum ether, or
carbon-tetrachloride or hexane or whatever your solvent preference is for oil
and grease extraction procedures, this method appears to have repeatability
and reproducability characteristics.
"5. Design original wastewater treatment systems for the petroleum
industry."
In the case of new grassroots refineries, scarce as they are, this is
being accomplished.
"6. Perform chronic (long term) toxicity studies on treated effluents."
Our Laboratory in conjunction with Dr. Mount's Laboratory in Duluth,
Minnesota, has started some screening studies in this area. Additionally,
some work has been done at Oklahoma State University.
"7. Identification of toxic components in petroleum wastewaters."
Refer to the forum program for further information.
"8. Develop efficient devices and techniques to remove oil spills on
diverse waters surfaces (i.e., swamps, rivers, and turbulent seas)."
One of Murphy's laws; if you're going to handle the stuff, pump it, or
process it, sooner or later you're going to spill some of it. A great deal of
money has been expended in developing spill clean-up equipment, procedures,
processes, and techniques. But a total answer approaching a respectable
efficiency for all problems of all types of spills is still not within our
reach.
"9. Assess environmental effects of spilled oils (i.e., volatile,
soluble, emulsified, floating, etc.,) and oil products."
My God, that's many life times of work for many biologists in one simple
statement.
24
-------
"10. Investigate use of cooling towers for treating selected refinery
wastewaters for recycling."
How much has been done? Well, some—we're far away from covering all
aspects though, nearly a decade later.
"11. Study water reuse within refinery."
Now certainly a good deal has been accomplished in this area. Not all
of it in the last 10 years—some of it was done prior to that time—and yet
there have been significant accomplishments in the last decade.
."12. Explore feasibility of phenol removal from wastewaters using
phenol-soluble oils."
There is more than one plant in the country today that is extracting phenol
using solvent extraction processes.
"13. Perform economic studies of brine treatment and disposal on land
and sea."
This is still a very sticky problem from the standpoint of what do you
do with it regardless of how you manipulate inorganic salts that constitute
brine,.whether they're from a producing well or a desalter, you still have
the salt left over, and you can't burn it, you can't eat it, you can't sell
it, you can't use it, and you can't dump it—unless you are lucky enough to
be on the coast. Most of the brines approximate seawater; admittedly indi-
vidual component concentrations range widely, but the components do approximate
seawater in terms of numbers and types of elements.
"14. Develop remote sensing techniques for detection of oil and brine
pollution."
Supposedly, the CIA can read a license plate from 400 miles out in orbit
via satellite, but nothing approaches that degree of precision and accuracy
with respect to this simple little recommendation of 10 years ago.
"15. Perform feasibility studies on by-product recovery from refinery
wastes."
The phenol recovery technique referred to earlier is, in fact, one of
these by-products recovery operations in actual practice.
"16. Devise a monitoring program to prevent subsurface pollution from
abandoned oil wells."
There has been a considerable expenditure of EPA funds on a GE project
called Tempo having to do with monitoring techniques of groundwater sources
and the presence, direction of flow, and source of groundwater pollution.
Yet, there remains a great deal to be done to put these into effect with
respect to individual plant areas; tank bottoms fail and buried transfer
lines will leak.
25
-------
"17. Examine pollution problems associated with extreme cold in the
Alaskan oil fields."
There's one where they really put some money into. Considerably more
apparently than any of these other 20 recommendations.
"18. Determine proper measures to collect and reuse waste oils from
U.S. vehicular and boat service stations."
This is something that has probably been beaten to death in terms of
rerefining of lube oils; nevertheless, from the amount produced each year
versus the amount that is used, even in cars that are as smoky as mine there
are volumes of lubricating oil that are not reclaimed and not reused—that
are dumped in places damaging to the environment.
"19. Design antipollution devices and management controls to insure
proper storage of crude oil."
As mentioned earlier, those tank bottoms will fail, transfer lines will
leak, and hoses will rupture. I cannot say that there hasn't been a great
deal of progress in 10 years in this particular area; but these excursions
still occur from time to time with great fanfare and headline publicity
usually connected with a tanker loss. Finally, this is the real "catch 22"
that will occupy much of your time this week—
"20. Assess toxicological aspects to man and to warm-blooded animals
ingesting oil and oily substances."
That was a rather obvious one, wasn't it. Looking back now it's rather
obvious to see the wisdom in that recommendation, but we're just now starting
to scratch the surface of it in reality. Admittedly, some of these may have
been limited by the lack of analytical equipment. Some 20 or 25 years ago,
the mass spectrometer was the answer to "on line" process control equipment
for refineries, but that didn't work too well. Then the gas liquid chromato-
graph came along and everybody dumped their mass spectrometers and bought
GLC's and hooked them up as process control instruments, and they worked, but
not too well. Finally, we have third generation interfaces between the two
instruments which we are utilizing to separate and identify these relatively
minute traces of organics in treated wastewater effluents.
A few examples of the conclusions in completed extramural reports include:
"Fluid Bed Incineration of Petroleum Refinery Wastes" - American Oil Co.
1. The fluid bed incineration process has been demonstrated to
be practical and effective for the disposal of petroleum refinery
generated spent caustic and oily sludge.
2. The process creates no atmospheric pollution problems,
emitting only carbon dioxide nitrogen and water vapor. The odor
of the off gas has been described by various observers as being
slight to non-existent.
26
-------
3. The ash produced contains sodium carbonate, sodium sulfate,
other soluble inorganic salts and inert material such as sand,
clay, rust, etc.
4. Operations are frequently prematurely terminated by the
loss of bed fluidity caused by excessive bed particle size.
The particle size growth rate can be controlled by including
salient design features and by exercising proper operating
techniques.
"Oily Waste Disposal By Soil Cultivation Process" - Shell Oil Co.
1. Disposal of oily sludges (hydrocarbon) by microbial
action in cultivated soil has been demonstrated at prevailing
soil and climatic conditions at Deer Park, Texas.
2. Three simultaneous experiments with three oils, i.e.,
crude oil, bunker C fuel oil, and waxy raffinate oil, indi-
cated decomposition rates for the three oils to be approxi-
mately equal and averaged about 0.5 pounds of oil per cubic
foot per month without adding nitrogen and phosphorus
nutrients and about 1.0 pound per cubic foot per month when
fertilizers were added. This is equivalent to about 70 bar-
rels of oil per acre per month.
3. Cost of the soil cultivation process based on the demon-
stration project expenses and a disposal rate of 70 barrels
of oil per acre was $7.00 per barrel of oil. Assuming oily
sludges and waste materials contain 33 percent oil, the
disposal cost by the soil cultivation process would be about
$3.00 per barrel.
4. An optimum fertilization program appears to be a) the
initial addition of chemicals, if needed based upon soil test
results, to attain a slight excess of nitrogen, potassium,
and phosphorus, and b) test at regular intervals, once per
month, for ammonia and nitrate contents of the soil and add
small dosages of ammonium nitrate as needed to maintain a
positive test result (10-50 ppm) for ammonium and/or nitrate
contents.
5. The major species of microorganisms present are members of
the genus Pseudomonas, Flavobacterium, Nocardia, Coryne-
bacterium, and Arthrobacter. The nature of the hydrocarbon
substrate did not appear to influence the type of organisms
present but did affect the number of bacteria in the soil.
Crude oil tank bottoms produced the highest count, waxy
raffinate oil produced an intermediate count, while bunker C
fuel oil exhibited the lowest microbial population. Temp-
erature appeared to have no effect upon the microbial count
and distribution. Addition of fertilizer did not affect the
microorganism distribution but appeared to be directly re-
lated with the total aerobic count.
27
-------
6. Oil decomposition rates were low when the concentration of
oil in the soil approached the starting condition of 10 percent
oil in the soil. Also, the low reaction period coincided with
the winter months and low temperature period.
7. Both aromatic and saturated hydrocarbons were reduced with
time in the soil for crude oil tank bottoms and bunker C fuel
oil. Only the saturate fraction of waxy raffinate oil appeared
to be reduced by soil microbial action at conditions existing
during the project period.
8. Infrared and gas chromatographic analyses of the oil added
to and extracted from soil indicated a) the absence of organic
acids in oils added to the soil and the presence of organic acids
in each of the extracted oils (oils from plots that were ferti-
lized showed higher concentrations of organic acids), b) the
organic acid increase coincided with a decrease in total satu-
rates, and c) the percent weight boiling less than 500°C gener-
ally was lower for the oil extracted from the soil at the finish
of the project than for either the oil added to or the oil ex-
tracted from the soil at the start of the project (the lowest
percent weight boiling up to 500PC was extracted from soil which
had received the largest quantity of fertilizer materials).
9. Oil and fertilizer chemicals did not infiltrate vertically
into the soil at the test location and condition.
10. Rainfall runoff water contained 30 to 100 ppm oil. This
oil appeared to be essentially naphthenic acids based upon
infrared inspection of oil fractions. Also, rainfall runoff
water contained ammonia (nitrogen nutrient) approximately pro-
portional to the excess ammonia content of the soil. Phosphorus
and nitrates were not found in runoff water.
11. Oil and nutrient contents of rainfall runoff water from
the soil cultivation process can be relatively high, and this
discharge water should receive treatment before entering public
waterways.
"Refinery Effluent Water Treatment Plant" - Atlantic Richfield
1. Reduction of Chemical Oxygen Demand (COD) content in re-
finery wastewater effluent has been demonstrated to be feasible
by using activated carbon as an adsorbent.
2. The system performed well in that it demonstrated an
ability to start up and shut down without delay or difficulty.
This gives the process a distinct advantage over biological
units for use in handling intermittent rainfall.
3. During the two-year project the unit was operated at an
overall average cost of 49 cents per 1000 gallons of water
treated or 24 cents per pound of COD removed from the
28
-------
effluent wastewater. The first year's operational costs were
62 cents per 1000 gallons of water treated, or 34 cents per
pound of COD removed. After improvements over the first
year's operation, the costs during the second year were reduced
to 40 cents per 1000 gallons of water treated, or 18 cents
per pound of COD removed.
4. The plant demonstrated excellent reliability.
5. The carbon adsorption plant has demonstrated that when the
feed COD is controlled to an average of 233 ppm, the plant can
treat refinery water using 8.5 pounds of carbon per 1000 gal-
lons treated and reduce the effluent COD to an average of 48
ppm with a high level of 95 ppm.
"Chemical Coagulation/Mixed-Media Filtration of Aerated Lagoon Effluent"
American Oil Co.
1. Chemical coagulation/mixed-media filtration has been demon-
strated to be an effective tertiary water treatment for accom-
plishing further reductions in suspended solids, oil content
and biochemical oxygen demand following a refinery end-of-pipe
treatment sequence consisting of an API separator and aerated
lagoon.
2. The most significant factors influencing the treatment
effectiveness in the scalping mode operation are incident con-
taminant concentration, the applied chemical treatment level,
and seasonal aerated lagoon conditions.
3. Backwashing with lagoon water will keep the filter media
satisfactorily free of oil and waste accumulation, and return-
ing the backwash waste to the lagoon can be handled in a way
to avoid unmanageable sludge buildup.
4. The filtration facilities require a minimum amount of
operator attention and generate no objectionable wastes or
odors.
5. The unit operation can be demonstrated easily, appearance
of the effluent is greatly improved and the facility there-
fore has considerable public relations appeal. Both the
process and the results are easily visible and comprehensible.
6. Unless adequate hydraulic controls are provided, the filter
media can be physically disturbed, rendering it less effective
than designed. Safeguards are necessary to prevent such
distrubances.
7. Optimized operation to achieve near potable water quality
clarity requires response to the required water chemistry
for destabilization of colloidal material. This response may
29
-------
require taking any, or all, of the following into account:
1) pH control; 2) two or three chemical destabilization
systems, and 3) reduced hydraulic loading at water tempera-
tures less than 15.5°C (60°F).
8. Brackish water has a major impact on the effectiveness
of destabilization chemicals.
9. Brackish water interferes with direct usage of zeta
potentials for determining optimal chemical pretreatment
because the reduction of zeta potential by double layer
repression must be detected. This is readily achieved by
determining zeta potentials on diluted samples.
10. Colloid entrapment and double layer repression are
destabilization mechanisms to be sorted out and avoided for
direct filtration.
11. Charge neutralization and bridging are required destabi-
lization mechanisms for optimal filter performance.
12. Weakly anionic polyelectrolytes are much more effective
than nonionics for filter aids in a three chemical system.
13. Even with optimized chemical pretreatment, filter loading
must be decreased with decreasing water temperatures.
14. Recommended filter hydraulic loadings are proportional
to the viscosity of water.
15. Incorporation of powdered activated carbon up to 150 mg/1
had no favorable impact on the destabilization chemistry of
biocolloids.
In-house reports and papers produced by program personnel include:
"Analytical Variability' of Five Wastewater Parameters - Petroleum Refining
Industry" - Petroleum-Organic Chemical Wastes Section, Robert S. Kerr Environ-
mental Research Laboratory.
1. The chemical oxygen demand (COD) test had a repeatability
expressed in terms of standard deviation of 9.5 milligrams per
liter (mg/1) for petroleum,refinery wastewater which had a COD
average of 134 mg/1. Reproducibility for this same refinery
wastewater exhibited a standard deviation of 15.0 mg/1.
2. Suspended solids with an average concentration of 19 mg/1
had a standard deviation for repeatability of 1.8 mg/1 and a
standard deviationifor reproducibility of 5.2 mg/1.
30
-------
3. Results obtained for the ammonia test with an average
concentration of 8.5 mg/1 exhibited a repeatability standard
deviation of 0.1 mg/1 and a reproducibility standard deviation
of 0.9 mg/1.
4. The repeatability standard deviation for phenolics was
0.2 mg/1 for a sample containing 5.5 mg/1; the standard
deviation for reproducibility of phenolics was 0.8 mg/1.
5. Oil and grease standard deviation for repeatability was
2.3 mg/1 and reproducibility was 2.9 mg/1 for a sample contain-
ing approximately 11 mg/1.
6. There does not appear to be any major differences in con-
centration between the results of hexane extraction and freon
extraction procedures.
7. The variance of the analysis for oil and grease appears
to be less for the freon method than the hexane method.
8. A better than 95 percent spike recovery for COD and
ammonia was achieved.
9. A comparison of results between Phases I and II indicate
the instruction seminar, which was held to achieve uniformity
of analytical procedures accomplished:
a. A significant reduction in arithmetic and extreme
outlier value errors;
b. Enhancement of uniformity of laboratory technique;
'•'<*. -j
c. Minimizing the COD mean values between intra-
laboratory and interlaboratory results;
d. Improvement of spike recovery for COD and ammonia.
10. The standard deviations for COD, ammonia, and phenolics
were decreased between Phase I and Phase II.
>,
11. Environmental Protection Agency (EPA) methodology for the
parameters studied appeared applicable for petroleum refinery
wastewater when the analysts were properly instructed.
"Acute Toxic Effects of Petroleum Refinery Wastewaters on Redear Sunfish"
John E. Matthews and Leon H. Myers, Robert S. Kerr Environmental Research
Laboratory.
1. Short-term static bioassays of 24-hours' duration can be
an effective tool for screening industrial process wastewaters
to locate sources of toxic agents; these tests can also be
used to evaluate effectiveness of industrial waste treatment
processes.
31
-------
2. Static bioassays cannot be used to obtain reliable 50 per-
cent tolerance limit (TL5Q) values for low toxicity wastes
that exert a high oxygen demand although a range can often be
established by an experienced observer. Dissolved oxygen (DO)
usually becomes critical at about 12 hours; low DO plays a maor
role in a mortaility after this time. Activated sludge treat-
ment processes tend to reduce the oxygen demand beyond the
critical stage.
3. Raw wastewaters from different oil refineries vary greatly
in their toxic characteristics; wastewaters from different
processes within a single refinery also vary greatly in their
toxic characteristics. Toxicity of oil refinery wastewaters
varies considerably at different treatment stages in the
activated sludge process; toxicity appears to decrease follow-
ing each stage of treatment.
4. Toxicity of oil refinery wastewaters cannot always be
predicted from results of chemical analyses; the toxic effect
of the waste is dependent on the synergistic or antagonistic
activity of toxicants present.
5. The most common toxic constituents of untreated oil
refinery wastewaters are: ammonia, sulfides, phenolic
compounds, and cyanides. Raw wastewaters also may contain
other toxic compounds including various hydrocarbons.
6. Due to the volatile and unstable nature of some toxic
components of oil refinery wastewaters which may have led
to a reduction in concentrations during sample transporta-
tion, storage, andhandling, TL^Q values obtained during these
tests may be higher than the actual value.
7. Acute toxic effects of raw wastewaters from oil refineries
are generally exerted within the first 12 hours of the static
test; therefore, a 24-hour test will provide good positive
results under static conditions. If samples containing toxic
compounds, other than those mentioned above, toxic effects may
be exerted over an extended period. Tests should then be
continued for at least 48 hours to obtain more positive
results.
8. Toxicity of the final clarifier effluent from oil
refineries with activated sludge treatment systems is depen-
dent on toxic constituents present in the influent and their
concentrations. Results of chemical analyses conducted dur-
ing these tests indicate that concentrations of most toxicants
other than ammonia are reduced by activated sludge treatment
systems and at least a four-fold decrease in toxicity can be
expected after treatment.
32
-------
9. Although activated sludge treatment systems have very
little effect on ammonia concentrations, the toxic effect
of ammonia is lessened due to a decrease in pH of the treated
effluent.
A listing of some of the current funded research may be of interest:
"Identification of Refractory Organic Compounds from Treated Refinery Waste-
waters" - ERDA, Division of Environment & Safety, Washington, D.C. Project
Site - Argonne, Illinois
"Cyanide Removal from Petroleum Refineries" - Illinois Institute of Technology
Research Institute, Chicago, Illinois.
"Refinery Sour Water Stripper Studies" - American Petroleum Institute,
Washington, D.C. Project Site - Ponca City, Oklahoma.
"Powdered Carbon-Activated Sludge-Filtration Processes for Petroleum Refinery
Wastewater" - Atlantic Richfield Co., Harvey, Illinois.
. <*
In closing, the lesson which I hope we^have,learned from this recap-
itulation of the Laboratory's research program history is that the list of
priority pollutants will not be an end in itself. We should apply the
available technology to its utmost to determine what is in an effluent,
where it came from, and how to remove or control it, if it is present in
environmentally significant amounts. Such approach may avoid or at least
ameliorate future crises.
Despite having disposed of soothsayers earlier, the urge to predict is
too strong to resist. Try these with respect to the priority pollutants
list.
1. Some compounds on the list ain't gonna be found in refining effluents
2. Some compounds not on the list will be found in refining effluents
3. If you continue to search long enough and hard enough for some-
thing you may find it anywhere—even though it ain't really there. Now
ponder on that one a while.
33
-------
DISCUSSION
Paul Mikolaj, Lion Oil Company; Is there any research work being done on radiation
techniques to treatwastewater?
Marvin Wood; There is, of course, some being done, but none that I am aware of or
at least none that we are funding in this specific area of refinery wasfes. There is
some being done in Hie case of other wastes—domestic and possibly a few cases of
industrial. It might involve come coblat 60 or something sources or radiation to change
the physical characteristics to make the wastes more amendable to handling.
BIOGRAPHY
Marvin L. Wood holds a B.S. in Chemical
Engineering from the University of Arkansas and an
M.S. 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.
34
-------
SESSION II
'PANEL DISCUSSION: RESEARCH SUPPORTED IN THE PRIVATE SECTOR"
Chairman
* Milton R. Beychok
Consulting Engineer
Irvine, California
Panel Members
Robert T. Denbo
Exxon
Baton Rouge, Louisiana
Kent G. Drummond
Marathon Oil Company
Findlay, Ohio
Francis S. Manning
University of Tulsa
Tulsa, Oklahoma
David C. Bomberger
Stanford Research Institute
Menlo Park, California
Judith G. Thatcher
API/DEA
Washington, D.C.
*Biography on Page 300
35
-------
"DISCUSSION OF RESEARCH RELATED TO PETROLEUM REFINERY
WASTEWATER SPONSORED BY THE API COMMITTEE
ON REFINERY ENVIRONMENTAL CONTROL"
Robert T. Denbo
Coordinator of Environmental Control, Exxon
Baton Rouge Refinery
Mr. Drummond's and Ms. Thatcher's papers discussed the organization of the
Division of Environmental Affairs of the American Petroleum Institute and research
sponsored by the Division of Environmental Affairs. This paper will cover the research
program sponsored by the Committee on Refinery Environmental Control (CREC).
First, some added background on the API. The API was incorporated in 1919.
The objectives were set forth as follows:
- afford a means of cooperation with the government in all matters of national
concern.
- foster foreign and domestic trade in American petroleum products.
- promote, in general, the interest of the petroleum industry in all its branches.
- promote the mutual improvement of the members and the study of the arts and
sciences connected with the petroleum industry.
Next, it may be appropriate to put the CREC Committee in proper prospective in
the American Petroleum Institute. The first chart shows the organizational setup in API
down to the divisional level.
The second chart shows the relationship of the CREC Committee to the General
Committee of the Division of Refining. The CREC Committee was set up to have the
following responsibilities:
- the development and publication of the API Manual on Disposal of Refinery
Wastes which covers atmospheric emissions, solid wastes and liquid wastes.
- the sponsorship of research related to refinery pollution control.
36
-------
AMERICAN PETROLEUM INSTITUTE
BOARD OF
DIRECTORS
1
ADMINISTRATION
MANAGEMENT
COMMITTEE
1
INDUSTRY
AFFAIRS
BUDGET ADVISORY
COMMITTEE
1
COMMITTEE ON
PUBLIC AFFAIRS
COMMITTEE ON
POLICY
DEVELOPMENT
1
DIVISIONOF
ENVIRONMENTAL
AFFAIRS
1
DIVISIONOF
.REFINING
ICOSH
DIVISIONOF
PRODUCTION
1
DIVISIONOF
MARKETING
1
DIVISIONOF
FINANCE AND
ACCOUNTING
1
DIVISIONOF
TRANSPORTATION
1
DIVISIONOF
MEDICINE AND
BIOLOGICALSC.
37
-------
DIVISION OF REFINING
INTERDIVI SIGNAL
COMMITTEE ON
OCCUPATIONAL
SAFETY AND HEALTH
GENERAL
COMMITTEE
COMMITTEE ON
LIQUEFIED
HYDROCARBON
GASES
COMMITTEE ON
REFINERY
ENVIRONMENTAL
CONTROL
COMMITTEE ON
REFINERY
EQUIPMENT
COMMITTEE ON
RESEARCH, DATA
AND INFORMATIO
SERVICES
N
OPERATING
COMMITTEE
COMMITTEE ON
TRAINING AND
DEVELOPMENT
OPERATING
PRACTICES
COMMITTEE
PETROLEUM
PRODUCTS
COMMITTEE
38
-------
- the dissemination of information on legislative developments concerning stream
and air pollution and solid waste disposal.
It is within this context that the R&D related to petroleum refinery waste water is
carried out. During the past few years, CREC has worked closely with the API Division
of Environmental Affairs and participated actively in many of the projects of the Water
Quality Committee. CREC's own research and development and related studies have
resulted In the publication of the following reports:
- the volumes on liquid wastes including:
o separator design parameters
o biotreatment design parameters
o stripping
o extraction
o adsorption
o and other areas
- a filtratfon volume
- a sour water stripper volume.
The following projects are currently in the planning stage tentatively for 1978:
- Enhancement of biological treatment - Phase III
- Studies of handling of storm runoff from petroleum refineries
- Studies of variability of treated effluents
- Benthic assays of treated refinery effluents.
The following two projects will be discussed briefly in this paper for purposes of
illustration, how research and related studies are developed in the CREC Committee:
- Sour water stripping
- Bioenhancement
Sour Water Stripping
Sour or foul water in petroleum refining is process water containing organics that
also contains more hydrogen sulfide and ammonia than can be handled effectively in a
biological treatment system. The types of organics present in this stream include phenolic
type compounds and other objectionable materials. This type of water is probably the
most difficult to handle of all process waste waters in most refineries. It is usually
39
-------
produced in such operations as a steam condensate from catalytic cracking, coking, hydro-
cracking and other hydroprocessing. Effective removal of the hydrogen sulfide and
ammonia are essential to allow meeting limits on these two compounds and to permit
effective biological oxidation of the phenols and other organics.
Ron Gantz of Continental Oil has been the guiding hand in the CREC-sponsored
studies by contractors to improve the efficiency of removal of these two constituents by
steam stripping which is the existing state of the art. The current program of study has
extended over a period of more than three years and has been, in part, a joint effort by
EPA and API CREC. The EPA Water Labs at Ada, represented by Leon Myers, played a
significant role in the development of this work. The program has consisted of extensive
surveys of actual plant operations to determine state of the art in the industry, plant
testing to obtain actual operational data, the operation of pilot facilities in laboratories,
and the correlation of all the results into a comprehensive, new equilibria calculation
procedure.
The results from these studies are to be published soon. The most significant
findings include:
- the identification of ammonia that cannot be removed by stripping from the
water as produced - so called "fixed ammonia". This fixed ammonia
apparently will not come out of solution by stripping if it exists as ionic
ammonia in combination with carboxylic acids and other such compounds when
slightly acidic. It was found that the presence of a slight excess of sodium
hydroxide over the stoichiometric amount will allow stripping to proceed
efficiently.
- another significant result has been the development of a new equilibria calcu-
lation procedure. New equations have been developed and tested which give
much more accurate results and can be used more flexibly than existing
accepted approaches. The procedure exists as a computer program which will
permit calculations including the effect of pH adjustments at selected points in
the stripper tower. This is not possible in the existing sour water stripping
computational approach.
Bioenhancement
The CREC study of bioenhancement will be covered in some detail in Session VIII
of this open forum by Lyn Crame of Texaco. Consequently, it is not appropriate to go
into the findings of this study in this paper.
However, it is pertinent to say that this study was started about a year ago prior
to the availability of initial results from other API studies which were designed to deter-
mine the presence of residual organic compounds in refinery effluents after biological
treatment. By "residuals" is meant organics present in the effluent which were introduced
in the raw waste load feed to biological treatment. Initial results of the study indicate
40
-------
only a small amount of residuals are present in biologically treated effluent. Apparently
the vast majority of organics in the treated effluent are metabolities, humic acids, and
other compounds associated with the biodegradation process. Almost no objectionable
compounds have been found in biologically treated effluents. At this point it would not
appear that it is practical to require further treatment on the basis of results of con-
ventional analyses or even results on the basis of analyses for the more exotic list of toxic
compounds.
Conclusion
The CREC program for the past few years has been consistent with the topics dis-
cussed at this open forum. The agenda of this open forum suggests the high degree of
sophistication that is employed in the treatment of refinery wastewater. It can also be
said that the performance of treatment facilities for petroleum refineries suggest that the
current quality of treated wastewater from these refineries is no longer a significant
problem in water pollution abatement. It appears that technology and research in this
area has advanced to the point where it is well ahead of other areas of environmental
concern for the industry.
BIOGRAPHY Robert T. Denbo
Robert T. (Bob) Denbo is the coordinator of
environmental control at Exxon's 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 University 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 environmental control for all domestic
refineries. He is chairman of the American Petroleum
Institute Committee on Refinery Environmental Control,
member of the API Division of Environmental Affairs,
former president of the Louisiana Water Pollution
Control Association, member of the Technical Advisory
Committee to the Louisiana Air Control Commission, and
member of the Air Pollution Control Association. Mr.
Denbo has served on a number of panels related to develop-
ment of information on target values for U.S. refinery
water effluents. He has served on a joint U. S.-Soviet
planning group for cooperative effort in refinery water
pollution controls.
41
-------
WATER QUALITY COMMITTEE OF API
Kent G. Drummond
Technical Coordinator, Environmental Control Div.,
Marathon Oil Company
The Water Quality Committee is one of eight standing technical committees under
the General Committee on Environmental Affairs (see Figure 1). The General Committee
Is composed of 28 oil company environmental coordinators. Under the General Committee
are 8 standing committees which have over 50 task forces involving more than 300 oil com-
pany personnel.
These committees meet on a regularly scheduled basis of twice a year; once in
the spring to review current projects and to generate new projects for the forthcoming
year and again in the fall to review current projects. The spring meeting is the normal
time for proposing the next year's budget.
In the case of the Water Quality Committee, if a project is accepted by our
committee, it is then proposed as part of the committee's budget, to the Planning
Budget Advisory Committee which is under the General Committee on Environmental
Affairs. Incidentally, each standing technical committee has a liaison representative
from the Planning Budget Advisory Committee.
If PBAC approves the project, it is then submitted to the parent Environmental
Affairs Committee for consideration. The project then becomes part of the proposed
budget of the Environmental Affairs Committee to be considered by the API Finance
Committee and, if approved, becomes a funded project for the new year's budget.
Figure 2 lists the members' company affiliation on the Water Quality Committee,.
Of the 18 members on the Committee, only 2 companies have 2 representatives and this
is because their respective representatives are either chairmen of more than one task
force or are a representative from the Committee on Refinery Environmental Control.
There are 3 main project areas of interest as shown in Table 1. The Water
Quality Committee is primarily interested in the quality of water as it leaves the
premises. We also become involved in projects which need immediate attention. At
other times, we get into projects which are beyond the monitory scope of the various
divisions.
43
-------
Later on this afternoon, Mrs. Judith Thatcher will discuss 2 of the projects which
come under the Water Quality Committee.
In 1977, we had 10 active projects. These projects are listed in Table 2. Their
names and the budgetary allocations are given. Our original budget of $240,000 was
approved in December, 1976, however, because of the changes in government regulations,
we came back with the supplementary budget of $450,000 in March, 1977. Project W-14
and the last 4 items on Table 2 were approved at that time.
Our 1978 proposed budget came to a total of $515,000. This is shown in Table 3.
For the past five years, the budgetary allocations for the Water Quality Com-
mittee have continued to grow. Back in 1974, we had a budget of only $60,000; whereas
in 1978, our proposed budget is $515,000 plus 40% of $450,000 which is $180,000;
making a total of $695,000 for the year of 1978. This is all shown on Figure 3.
In the past six years, the Water Quality Committee has attempted to learn
more about our own refinery operations. With this better understanding, we have been
able to defned ourselves against unreasonable regulations and to discuss with EPA regu-
lations which we feel attainable.
The five projects which we have proposed for 1978 will help us to better under-
stand some of the problems within our refineries and hopefully, this will lead to guide-
lines in 1983 which are attainable and will give best valuable treatment economically
achievable.
BIOGRAPHY Kent G. Drummond
Kent G. Drummond is Technical Coordi-
nator in the Environmental Control Division of
Marathon Oil Company, Findlay, Ohio. He
has a B.S. degree in Civil Engineering and a
M.S. degree in Sanitary Engineering from Iowa
State University. He is currently Chairman of
API's Water Quality Committee and Chairman
of the 1979 Oil Spills Conference.
44
-------
TABLE 1
WATER QUALITY COMMITTEE PROJECT AREAS OF INTEREST
1. Quality of wafer leaving premises
2. Projects which need immediate attention
3. Projects beyond monetary scope of divisions
TABLE 2
WATER QUALITY COMMITTEE 1977 PROJECTS
No. Title Budget $k
W-12 Bioassays of Refining Effluents $ 90k
W-14 Toxic Pollutant Effluent Standards $ 25k*
W-15 BPT Vs BAT Petroleum Refining Guidelines $100k
W-19 Non-process Effluent Standards
W-20 Analysis of Refinery Effluents (BPCTCA) $ 50k
W-21 Amendments to 92-500
W-22 Sampling & Testing Protocol for Toxic Pollutants $200k*
W-23 Refinery Questionnaire (EPA) Analysis $100k*
W-24 Refinery Effluent Pollution Contribution in Perspective $100k*
W-25 Economic Study BPT to BAT $ 25k"
,*
Total $240k $450k*
*Approved March '77
TABLE 3
WATER QUALITY COMMITTEE 1978 PROPOSED BUDGET
No. Title Budget
W-12P (cont'd) Bioassays on Refiner/ Effluents $ 40k
W-22P (cont'd) Sampling & Testing Protocol for Toxic Pollutants $200k
W-23P (cont'd) Refinery Questionnaire Follow up $100k
*W-26P (new) Removal of Toxics from Refinery Effluents - Pilot $ 75k
Plant (Joint Study)
*W-27P (new) Evaluation of Selected Hazardous Refinery $100k
Chemicals to F.W. Organisms
Total $515k
* Projects to be carried out if significant quantities
are found in effluents.
45
-------
COMMITTEE ON ENVIRONMENTAL AFFAIRS
I
PBAC
GOVERNMENT
LIAISON
COMMITTEES
1. Fate & Effects of Oil
2. Oilspill Prevention & Control
3. Mobile Source Emissions
4. Environmental Economics
5. Stationary Source Emissions
6. Solid Waste Management
7. Air Quality
8. Water Quality
ENVIRONMENTAL AFFAIRS ORGANIZATION CHART
API Environmental Affairs
Figure 1
WATER QUALITY COMMITTEE
MEMBERSHIP BY COMPANY
Chairman - Marathon
Vice Chairman - Standard of Indiana
Members
Chairman of CREC - Exxon Refining
Chairman of Liquid Wastes - Union Oil
Task Force Chairman
1. Exxon Research & Engineering
2. Union Oil Co.
3. Gulf Oil Co. Others
4. Standard Oil of Ohio 1.
5. Shell Oil Co. 2,
6. Shell Oil Co. 3.
7. Sun Oil Co. 4.
8. Mobil Oil Co.
Atlantic Richfield Co,
Getty Oil Co.
Texaco, Inc.
Chevron Oil Co.
5. API Representative
Figure 2
46
-------
$600k-i
$500k
w $400k
DC
5
0 $300k
O
$200k
$100k
0
o I
to I
to
U)
M
S
CM
|
(O
O
CM
O
U)
u> to
5 U)
1974
1975
1977 1978P
1976
YEAR
WATER QUALITY COMMITTEE
PAST BUDGETS 1974 - 1978P
Figure 3
47
-------
OVERVIEW OF RESEARCH ON PETROLEUM REFINERY
WASTE WATERS AT U.S. UNIVERSITIES
Francis S. Manning
Professor of Chemical Engineering
University of Tulsa
This overview addresses research at United States universities not covered by the
previous papers. In other words, research sponsored by federal agencies such as the
E.P.A.; the A.P.I.; and research institutes has been summarized elsewhere and will not
be repeated. Obviously it is impossible to canvas every potential department in every
university and hence the following list will be incomplete. The author apologizes to
those researchers whose efforts have been omitted.
i
This overview will not discuss the following studies because they are described else-
where in this symposium: -
o the Brigham Young study of liquid-vapor equilibrium of H«S-NH«-CO«
which is reviewed by Milton Beychok in Session VI
o the University of California at Berkeley work on stripping of organics using
volatile solvents - see Marvin Wood in Session I
o the University of Texas at Austin investigation of the effect of temperature
on critical sludge age as this is described by Davis Ford in Session V
o the University of Delaware experiments on PAC in biological treatment
which is covered by Francis Robertaccio in Session VII.
This survey uncovered three long-term research programs which, surprisingly, have
many common features. All three are sponsored by groups of petroleum and petroleum
related industries and all three address very specific research objectives. These three
programs at Louisiana State University, Oklahoma State University, and the University of
Tulsa are now described.
LOUISIANA STATE UNIVERSITY
The Petroleum Refiners Environmental Council of Louisiana (PRECOL) group has a
long history of financial support for environmental research in the Zoology and Physiology
48
-------
Department at LSD. Initial support occurred in 1946 when the Louisiana Petroleum
Refiners Waste Control Council provided two fellowships for the determination of
immediate and accumulative toxicity of refinery wastes to fish. Participating companies
included Chalmette Petroleum Corporation, Cities Service, Continental Oil Company,
Pan American Petroleum Corporation, Shell and Standard Oil.
Bioassay techniques were developed in the 1950's and pesticide toxicity work
became the focal point of the laboratory during the 1960's. Emphasis shifted to toxicity
studies on the effects of crude oil, emulsifying agents and interactions between them on
aquatic fauna during the early 1970's. The organizational name was changed to PRECOL
in 1973. The community structure of the Calcasieu estuary began in 1974 along with an
investigation of heavy metal movement through the estuary. Two new projects were
initiated in 1977. L.S.U. is investigating the accumulation of heavy metals and organic
pollutants by water snakes and benthic invertebrates from the Mississippi River near the
Baton Rouge industrial complex. The effects of low oxygen tension on the survival and
metabolic rate adjustments of aquatic fauna are also being investigated.
Financial support for research efforts has continuously grown as has the number of
oil companies providing it. The PRECOL group now consists of: Cities Service Oil
Company, Continental Oil Company, Exxon Company, Gulf Oil Company, Murphy Oil
Corporation, Shell Oil Company, Tenneco Oil Company and Texaco, Inc. The L.S.U.
laboratory has always enjoyed a close working relationship with the Louisiana Division of
Water Pollution Control which provides technical support for graduate students.
OKLAHOMA STATE UNIVERSITY
Since 1956 the Oil Refiners Waste Control Council of Oklahoma has sponsored bio-
logical investigations on petroleum refinery waste waters in Oklahoma. This work,
originally directed by Professor Troy Dorris and directed since 1970 by Professor S. L.
Burks, is currently studying:
o The Effects of Residual Volatile Toxins in Oil Refinery Waste Waters
o The Biological Evaluation of BPTCT and BATCT for Refinery Waste Waters
o The Development of Benthic Bioassay Techniques.
At present the Oil Refiners Waste Control Council of Oklahoma includes the following
members: APCO, Champlin, Conoco, Hudson Oil Company, Kerr-McGee, OKC
Refining Company, Sun Oil, Texaco and Vickers.
The recent biological evaluation of activated sludge, sequential activated sludge-
dual media filtration (equivalent to Best Practicable Treatment Control Technology,
BPTCT) and sequential activated sludge-dual media filtration-activated carbon adsorption
(equivalent to B_est Available Treatment Control Technology) at the ETU refinery showed
that;
49
-------
1. Activated sludge and BPTCT would not produce a non-toxic effluent.
2. BATCT produced an effluent which did not cause any significant deleterious
effects upon fish, benthic macroinvertebrates, and periphyton during 32-
day exposures.
Chemical analyses of the treatment effluent streams indicated that most of the toxicity was
associated with the organic substances as measured by TOC and COD analyses.
A new bioassay method for determining the effects of environmental contaminants on
populations of benthic macroinvertebrates has been developed. Colonized Hester-Dendy
samplers were transported from a natural stream to artificial streams and exposed to
industrial waste water. Species diversity, number of taxa, and density of the aquatic
organisms were measured before and after selected time intervals of exposure. A 30 and
a 32-day continuous-flow exposure test with the benthic macroinvertebrates showed that
activated sludge treated petroleum refinery wastewater caused a greater decrease in
species diversity than the sequential activated sludge-dual media-activated carbon
treated effluent. The effect upon number of taxa and mean density of individuals was
even greater. This procedure has permitted exposure of several species of aquatic
invertebrates to the test solution and thus measured the effect upon pollution sensitive and
tolerant organisms.
UNIVERSITY OF TULSA
The University of Tulsa Environmental Protection Project (UTEPP) began January 1,
1974. The UTEPP program, which is directed by Professor Nicholas D. Sylvester of
Chemical Engineering, has received support from the following companies: Ameron,
Amoco Production Company, Aramco, Bechtel Corporation, Calgon Corporation, Chevron
Research Company, Crest Engineering, Compagnie Francaise des Petroles, Dow Chemical,
U.S.A., Exxon Production Research Company, Foster Wheeler Corporation, Getty Oil
Company, Iranian Oil Services, Ltd., Lummus Company, Marathon Oil Company, Mobil
Research and Development Corporation, and Pullman Kellogg.
Currently the program includes the following projects:
1. Chromate Removal From Cooling Tower Slowdown
Commercially available ion exchange resins for chromate removal from
cooling tower blowdown are being evaluated. The effectiveness of
chromate removal, the regenerability of the resins and chromate recovery
are being studied.
2. Oil Removal From Wastewaters by Induced Air Flotation
The effects of oil type, concentration, drop size and size distribution;
process residence time, air flowrate, bubble size; and polyelectrolytes on
the performance of induced air flotation is being studied.
50
-------
3. Oil Removal From Wastewaters by Ultrafiltration
The removal of gulf coast crude oil from wastewaters by ultrafiltration is
being investigated. The effects of oil type and concentration, flowrate
and salt concentration on oil removal efficiency are being studied.
4. Pollutant Removal by Adsorption and Reaction in Aqueous Slurries of
Powdered Activated Carbon
The adsorption and oxidation reaction characteristics of pollutants (SOo
and organics; e.g., benzene) in aqueous slurries of high surface area
powdered activated carbon is being studied, In conjunction, the effects
of the mass transport processes attendant to slurry reactors on adsorption
and oxidation are being determined. In addition, the catalytic activity
and mechanisms of activated carbon towards oxidative pollutant removal
will be elucidated.
5. Urea Removal From Industrial Wastewaters
A state-of-the-art review of urea removal from industrial wastewaters is
being prepared. Although urea is synthesized on a large scale for use in
the manufacture of urea-formaldehyde resins, the primary emphasis of
the review will be to its manufacture for use as a fertilizer.
6. Activated Sludge Enhancement With Powdered Activated Carbon
An experimental study is being initiated to investigate the mechanisms
involved in activated sludge enhancement with powdered activated carbon.
In particular, answers will be sought to questions such as:
(i) How much does the blocking of the carbon surface bacteria and their
by-products reduce the adsorptive capacity of the carbon?
(ii) What is the mechanism of carbon surface renewal by the action of
microorganisms?
In addition, appropriate mathematical models of the process will be
developed.
SUMMARY
It is hoped that more universities will become involved in the treatment of refinery
wastewaters. While considerable progress has been made due to encouragement and
financial support from the U.S. EPA, API, etc., several questions remain to be answered.
Some key topics are: -
o the effect of temperature on biological degradation operating performance
during the severe winter of 1976-77 has questioned the accuracy of the
traditional Streeter-Phelps equation
o the effect of PAC on nitrification
o identifying inhibitors of biological kinetics
51
-------
dewatering sludges especially when the contributions from refinery API,
DAF, and ASP units are combined.
ACKNOWLEDGEMENTS
Professors Bud Burks; Bill Stickle and Nick Sylvester provided the descriptions of
the major programs at O.S.U., L.S.U. and T.U. respectively.
BIOGRAPHY
Francis S. Manning is the Director of the
Petroleum Energy Research Institute (PERI) and
Professor 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
University for 9 years before Joining the University
of Tulsa in 1968. The author of one book and over
60 papers, Frank's current research interests lie in
thermodynamics, reaction kinetics and industrial
pollution control. In 1969 he received the R. W.
Hunt Silver Medal from the AIME.
52
-------
OVERVIEW OF RESEARCH ON PETROLEUM REFINING WASTEWATERS
AT INDEPENDENT CONTRACT RESEARCH ORGANIZATIONS
David C. Bomberger
Chemical Engineer, Environmental Control Group
SRI International
The major contract research organizations in the United States
were surveyed to determine what research is being done on the waste-
waters from petroleum refineries. Companies that were engaged primarily
in engineering and design were not contacted. Any research relating
to identification of priority pollutants identified in the 1976 Consent
Decree is not reported here because it will be covered by other panel-
ists.
Very little research relating directly to refinery wastewater is
being done at contract research organizations. Only ten ongoing or
recently completed projects were identified, and only four of these
were concerned directly with refinery wastewater. The other six proj-
ects covered aspects of synthetic fuel production and utility water
usage. The results of these projects could be utilized by petroleum
refiners through technology transfer.
Most of the research funding is from government agencies, princi-
pally ERDA and EPA. Only two projects were funded by private industry,
one by a single company and the other by the American Petroleum Insti-
tute.
The ten research projects cover four general topics: the cost
of effluent treatment, wastewater treatment technology, measurement of
wastewater components, and reduction of effluent volume, as summarized
below.
Cost of effluent treatment:
" Battelle has studied the economic impact of environmental
regulations on the petroleum industry, for the American
Petroleum Institute.
53
-------
Treatment technology:
• SRI International has studied ammonia fixation in sour
water strippers, for the American Petroleum Institute
(EPA grant).
• Midwest Research Institute has examined a petroleum
refinery treatment system, for a refinery.
• Gulf South Research Institute is studying the technology
for removal and destruction of organic compounds from
wastewater, for the EPA, the State of Mississippi, and
private industry.
Measurement of wastewater components:
• Radian is conducting two projects on the development of
sampling technology and on analysis of effluents from
synthetic fuel production, for ERDA.
• Battelle Northwest is studying the wastewater produced
in shale oil retorting, for ERDA.
• Radian is studying fugitive hydrocarbon emissions from
petroleum refining, for the EPA.
Reduction in effluent volume:
• Radian is investigating optimization of water use in the
utility industry, for the EPA.
• Radian .is also studying saline water use in energy
facilities, for ERDA.
these last two projects are applicable only indirectly and would re-
require technology transfer.
54
-------
BIOGRAPHY
David C. Bomberger is a chemical engineer
in the Environmental Control Group at SRI
International. He has a Ph.D. in chemical engineering
from Princeton University. Before joing SRI Inter-
national, he worked on environmentally related
projects for Bechtel Corporation and The Shell
Development Company.
55
-------
"CURRENT API STUDIES OF RESIDUALS IN REFINERY EFFLUENTS"
Judith G. Thatcher
Environmental Associate, American Petroleum Institute
Good afternoon, ladies and gentlemen.
Mr. Drummond has described the general structure of API's Environmental Affairs
Department and the type of research carried out under the Water Quality Committee. I
would like to discuss two current projects relating specifically to one of the major issues
being discussed at this forum, that is, toxic pollutants.
In late 1974, a widely publicized study focused national attention on the
presence of a number of allegedly toxic organic compounds in New Orleans' public
drinking water. Although the validity of the studies attempting to relate New Orleans'
drinking water to disease has been questioned, the studies understandably generated much
publicity and public concern. There are many possible sources of organics in public water
intakes, including effluents from industrial, chemical, and municipal sewage plants, in
addition to surface run-off.
Although much was known in 1974 about refinery effluents with respect to con-
ventional waste water parameters such as BOD and COD, little was known of the nature
of the organic residuals contributing to the concentration of these parameters in the final
discharge of treated refinery effluents. In order to begin to fill this data gap, funds were
made available to begin work in this area. The project was entitled "Analysis of Residuals
in Refinery Effluents" and is under the direction of API staff and the W-20 Task Force.
The objective of the first phase of the work was to determine the types and
amounts of residuals in the effluent from a refinery meeting BPCTCA, and the potential
effects of these residuals on public drinking waters. During Phase I, samples of the intake
water and the effluent water from a Class B refinery were obtained, along with samples of
the effluent from a municipal sewage treatment plant. It was felt that obtaining and
analyzing samples from the treatment plant would lend some perspective in evaluating any
residuals found in the refinery effluent. The types of analyses that were run on the
samples were metals, polynuclear aromatics, organo-halides and volatile and nonvolatile
hydrocarbons. In addition, classical parameters such as COD, BOD, and total suspended
solids were run. In order to determine whether any precursors of organo-halides were
present in the refinery effluent and intake waters, portions of the refinery intake and
effluent samples were chlorinated and then analyzed. The analyses of these samples were
then compared to the chlorinated municipal plant samples.
56
-------
The objective of Phase II is to determine the types and amounts of residuals in a
second BPCTCA refinery, and the removal of these residuals across various pilot plant
treatment trains. Samples from a second Class B refinery were taken as well as samples
from another municipal treatment plant. The Class B refinery is the same refinery sampled
for the EPA-Argonne study. This study will be discussed in detail on Thursday by Mr.
Fred Pfeffer of EPA's R.S. Kerr Laboratory. In this study, a pilot unit mixed media filter
and activated carbon column were set up at the refinery to treat a slip-stream from the
refinery activated sludge unit. Samples of the refinery's intake water, DAF and ASU
effluents were taken, along with samples of the two pilot units, over a four-day period.
API took samples along with EPA at the same time and place. Our analyses were primarily
for polynuclear aromatics and organo-halides, whereas EPA determined trace organics
present in the samples using gc-ms techniques.
During the planning of the Phase II study, the task force became aware of a pilot
study that was to be carried out by Texaco under contract to API's Refining Department.
This study consisted of evaluating the performance of several different pilot scale treatment
trains — including powered and granular activated carbon — on refinery wastewater. The
W-20 Task Force decided to take samples across the various treatment trains and analyze
them for the presence of polynuclear aromatics and organo-halides. Conventional
parameters were measured by Texaco. Details of this study will be given later in the
program by Mr. Len Crame of Texaco, Inc.
About the time the scope of the Phase II study was being finalized, the toxics
settlement agreement was signed by EPA and several environmental organizations. Under
this agreement, EPA is undertaking a program to determine which contaminants out of a
list of 65 allegedly toxic compounds and classes of compounds are present in the effluents
of 295 industries in 21 major categories, including petroleum refining. As a result of this
agreement, a new task force, W-22, was formed at API. This task force was given the
objectives of (1) reviewing, evaluating, and critiquing analytical methods selected by
regulatory agencies for qualitative and quantitative determination of the presence or
absence of trace toxic pollutants in refinery waste waters; and (2) conducting sampling
and analyses for these trace toxic pollutants at selected refineries.
This task force has been extremely active since it was formed early last fall, and
members and API staff have met with EPA personnel on several occasions. We have
reviewed and critiqued the sampling and analytical protocol being used by EPA in its
analyses of refinery and other industrial wastewaters. The task force at the present time
has grave concerns about the capability of the sampling and analytical protocols to detect
-- and perhaps quantify — the presence of these toxic compounds at exceedingly low
levels, and of the accuracy, repeatability and reproducibility of the methods being
employed.
As Marvin Wood mentioned this morning, Kerr Laboratory personnel are presently
sampling and analyzing waters from 12 refineries. These include intake waters, separator
or DAF effluents, and final effluents. In addition, EPA is conducting pilot scale powdered
activated carbon enhancement of activated sludge units, and granular activated carbon
57
-------
end-of-pipe treatment, at some of these 12 refineries. The W-22 Task Force is
monitoring this effort and has also selected three of the 12 refineries for its own parallel
study. Samples are being taken at the same locations and times as those taken by EPA.
The samples are being analyzed for the 129 priority pollutants using analytical techniques
similar to those being used by EPA contractors. In addition, more detailed and specific
analyses of PNA's are being run. We feel that the results of this study will give us
valuable information concerning sampling and analytical variability as regards the
priority pollutants in refinery wastewaters.
Phase I of the W-20 study is presently in final review stage, and the last of the
analyses for Phase II are now being carried out. As a result, no reports have been issued
and many of the findings are preliminary. In addition, I should emphasize that in view of
the limited number of facilities involved (two refineries and two municipalities), the
study results will not necessarily be typical of either the refining industry or of all
municipal treatment plants. Rather, they were carried out to build a data base where
none existed.
Preliminary results from both the W-20 efforts and the W-22 refinery analyses do,
however, look very encouraging. Although a few of the so-called priority pollutants
were found to be present at levels greater than 10 ppb in separator or DAF effluents, even
these appear to be significantly reduced in the refineries' biological treatment systems.
Available results indicate that there are no significant quantities of the allegedly toxic
pollutants in the final effluent samples of the refineries analyzed to date. In addition,
the W-20 study revealed that there were no significant quantities of organo-halide
precursors in the effluents of the two refineries sampled.
One very interesting result of both studies is that a very poor carbon material
balance is obtained when comparing ppm total organic carbon (as determined by a TOC
analyzer) to the sum of the ppb concentrations of identifiable, extractable organics. One
would expect to find, however, high molecular weight compounds such as cell metabolites,
proteins, carbohydrates, and oxygenated aromatics in biologically treated effluents and
these would not be identified or measured by the analytical methods used. It is quite
possible that these types of compounds account for much of the TOC in refinery effluents
meeting BPCTCA and 'in well operated municipal treatment plants.
It would appear logical, then, that expensive, cost-ineffective technologies
should not be required by regulation just because considerable TOC removals are obtained.
Rather, it should first be demonstrated that the TOC removed is actually harmful to man or
to the aquatic environment before removal is required.
58
-------
BIOGRAPHY Judith G. Thatcher
Judith G. Thatcher holds a B.S. in Chemistry
from Southeastern Massachusetts University. She is
currently an Environmental Associate with the Environ-
mental Affairs Department of the American Petroleum
Institute. Prior to joining API in February of 1976, Ms.
Thatcher worked for 10 years in the Research and
Technical Department of Texaco Inc.
59
-------
DISCUSSION
Ned F. Seppi - Marathon Oil Co. - I wondered if Mr. Denbo would discuss bio enhance-
ment a little more thoroughly. "
R. T. Denbo- I think the subject will be discussed thoroughly in a later formally-developed
session. However, let me just mention generally what we had in mind when we set up a
research project. PL 92-500 includes requirements for BAT - Best Available Technology.
As time has passed, there have been a number of developments that indicate that toxics
have become the important thing as far as water quality was concerned. This began to
become obvious to us a couple of years back. The first question is - do we have a toxics
problem in BPT treated refinery effluent? And the first step in attempting to ascertain
that was the program that Judy Thatcher and Kent Drummond talked about and that was
the analysis of BPT treated refinery effluents for residuals. But before the results became
available we began to look for ways to get rid of compounds of concern if they are, in
fact, present. We decided to look at what can be done along the line of upgrading
existing bio systems. Jim Grutsch and others had done work on higher sludge ages and
how effective this is in further reducing organics in treated effluent. There had been
work done by many people on powdered carbon. We decided to undertake the controlled
experiment that Len Crame will talk about in detail. That's what is meant by bio enhance-
ment or enhancement of biological treatment systems.
Paul Mikolaj - Lion Oil Co. - You mentioned that you're monitoring three refineries
along with the 12 that EPA is monitoring. Are EPA's results coming out the same in terms
of the removal of these priority pollutants?
Judy Thatcher - We have not seen any of EPA's results to date, Paul, and we only have
some limited results from our own studies right now. I was basing my discussion on what
we had found in the two phases of W-20 and on our limited results to date on these three
refineries. We're obtaining samples from the third refinery this week. We would hope
that EPA's results come out the same.
E. A. Buckley - Lion Oil Co. - I would like to ask this question of either Mr. Beychok
or Mr. Denbo. First, Mr. Beychok, I assume that you will cover the injection of caustic
into sour water strippers. My question is: at the present time, do we have enough infor-
mation to use caustic injection in a stripper being built now?
M. R. Beychok - A preliminary report on SRI's work covering that subject was given at
the API meeting in Chicago a few weeks ago. I think that SRI did a very good job of
quantifying that the optimum caustic injection point is at the top of the stripper with the
feed. They also defined the amount of caustic required. That report is available to you
to use now. The final report may not be issued for some months, but the preliminary
report is available and provides enough information to be used now in designing for
caustic injection.
60
-------
William C. Galegar - R. S. Kerr Environmental Research Lab - I would like to turn
around the question that was asked a little earlier this morning. If I include Mr. Beychok
in the group sitting at the table, you broadly represent the group that we have turned to
for handling the research programs that have been undertaken to provide answers. The
question I would like to ask the group is: are the research programs related to the refining
industry wastewater treatment providing adequate information? If not, how can they be
improved? I am talking to your group as a whole since you represent universities, industry
and consultants.
M. R. Beychok - I'll give everyone a chance to catch their breath while I try to respond
to that as an individual private consultant. My opinion is somewhat as stated earlier by
Bob Denbo. We've worked the problems of refinery wastewater technology quite a bit
and I think we now have a good data base. My experience with some of the research
programs such as the SRI work for the API and the Carnegie-Mellon work for ERDA indi-
cates that our emphasis and attention should now be devoted to laboratory analytical
procedures. We haven't defined those procedures well enough yet and there are many
problems with procedures that we've always considered we 11-proven and adequate. We
don't yet really know how to analyze for ammonia or for cyanides. At least, that's what
SRI found out. Dave, do you agree with that statement? """
Dave Bomberger - Yes, I think so.
F. S. Manning - The RSKERL has measured the analytical variability of 5 wastewater
parameters for refinery effluents. This data can be obtained from EPA Document 600/2-
76-234 (September 1976).
Arthur J. Raymond - Sun Oil Co. - In your talk you talked about your analysis of poly-
nuclear aromatics and chlorinated hydrocarbons. You said the EPA used gc-ms, what
methods are you using?
Judy Thatcher - Well, the W-20 study was started before the toxic settlement agreement
and before the EPA sampling protocol came out, so we used a variety of methods. The
PNA method was developed by Exxon R&E and involves solvent extraction, clean up of
the sample over an alumina column, elution, separation of individual PNA's by gas
chromatography, trapping of each peak, and quantitative measurement by UV spectro-
photometer with results based on ^C labeled internal standards. For the organo-halides,
that involved purging of the volatile organo-halides and passage through a gas chromato-
graph. The detector was a microcoulomenter which only measures halogenated hydro-
carbons. The volatile organics were purged from the sample and analyzed by gc; the non-
volatiles were solvent extracted, I think mainly by carbon tetrachloride. There was some
preliminary gc-ms work done on the non-volatiles. With regard to the W-22 studies, we
have Radian Corp. who is running the gc-ms protocol as specified by EPA. We've made a
couple of changes, one involves the final effluent samples from each one of the refineries.
We'll be spiking each sample with certain of the pollutants. As far as I know EPA is not
doing any spiking of the actual water samples. We will run a final effluent sample and
then spike a duplicate of this, and run that and try to get some feeling for the accuracy
61
-------
of the method. And then in addition we're looking at polynuclear aromatics by the
Exxon method which can measure, I believe, up to 18 individual PNA's. We know al-
ready from preliminary results that the gc-ms method in the protocol does not separate all
the PNA's and so you end up measuring three PNA's under one peak.
62
-------
SESSION III
OPEN QUESTION AND ANSWER SESSION
INDIVIDUAL PROBLEMS IN MANAGEMENT OF REFINERY WASTEWATER
Chairman
* Francis S. Manning
University of Tulsa
Tulsa, Oklahoma
Panel Members
Milton Beychok
Consulting Engineer
Irvine, California
W. Wesley Eckenfelder
Vanderbilt University
Nashville, Tennessee
Davis Ford
Engineering Science, Inc.
Austin, Texas
H.E. Knowlton
Chevron Research Company
Richmond, California
R.N. Simonsen
Standard Oil Company of Ohio
Cleveland, Ohio
''Biography on Page 52
63
-------
OPEN QUESTION AND ANSWER SESSION -
INDIVIDUAL PROBLEMS IN MANAGEMENT OF REFINERY WASTEWATER
Milton R. Beychok
Consulting Engineer, Irvine, California
W. Wesley Eckenfelder, Jr.
Vanderbilt University, Nashville, Tennessee
Davis L. Ford
Engineering Science Inc., Austin, Texas
, *, ,,.. H.SE. Knowlton
Chevron Research Company, Richmond, California
R. N. Simonsen
Standard Oil Co. of Ohio, Cleveland, Ohio
N. Seppi - Marathon Oil Co. - I would like to ask Mr. Eckenfelder if he would discuss
nitrification and go into the prevention of upsets which result in loss of nitrifying bacteria
and possibly talk about methods of recovery from the loss of nitrifying bacteria.
W. W. Eckenfelder - Our experience to date has tended to indicate that dependable
nitrification plus carbonaceous BOD removal can be achieved in the treatment of domestic
wastewater in one stage. Highly variable success has been attained on most industrial
wastewaters. Nitrification is better considered in a second stage after removal of
inhibition in a first stage. The second stage could be either activated sludge or rotating
biological contactor or an up-flow aerobic filter. It also appears that the use of powdered
activated carbon in the activated sludge process tends to greatly enhance nitrification.
The carbon tends to remove compounds and materials which would tend to be inhibitory to
the nitrification process. The case to which I refer while not a petroleum-refining
effluent, indicated that straight nitrification required an average sludge age of about 45
days and that with the addition of the powdered activated carbon, was reduced down to
about 8 days. Again, of course, there is an economic trade off here, but apart from the
economics, having a greater dependability of operation would mitigate towards powdered
carbon or possibly towards a two-stage process as opposed to one.
Davis Ford - I think the question was asked regarding recovery of nitrifiers, and it is a
difficult question, but I might communicate one or two things. The first step, and we've
had some experience in this area, is if you have low nitrification periods which had been
64
-------
preceded by good nitrification, temperature may be the problem because as you know
nitrification tends to drop off in winter climates. When lower temperature prevails, we
try to reseed from plants which have had good nitrification. We've done this down'in the
South Texas area with some success. Of course we know all the other environmental
factors which we can correct, namely the pH sensitivity, the sludge sensitivity, and I
certainly agree with Wes on the sludge age. Take sludge age and look at some of the old
graphs and we expect certain percent nitrification that just isn't there. And that's
particularly true when you have a high concentration of organic nitrogen, TKN, particu-
larly amines. I think in plants wherever applicable, control of amines could be an answer
to nitrification in biological processes. And of course, the third response as far as
correction goes is to get a good lawyer. We get into very strict ammonia nitrogen con-
centration requirements on NPDES permits. It is my opinion that many of the permits that
go into effect on July 1, 1977 are going to be noncompliant in terms of ammonia. So it
takes the combination of a lot of negotiation with regulatory authorities and some of the
steps I outlined.
Eldon Rucker - API - There was some mention this morning about the fact that the refining
industry appears to be well on its way to solving many of the problems of the priority
pollutants. I wonder if any particular members of the panel, or the entire panel, might
comment on whether some of these problems could be transferred into the sludge, and if
this is the case, what are the accepted methods for handling sludges generated in
refineries?
H. E. Knowlton - As a member of the API Solid Waste Committee, one of our concerns is
the new Resource Conservation and Recovery Act which affects handling and disposal of
oily sludges. Many people recognize that land farming is an excellent way of treating
oily solids and biosolids. Land farming of oily solids and biosolids done properly will not
put contaminants into the ground water, there is not a run off problem because you
normally dike these installations so any net water is put back into the wastewater treat-
ment system. The studies so far show that the net water from a land farming area has very
low amounts of contaminants such as oil. We also believe we should recover as much oil
as economically feasible so that the amount going to land farm will be relatively small.
Davis Ford - I might just make one comment with respect to sludges from refineries.
Speaking now as a consulting engineer, you start off with land farming and then back off.
Because of process applicability in many areas in the U.S., successful land farming can
be achieved for oily and biological sludges. Now this normally forces digestion to satisfy
most state regulations, then combining with biological sludges, and then to land farming.
Another comment is that the key here is dewatering. I think technology has certainly
progressed a long way in dewatering of oily sludges. I know.that EXXON and others are
having good success with belt filters, so if you can successfully dewater through belt-
press, or other filters, or other comparable processes, and combine that with the digested
activated sludge, land farming is the place to start. This is far better than less palatable
methods such as incineration, so I think that's our starting point.
65
-------
E. A. Buckley - Lion Oil Co. - On your land farming of the sludge, how do you deter-
mine how much oil you've got say in an alum sludge where you're coming off a DAF
system? Of course you have floating oils which you can separate, but how about
separation of the alum sludge?
H. E. Knowlton - My answer to you is that at Chevron we don't have any alum sludge.
And we don't have any for a very real reason. About 3 or 4 years ago we said, let's go
to the new induced air flotation process; we don't want the problem of handling-disposing
of that sloppy mess. And it is a mess. So we don't have this problem and I can't give you
any solutions for a non-existent problem.
N. Sepp? - Marathon Oil Co. - What's a rule of thumb as far as the barrels of oil per
acre that you can soil farm?
H. E. Knowlton - In a recent paper on land farming prepared by Sun Tech and given at
the CREC meeting in Chicago, they were land farming 600 barrels of oil per acre per year.
I think the answer is, where are you? If you're in an area where it's quite warm such as
in Southern U.S.A. where on a hot day (95 F) we measured 130 F one inch down !n our
soil of the land farm, then you can expect the oil to biodegrade rapidly. This assumes
that your land farm is set up properly. In the northern areas obviously your working time
is shorter and your temperatures lower, so biodegradation potential per acre is less.
Ben B. Buchanan - Phillips Petroleum Co. - I would like to ask the panel if they could
comment on the revolving disc biological process as compared to activated sludge, and if
they know of differences in the types of bacteria used for both processes? Of course the
slime forming bacteria and the revolving disc process I don't know how different they are
from the activated sludge bacteria. Can you have nitrifyers in both systems, or are there
limitations to what bacteria on revolving discs can do compared to what they do in
activated sludge?
R. N. Simonsen - We have had pilot plant experience at one location having a well
operated aerated lagoon system with several days aeration time. We are interested in
getting nitrification .and ran both activated sludge and rotating biological surface pilot
units using aeration basin effluent for feed. Although we got effective nitrification with
activated sludge, there were frequent upsets and recovery was very slow. The RBS unit
in the same application worked quite well. We don't know what differences in bacteria
type there may have been. Dave Rulison who is in the audience and did the work might
comment further.
Dave Rulison - Sohio - Only that in the case of activated sludge the variability of the
feed coupled with the difficulty in retaining nitrifyers in the system meant that small up-
sets would ruin nitrification. With the RBS unit, even significant changes in feed quality
didn't make much difference.
66
-------
H. E. Knowlton - One of your questions was comparison of activated sludge vs. biodiscs.
To answer that, in six Chevron refineries, one is using activated sludge, two are using
biodiscs and three are using ponds. The criteria used to pick a biosystem for these
individual situations was which system will cost the least. I will make one comment on
what we see with one of our biodisc systems that went on stream about December 1976. A
month ago an operator dumped 11 tons of acid into the system. This was not discovered
until morning, by which time the acid removed most of the biosolids from the eighteen 11
foot diameter by 25 foot long discs. The acid cleaned those discs of that material. The
item that really made us quite happy was that at the end of two days we were back meeting
our phenol spec. In other words the discs had recovered sufficiently so that we were
meeting our phenol specification in the final effluent. We thought this was quite remark-
able, because of the severity of the upset. We have found by other experiences that of the
contaminants removed in the biodisc process that phenol removal is the last to recover.
Davis Ford - I would like to make a few comments on the RBS-activated sludge comparison.
Let me use RBS (rotating biological surface) because "Bio Disc" is a trade name and I
don't want to advertise. Let .me say that I have personally come to full cycle about 4 or
5 years ago, inherently I was rather negative on the whole RBS concept. I think today I
have a much different view about that. Chevron Research was partly responsible. Let me
make a few specific comments, and I'll discuss that in more detail Wednesday morning.
First of all, from an operator's point of view and from a cost point of view, RBS or bio
disc has a lot of innate advantages, mainly on power-connected horsepower as compared
to activated sludge. It's much simpler to operate, and I think the capital costs with
BOD's below a certain level - I'm not sure what that level is - but capital costs are
probably more cost-effective or cheaper than the activated sludge. I don't want to make
these positive comments at the expense of activated sludge; that's been the "modus
operand!" in the petroleum refining industry for a long time and it's really the basis for
BPT and will probably be the basis to some extent for BAT in the remaining guidelines. I
believe that as the influent BOD gets higher, let's say 750 mg/l or higher, then activated
sludge might be equally or more attractive than RBS systems. Again, I don't have
specific numbers for this but when you get into high BOD's you'd better check the cost-
effectiveness quite carefully before making the process decision. We designed both RBS
systems and activated sludge so we have no inherent process bias. But let me make a few
comments on questions that I still have about RBS systems. One, I think you have to be
careful when comparing the ability to withstand upsets between RBS and activated sludge
because you're getting full raw waste load impact on those first discs; for example, an
acid spill. Whereas in a completely-mixed activated sludge system the acid equivalent
or potential bio-toxic or bio-static level per bacteria is reduced, it is less than in RBS
systems. I think the jury is still out on the ability to withstand upsets, recognizing too
that you have the ability to operate at very high sludge inventory on any activated sludge
system, and you kind of ride the line on the RBS system. It has been proposed, primarily
by the vendors, that sludge settleability in RBS systems is better than that of activated
sludge. I think that would be the case, although I have yet to see it proven. So, I still
have a question on the effluent TSS ability of an RBS system as compared to activated
sludge. With respect to BOD less than 750 or 500 mg/l, we found that activated sludge
67
-------
systems take out about 1.5 to 1.6 Ibs of BOD per horsepower hour. If you can use that
parameter to determine the power efficiency or power utilization removing BOD, whereas
RBS systems on the paper studies that we have done preliminary to the detailed design
have been about 2.3, so it is more energy-effective. At low BOD concentrations you
have an inherent incentive to consider RBS systems, particularly in land-limited situations.
One other comment - RBS has been proposed by the organic chemicals industry and we're
a bit reticent to accept that system carte blanche when you have COD problems because
your contact time between the bacteria and the sludge in RBS systems is going to be less
than it is, for example for extended aeration. You can solve the BOD problem but
possibly not the COD because the hydraulic contact time is quite low. There is no way
to accurately calculate sludge age, which you can of course in an activated sludge system.
One quick comment on nitrification - I believe that the RBS system offers advantages in
nitrification as compared to activated sludge, because of the ability of the latter stages
in the RBS system to develop a good nitrifying bacteria. And also, it has been quite
successful in adjusting the pH, raising the pH in the latter stages to the optimum level for
nitrification.
Bill Ruggles - Phillips Petroleum (Bartlesville) - With the increasing emphasis on hydro-
carbons in the atmosphere, I am wondering if any member of the panel has experience
with possible effects on this situation that has resulted from the land farming of sludge?
H. E. Knowlton - There are no numbers available on this to our knowledge. In fact, one
of the items on the API budget for next year is to measure hydrocarbon emissions from a
land farming area. We expect very little as we've spread oily sludges and rototilled them
in and have not been able to smell any odor in the area.
E. A. Buckley - Lion Oil Co. - One other question. What are the most feasible as well
as most economic ways of reusing treated wastewater?
M. R. Beychok - There are literally hundreds of answers to that question and they are all
refinery or site specific. The first answer that comes to mind is the reuse of treated waste-
water as cool ing tower makeup. Another answer is the reuse of treated wastewater to
produce low pressure steam (rather than high pressure steam which requires relatively pure
water to avoid excessive and uneconomic blowdown). The low pressure steam might be
used in atomizing the oil fired in process furnaces, or as flare steam, or as stripping steam
in crude unit sidecut reboilers and in other process steam uses. Using treated wastewater
for low pressure steam generation may require 10-15% blowdown to avoid fouling
problems in the steam generation units. If you attempt to reuse treated wastewater for
high pressure steam generation, you can't afford such high blowdown rates and you will
need tertiary treatment following your secondary treatment to make the water suitable for
high pressure boiler feedwater. As you may know, the API CREC funded a pilot program
studying the reuse of treated wastewater in cooling towers to take advantage of the
evaporation in the cooling tower so as to result in a more concentrated waste in the form
of the cooling tower blowdown.
68
-------
E. A. Buckley - Lion Oil Co. - What parameters would you put on the IDS in evaporated
cooling water blowdown?
M. R. Beychok - It depends upon the cycles of concentration at which you operate the
cooling tower. It also depends upon the economics and whether you can justify a slip
stream filter or a slip stream softener on the circulating cooling water. Some refineries
have so-called "clean water" and "dirty water" cooling towers which adds another
parameter of flexibility for reusing treated wastewaters.
H. E. Knowlton - I can give you an example of reuse recently completed which worked
out a lot better than we hoped. This refinery had 800 GPM of effluent; it now reuses 400
GPM of filtered effluent 250 GPM into the cooling towers and 150 GPM into the process
water fire-water system. We've been running only a month and a half or so but this has
worked out very well. We see no problems so far and the effluent is reduced to 400 GPM.
We do not have a lot of analyses as yet to tell us if anything is unusual, but we don't see
anything. Also we're surprised, we thought that maybe the calcium might double but so
far we can't see it changing much. Some things happen in a practical system that when
we plan we don't anticipate. Other refiners, by the way, have used API-separator water
in their process water and fire water systems for 25 years. Their water is very poor
quality compared to that we are reusing which has a BOD of less than 10 PPM and an oil
and grease content of less than 10 PPM.
R. N. Simonsen - We have used effluent in fire water systems, but I should relate what
happened when this was tried at one refinery which has since been shut down. Effluent
had been used to pressure the fire water system for several months. Then, during a plant
fire fighting training exercise, attempts to extinguish a fire with a foam generator failed
because the powder would not make foam. New powder from the storehouse also failed
to make foam. This was how the refinery discovered it had been operating several months
without fire foam protection. Neither we nor our supplier ever learned why and, of course,
the refinery changed back to its former fire water supply. A number of refineries use
stripped foul condensate and blowdown from oily cooling towers in desalters as another
form of recycle.
M. R. Beychok - This isn't another specific example, but rather a general caution. When
designing any process unit, you must be very careful about recycling. You must avoid a
closed loop from which there is no way to bleed out any buildup of a recycled impurity.
The same thing holds true for reuse of treated wastewater. You don't want to recycle in
such a manner that you create a closed loop and build up an intolerable level of dissolved
solids or other contaminants. So you want to thing in terms of cascading systems. If your
treated wastewater originated from steam condensate, reuse it to generate steam which
will be used in other services. Or if you reuse the water in a cooling tower, be sure not
to lock yourself into a closed loop which has no way to bleed out or blowdown impurities.
Ed Sebesta - Brown & Root, Inc. - I would like to go back to the subject of biological
rotating discs reactors, and mainly some comments on the maintenance - good or bad, on
some of these systems that have been in operation for some time.
69
-------
R. N. Simonsen - We haven't had any maintenance problems since oOr full-sized RBS
units haven't been installed yet.
H. E. Knowlton - We have had one in operation for about 2Ł years, it's a styrofoam disc
unit. We've had a few things happen, like the carbon steel nuts on the ends of tie rods
corroded off, and then the styrofoam discs moved out of position. One of the problems
with rotating disc units is that people tend to never look at them, tt sits there and runs
with very little attention. Mainly one tries to control feed quality to it because what
you put in determines what comes out. So the main effort is to keep a reasonable plant
control on your input quality. We suggest now that every three months they go to and
actually check eqdh disc and make sure that nothing is wrong. There should also be some
surveillance. The operator should walk by each of the discs once per shift ro see if there
is noise or anything indicating a problem. A second unit we have in operation has been
operating roughly six months. We see no problems there; it is a large system of 18 shafts.
There was a recent NPRA paper on rotating discs; (AM-77-27), Refinery Use of Rotating
Biological Surfaces in Waste Water Treating. In this paper we gathered the available
information on four existing commercial size installations that were! processing refinery
waste water in the U.S. in February 1977. You can ask Herb Bruch, NPRA Technical
Director, for a copy. There is a fair amount of detail in it on this subject. Are there any
other questions on this one ?
Ed Sebesta ••* Brown & Root, Inc. - A follow up while you're up. When ypU did have that
problem with the one at Salt Lake City/ to redo it how did you operate while you're dding
repairs, or what did you have to do to repair it?
H. E. Knowlton - They stop the rotation and bypass the water to the next shaft. Then
they patched the styrofoam discs that were torn. You can by-pass these things. This
system is set up.with four shafts in a row and has two rows; if necessary you can by-pass
a whole row, but normally it isn't necessary to do that, you can just by-pass the one
involved and work on it.
L. D. Erchull - Union Oil of California - I'd like to address this question to the panel as
a whole. Oh nitrogen removal from refinery waste water streams after they have been
steam stripped, are there any installations that you know of where some proven technology
has been applied where it can work year round, say in the northern climates of the country?
R. N. Simonsen - Are you talking about treating stripped foul condensate by itself? We
hqveh't tried that,-but nitrification of refinery wastewater containing stripped foul con-
densate can be accomplished with the RBS in northern Ohio. But temperdtute is Very
important and the colder it is the more surfdce is needed.
Davis Ford - I would like to add one more comment to that. I mentioned a case history
where the nitrification has been consistently 85%. I might mention that the ammonia
coming in about 10-15 g/liter, that's a pretty low concentration for a raw wdste load
going through a biological system just to put that in perspective. I'm hot sure that we
70
-------
could get that much nitrification if we had higher concentrations coming in. Another
thing that we haven't mentioned yet on the RBS activated sludge system is that you've got
to remember that the heat loss in an RBS system is much, much less than in activated
sludge, which is in effect a cooling tower with mechanical aerators. So, that would give
us an added advantage on nitrification which is so temperature-sensitive in the northern
climates, using RBS compared to activated sludge. With respect to pH, it has been indi-
cated that about 7.5 Ibs of alkalinity as calcium carbonate is destroyed per Ib of nitrogen
nitrified. I think perhaps we would have a little better pH control in an RBS system using
the approach I discussed earlier where you add alkalinity to the latter stages that have
already experienced an initial pH drop through the production of CO«. So even though
you have a pH problem in both activated sludge and RBS, you might have a little more
positive control on the RBS system by controlling pH through the various stages.
W. W. Eckenfelder -,| think one of the benefits of the rotating contactor for nitrification
as opposed to activated sludge stems from the fact that you will generate roughly 50 Ibs of
bio mass per 100 Ibs of BOD removed, but only about 15 Ibs of bio mass per 100 Ibs of
nitrogen oxidized. < What this means is in a completely-mixed activated sludge where both
BOD and ammonia are to be oxidized the population of nitrifyers is going to be very very
low; in the order of probably 1 to 4%. In a rotating contactor, admittedly in your initial
stages where you are primarily rmoving carbonaceous organics you will have a very very
low population of nitrifyers, but once the carbonaceous organics are essentially gone,
then 40 to 50% of the bio mass will be effective nitrifying culture. This provides a
cushion against both changes in concentration and also changes in operating temperature.
Where nitrification is important, and this refers to my earlier comments on one stage vs
multistage activated sludge, that you should consider the relative concentration of
nitrifyers as opposed to the concentration of other organisms. With respect to pH, 7.15
Ibs of alkalinity are required per Ib of nitrogen oxidized. Normally the amount of
nitrogen to be oxidized should not probably pose a major pH problem. But it could be a
problem if the available alkalinity is low and your amount of nitrogen to be oxidized is
high.
H. E. Knowlton - I would like to make a comment to the Union man. If we find we have
an ammonia problem we go back into the process units and find the ammonia sources.
Then we can strip it better; or reduce the volume of the stream before stripping it. We've
done both and we end up with a significant NHo reduction. We also find high content
NhL streams going into waste water treating from sources not previously recognized. We
of course send them to NHk recovery.
!;.." i •
M. R. Beychok - Well, Buzz beat me to it because he didn't play tennis this morning and
he can get up faster than I can. But I want to second his comments. I think that very
few refineries really have good nitrogen balance data across their wastewater systems.
The first thing is to find out where your,ammonia is coming from and find out what ammonia
containing streams are bypassing your stripper. Then.you should provide better ammonia
stripping or reduce the volumes of those streams or both. The end-of-the-pipe treatment
for nitrogen ought to be a last resort. ,
71
-------
Davis Ford - Just this comment - I don't want to mention names, but there certainly are
plants. One plant in particular that I'm thinking about was not designed for nitrification
but its nitrification level has been consistently at about 85% through the whole 12 months.
It is in the Southern part of the U.S. so the temperature is favorable. Right next door,
however, there is a problem in nitrification. To get back to my earlier comments, there
are so many variables - specific toxicity of certain organic compounds, etc. - so that it
is just not that predictable. But there are case histories of good nitrification throughout
the year. You're from Union, I guess. I know when you get up around the Chicago area
and you have a significant drop in temperature, that tendsQto play havoc with nitrification
systems - and I'm talking now of just around 15, 10, or 8 C.
Bob Carloni - Lion Oil Co. - We've just recently put on a treatment plant which uses a
mixed media gravity filter and we weren't getting very good solids removal until we
started using a polyelectrolyte. Is that normal experience? If it is what sort of concen-
trations of polyelectrolytes should we be considering? We're currently using 2 ppm and
we've heard that we should be down to around .5 ppm or lower.
Davis Ford - What type of system are you talking about?
Bob Carloni - This is a filter which follows a clarifier which follows an RBS system.
Davis Ford - The first comment is, yes, polyelectrolyte is being used to keep people out
of jail all over the country right now. Concerning control of effluent TSS levels without
filters, and I can cite you example after example, it is not free. In some cases the con-
centration requirements are 2, 4 and 5 mg/l and I know the polyelectrolyte salesmen are
going to be happy. It is about $1.00 - $1.50 per Ib, so if you have a high flow it's an
expensive operational procedure. It's been working quite well, however, just for final
clarifiers where just the poly is added to the activated sludge or RBS effluent and is mixed
going to the final clarifier. With respect to post filtration, it is my opinion that you have
got to have at least a poly addition capability prior to post filtration. Once you get into
the operation you might want to back off, but as a capital expenditure you'd better have
the ability to add a coagulant or coagulant aid to enhance filtration. By the way, the
enhancement of effluent TSS quality by virtue of adding poly, which is normally 1 - lŁ
ppm, but sometimes a little higher, has often been around 30% to 50% improvement in
equivalent TSS.
H. E. Knowlton - Chevron has only one refinery with a filter on a final effluent. On
this one final filter we do not use polyelectrolyte. As I recall we get roughly a 50% drop
in BOD. We possibly could get more removal with polyelectrolyte but we don't use it
because we have no need. We see these results of the filter—approximately 50% oil
reduction, and 30% BOD reduction. We run this filter on a time cycle only because we
find that the AP control system was not practical.
R. J. Churchill - Tretolite - We're one of the polymer companies that are getting rich
according to Davis. Polymers are useful but only as a tool. They are not a panacea.
72
-------
y^e've been called into places where people have bulking type problems and ask us to add
a polymer to solve that. You're treating a symptom you're not solving a problem. We
have had success in the application of filters by going into the clarifier, that is adding to
the mixed liquor on the way to the second clarifier, which does two things. One, it
knocks down the TSS leaving the clarifier going to the filter; second, almost uniformly or
almost characteristically we have seen it knocks down the variability so that you don't
slug the filter with a high TSS concentration episodically. That does two things and it
reduces the total solids load put onto the filter which should extend your run life. It
should also then give you better net TSS out of the filters since you're starting with a
lower TSS in. It also may turn out to be more economical to go to a split addition rather
than trying to do it all in one step. I think the key to the polymer is simply to use it as
a tool that is to take out the fluctuations and the total amount of TSS leaving the clarifier
going to the filter. To answer the point about one-half a part or 2 parts or 10 ppm, we've
treated systems with less than a part very low concentrations at the secondary clarifier.
We've also treated refinery systems as high as ten parts at the secondary clarifier. I think
that relates back to some of the points that Buzz was mentioning. You can have bulking
or nonsettling type sludges that's not a problem to be solved by a polymer. You can oil
carry over into your biological system and into the clarifier which affects settleability.
Again, that's not a polymer problem. It's a process problem or a waste management
problem. I think it's hard to draw any hard fast rules but excluding my bias I do think it's
a tool, but it's a tool that only can be used wisely.
Rich Sheridan - Brown and Caldwell - I think it was Mr. Ford that mentioned the appropri-
ateness of the bio discs in situations where nitrification would be required in cold climates
and therefore the bio discs have the advantage of less heat loss. I would like for someone
to comment on the requirement to nitrify in cold climates, as to its appropriateness, cost
effectiveness, these types of things.
W. W. Eckenfelder - I think it's obvious to all that in most cases involving cold climates,
as far as water quality goes, nitrification is neither justified nor is it required. My own
opinion is that these cases should consider a two tiered permit in order to avoid a large
economic penalty for nitrification in cold weather conditions. As far as water quality
goes, the fact that the process of nitrification radically slows down in the treatment
process, it equally slows down In the receiving water.
Eldon Rucker - API - In view of the existing widespread use of lagoon systems in the
refining industry, would the members of the panel comment on the efficiency of these
systems compared with some of the RBS and activated sludge which have been previously
discussed ?
M. R. Beychok - It depends upon what sort of lagoon system you're talking about. I will
be very specific and talk about surface-aerated systems using mechanical aeration as
against those that depend simply upon photosynthesis. I'm talking about lagoons with 3-
4 days retention, 12-15 feet deep, and using up to 50 hp per million gallons of retention
for mixing and aeration. About 4-5 years ago, an AlChE committee surveyed existing
73
-------
aerated lagoons In the refining, petrochemical and the pulp and paper industries. I don't
know if Wes Eckenfelder remembers it or not, but one of his colleagues reviewed our final
conclusions from that AlChE survey and disagreed with us. In any event, the data showed
that 85-90% BOD removal could be achieved in those lagoons, within the parameters
mentioned: 12-15 feet depth, 3-4 days retention and up to 50 hp per million gallons of
retention. That contrasts with about 90-95% BOD removal for activated sludge systems.
I'm not sure what the average percent BOD removal in an RBS system is, but I would
imagine it is somewhere between the aerated lagoon and the activated sludge systems.
W. W. Eckenfelder - I think one thing is pertinent and that is where you have an aerated
lagoon system, or a waste stabilization pond, you are going to be much more subject to
temperature effects during the colder weather than you would be from either an RBS or
activated sludge process. The variability month by month would generally be higher from
an aerated lagoon than it would be from those processes that tend to minimize heat loss
during the colder climates. I would agree from all of the data that I have seen that a
properly designed and operated aerated lagoon process should be capable of doing about
the same thing as the other processes relative to BOD removal. One of the obvious
problems with an aerated lagoon is the effluent suspended solids.
M. R. Beychok - To further qualify my earlier answer regarding 85-90% BOD removal in
aerated lagoons, we adjusted for the variability of temperature by using a power law
function. We converted all of the data to a base temperature of 77 F. We also
recommended that aerated lagoons be designed for the coldest temperature in order to
achieve the required BOD removal during the winter. Then one could take advantage of
the oversized system capabilities during the summer months to schedule maintenance and
turnarounds.
John C. Doolittle - Shell Oil Co. - Relative to the aerated lagoons, what effect would
an occasional oil film on the lagoon surface have upon BOD removal?
M. R. Beychok - I suspect it would depend upon whether you have 15 hp or 50 hp per
million gallons of retention. The key thing in my opinion is mixing and not aeration. A
high degree of mixing will offset any effect that an oil film might have on oxygen
absorption. 50 hp per million gallons of retention will more than provide the oxygen
required.
H. E. Knowlton - Eckenfelder said that the lagoon system is very dependent on temper-
ature. And in certain areas of the country the lagoon system wouldn't be effective. This
also has been our experience. In the southern part of the U.S. or Hawaii we find that the
lagoon system is excellent even though we have some TSS problems. In fact in one
refinery we exceed comfortably "77" permit conditions with a lagoon system. I would
like to mention that in the first refinery in the U.S.A. to use biodiscs, we did have a
lagoon system. The only reason we put in biodiscs was that in the winter time when the
effluent water was 33 F, we couldn't get enough BOD reduction. We put in the bio-
discs and even though the water into the biodisc is at 50 F, it drops only about 1 F at
74
-------
an ambient temperature of 20 F. In our second installation during the severe cold
weather of last year, we saw only a 6 to 8 drop across the bio disc system with ambient
temperatures of 0 F which is a pretty minimal drop. We did not suffer at all from the
severe cold weather. This is one of the pluses we see for bio discs.
Davis Ford - I have one other comment on the aerated lagoon waste stabilization pond.
I'm not really sure that your question is on waste stabilization ponds or aerated lagoons,
but let me add a postscript to the aerated lagoon discussion. I really agree, and some-
times it's difficult to convince regulatory authorities, that the aerated lagoon system is a
BPT system in many cases, and I really believe that. Of course, this depends on your
raw waste load and how far under the guideline number you are on RWL. The TSS comes
back to haunt us. It's going to be interesting to see how the TSS numbers come out in the
remanded BAT guidelines for the petroleum refining industry in September or October,
because that's really the key to how applicable these systems are going to be. In many
cases an aerated lagoon or waste stabilization pond concept can produce, from refinery
wastewater, a much lower COD than can an activated sludge system. So there are
inherent advantages. We just have the problem of algae proliferation and biological TSS
to contend with, and in effect the remanded guidelines and the implementation thereof
will dictate just how applicable these systems are. In addition of course we have some pH
problems with waste stabilization ponds; 6 to 9 being the normal pH on those permits; and
as we undergo photosynthesis using CO« the pH goes over 9. We have been successful,
and others have too in some cases, in having a time stipulated on the permit as to when
the pH is to be taken, namely, early in the morning to exclude that CO« utilization by
algae.
E. G. Kominek - Environtech Corp. - It is apparent that there is a lot of interest on the
part of this group in discussing nitrification, and it has been my experience recently that
in many areas that denitrification is becoming more and more of a problem. Is this some-
thing, in the opinion of the panel, that the refineries have to be thinking about at this
time?
Davis Ford - I don't think so unless your effluent discharges into a drinking water source
so that you'll have to comply with the Safe Drinking Water Act which has, I believe, a
10 mg/l as N maximum. If you discharge into a body of water that is used for drinking
water it could be a problem. None of the permits we're involved with have nitrates on it
and I don't anticipate any. So given that drinking water exclusion I don't think you'll
have any problem there.
R. N. SImonsen - One other comment on aerated lagoons. I believe the results of a
recent API survey on cyanides in refinery effluents shows that the lowest concentrations
are from refineries with aerated lagoons or long residence time oxidation ponds. This
parameter is one of the 129 toxics and could become a serious problem for refineries dis-
charging to low flow streams.
75
-------
F. J. Kuserk - Texaco - We have an activated sludge system that does a very good job
with most parameters, except we do have one problem with it. It seems to increase the
one pollutant, hexavalent chrome. I was wondering if one of the panelists could tell us
what's happening here? We're coming out of our API separator at hexavalent chrome
levels of less than 20 parts per billion, we go through our waste water treatment plant and
we're coming out in the range of 60 to 80 parts per billion.
M. R. Beychok - I don't have an answer to that. In fact, it amazes me because I recall
that the API -EPA 1972 raw waste characterization surveys actually showed a decrease in
total chromium and heavy metals across activated sludge units. Are you in effect saying
that some total chromium is being created?
F. J. Kuserk - No we're not creating total chrome we're just going from the trivalent
state to the hexavalent state.
Robert Carloni - Lion Oil Co. - Are.any of the panel members familiar with any
installations using ozone to improve the treatability of refinery waste waters?
R. N. Simonsen - I don't know of any in the U.S., but when Cities Service operated
their Trafalgar refinery in Ontario, ozone was one of a great number of treatment steps
used. I don't believe ozone is being used there now.
Ralph Churchill - Tretolite - Could I get some discussion from the panel relative to the
chrome problem and the alternatives to meeting chrome guidelines, namely chrome
removal from cooling tower systems, or closer control of chrome in cooling towers, or non-
chromate inhibitor programs?
H. E. Knowlton - Ralph, I can tell you what our status is. We do not see any problems
with hexavalent or total chrome contaminant levels in our effluents. The hexavalent
chrome is reduced and total chrome removed by our waste water treating system operation.
So we don't really see the need at this point of changing corrosion inhibitor or installing
chrome removal systems. Especially if we can reduce our final effluent to a trickle, in
the future, which we feel is desirable to get out of the clutches of EPA.
R. N. Simonsen - We have a refinery and a chemical plant side by side at one location.
Both have similar water supplies and cooling towers and both used chromate inhibitors.
Chromium is reduced and seems to drop out to some extent in the refinery system and is
really not a problem. Chrome has been a problem at the chemical plant and it has been
necessary to switch to non-chromate treatment.
M. R. Beychok - I would like to respond for those cases where you do have a problem.
For example, industrial plants in California are faced with a coastal water discharge
regulation limiting total chromium to 0.01 ppm (10% of time) and 0.005 ppm (50% of
time). The treatment alternatives to consider include chemical reduction and precipitation
which creates a gelatinous sludge which is as difficult a problem as the original chromium
problem. The other alternative is ion exchange, of which there are two types which
76
-------
might be used. One ion exchange process is the fixed-bed type which recovers the
chromium for reuse in the cooling water. Cities Service has such units in a petrochemical
plant in Louisiana and I assume that they are working well. The other ion exchange
process is a fluidized bed system with the ion exchange resins regenerated in a U-bend
type system. As for using non-chromate corrosion inhibitors, it has been my experience
that the petroleum industry doesn't think that the phosphate or zinc inhibitors work as well
as chromate.
Arthur J. Raymond - Sun Oil Co. - Is there any data available on the correlation of BOD
vs. TOC analysis?
Davis Ford - To answer that, yes, a lot of it but it's very site specific and many things
influence that correlation. First, just the severity of the process will have one impact on
the ratio and then the degree of treatment through a biological system will have another
impact. So I would say in order to be substituting TOC in terms of BOD you will have to
develop that correlation for your own site. You get into a lot of trouble by extrapolating
that correlation to your own site and accepting a permit on that basis.
J. Dewell - Phillips Petroleum Co. - Does anybody have an experience with a new vendor
process called dispersed air flotation as compared to what's called induced air flotation,
and do the same chemical additives work?
H. E. Knowlton - We have a couple of these in operation. For four years we've been
using induced air flotation in various refineries right after the API separator. I think the
dispersed air flotation uses less horsepower. You might get slightly less removal too, but
you still use the chemicals. We see it as simpler to operate and maintain than induced
air flotation.
J. Dewell - What size of air bubbles do you get with the induced air flotation?
H. E. Knowlton - I haven't worried too much about that detail since the system works so
well in practice.
John Byeseda - Tulsa University - I am working on induced air flotation now and in general
the bubble size depends on the surface tension between the liquid and the vapor. If
you've got anywhere from about Ł% of salt the bubble size will be in the range of 1 mm.
For dissolved air flotation you'll get somewhere around 100 microns practically independent
of the interfacial tensions. And for dispersed air it's about the same size as induced air -
it will be in the range of 1 mm, if there's very little salt in the water the bubble size will
go up to about 2 or 3 or 4 millimeters. There is approximately an order of magnitude
difference, about 100 microns.
N. F. Seppi - Marathon Oil - I had another question about soil farming. I was wondering
how important is fertilization to the rate of degradation, and what effect do metals have
on the bacteria degradation of oily sludges?
77
-------
H. E. Knowlton - There are some Shell reports out on this, on both topics. To fix the
application rate you'd have to check the soil and see how much you need. In fact in one
of our refineries we did not add any nitrogen and still got sufficient activity. So, it's a
matter of finding out how much is there and calculate how much you need to add if any to
achieve a desired result.
John R. Kampfhenkel - Sun Oil Co. - I have a question for Mr. Kuserk of Texaco. Are
your samples on the chrome taken or analyzed on a composite sample?
F. J. Kuserk - The last samples we had were done by an outside analyst.
John R. Kampfhenkel - And you got the same increase in your hex chrome? We've found
down in Corpus that we were getting the same problem but we were getting it on a
composite sample. It seems that we were getting a reduction of our hex Cr chrome in our
sample bottles. Our influent water had a larger amount of oil in it naturally than our
effluent did and we were getting reduction because of the oil in our sample containers.
On a grab sample we didn't find that to be true, we found that the hex in and out was the
same. I don't know whether it was just typical of our particular facility or not but we did
find that to be the case in the sample container because we went back and spiked some of
our samples and found that was what was happening.
Robert Carloni - Lion Oil Co. - My question has two parts. First, what is the state-of-
the-art now in automatic pH control systems? And, second, would a system which con-
sisted of one well-mix basin with feed-back control be satisfactory? In this particular
situation we could tolerate a certain amount of swing in the pH of the effluent from the
control system.
Davis Ford - I can't really answer that question specifically, but yes, the state-of-the-art
is feed forward, feed back in pH control systems, but I have yet to see one work very
well. Every time you go to the plant, the first thing the operator cusses is the pH control
system, so I guess that is the state-of-the-art. A two-stage control system is good when
your titration curve indicates you can fine-tune on the second stage and satisfy most of
the neutralization requirements in the first stage. That probably helps. With respect to
swings in the PH of biological systems, of course you're going to have a normal swing
there. As for the petroleum refining industry, I believe pH compliance has not been a
major problem except for the algae effects, at least in the plants I've been involved with.
We have pH swings - you know what you normally have in wastewaters, unless you have
an acid dump or acid spill - and you try to control it in an off-spec basin or you try to
control it at the source. I just haven't much experience with pH excursions from the 6 to
9 permit limits.
78
-------
BIOGRAPHY M.E. Knowlron
H.E. (Buzz) Knowlton is a Senior
Staff Engineer of Manufacturing Process En-
gineering, Chevron Research Company, Richmond,
California. He served in process design, process
development, technical service and process planning
at Chevron. Previous experience includes Chief
Process Engineer at Chevron East Refinery, Perrh
Amboy> New Jersey from 1967-1972 when he
assumed his present position assisting corporate
refineries on environmental problems. He holds a P
a Ph.D. degree in Chemical Engineering from
Ohio State University. Mr. Knowlton is also
a member of the API committee on Refinery Control
and Chairman of the API Solid Waste Management
Committee.
BIOGRAPHY Robert N. Slmonsen
Robert N. Sfmonsen is Senior Environmental Con-
sultant with The Standard Oil Co. (Ohio) where he has
been involved in environmental affairs since 1950. He is
a past chairman of API's Committee on Refinery Environ-
mental Control, and is a member of API's Environmental
Affairs Water Quality Committee. He received his B.S.
in Chemical Engineering from Lehigh University.
Milton Beychok Blogrdphy on Page 300
W. Wesley Eckehfefdet Btograghy on Page 476
Davis Ford Biography on Page
79
-------
BANQUET
Master of Ceremony — Dr. Allen R. Soltow
Dean of the Graduate School and Director of Research
The University of Tulsa, Tulsa, Oklahoma
Allen Soltow; "Contrary to your program, my name is Allen Soltow. I am pinch-
hitting for Dr. John Dowgray, who had to attend a funeral in Kansas City. First, as
representative of the University of Tulsa, it is my privilege and pleasure to welcome you
to this obviously successful Open Forum and to this banquet. Before I introduce the head
table, may I tell you how much the University of Tulsa values its good relations with the
petroleum industry. We have taken two recent steps to expand and enhance these
relationships. First, we have created a National Capital in Energy-Law and Policy
Institute (NELPI). We are one of the few university institutes addressing the new and very
complex field of energy litigation. Dr. Kent Frizzell, former Secretary of the Interior in
the Nixon and Ford administrations, is the director of NELPI. Second, we have formed a
Petroleum and Energy Research Institute (PERI) which serves as an umbrella organization
for extensive research activities in petroleum production and refining. Frank Manning is
our PERI director.
As your M.C. for this banquet it is my pleasure to introduce the
head table. Starting at my extreme right, that is your left: -
Milton Beychok, Consulting Engineer, Irvine, California
Judy Thatcher, Representing the API, one of the sponsors of this symposium
Bill Galegar, Director, R. S. Kerr Environmental Research Laboratory, Ada,
Oklahoma. Bill, as you remember welcomed us this morning.
Bill is also representing the EPA
Dale Kingsley, Our banquet speaker
Herb Bruck, Technical Director, N.P.R.A. Herb is representing the NPRA,
who is, of course, one of the sponsors of this symposium
Frank Manning, Professor of Chemical Engineering at Tulsa University. Frank is
Project Director of the EPA Grant financing this Open Forum.
May I also acknowledge two other people not at the head table. Ridgway Hall, Associate
Geneml Counsel for Water, EPA, and 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. Fred deserves the credit for engaging the speakers at this
Open Forum."
80
-------
BIOGRAPHY Allen R. Soltow
Dr. Soltow holds a B.A. in Mathematics and
Business Administration from Luther College, and M.S.
and Ph.D. degrees in Economics from Iowa State
University. Dr. Soltow is currently Dean of the Graduate
School and Director of Research. Prior to this, Dr.
Soltow has been Chairman of the Division of Economics,
Acting Dean and Acting Associate Dean of the College of
Business Administration at the University of Tulsa. Dr.
Soltow has been most involved in both national and local
community activities and also participates frequently in
both local and national economic conferences.
81
-------
SPEECH PRESENTED AT THE WASTE WATER SYMPOSIUM
Dale L. Kingsley
Vice President of Refining, CRA, Inc.
Thank you, Mr. Chairman, and I appreciate the honor to visit with this group
tonight. It is gratifying to see this large attendance, and it strengthens my belief that
we in business need to meet and exchange ideas, and to determine the best way to
effectively plan and act to improve the quality of our environment.
We at CRA, and I believe the other refiners that are represented here tonight,
build plants to supply energy at a competitive cost, but at the same time we realize that
these facilities must operate in a manner that will not degrade the quality of air and
water in our neighborhoods. Industry must be a good environmental neighbor because it
is not only good business, but it is also our public responsibility. We continually meet
and talk about the environment as if it were a new world, but actually the earth has been
in a changing climate since its formation. It is only in the last two decades that it has
become obvious that man must correct his wasteful ways if we are to have a good lifestyle
in the future. Since man responds slowly to change, laws have been passed that will
affect our destiny. One early law — The River and Harbor Act of 1886 — gently started
us on the road to improved environment. The Federal Water Pollution Control Act of
1948 was another milestone along this road, and the 1972 Amendment to the Act of 1948
was the atomic blast that really got our attention. We now respond to EPA, and Public
Law 92-500, and others that provide guidelines and goals that we strive to meet. The
requirements on waste water quality were established to help, aid and protect you and me.
But, is it really working out this way? And, can we afford this protection? I am
speaking of "we" as industry, as a country, and as individuals. I suggest we immediately
take a serious look at the pot at the end of the rainbow as related to the 1985 water
quality regulations. I am wondering when we get there if there will really be a pot of
gold at the end of the rainbow, or will it be a pot of I. O. U.s? The cost of zero
pollutant discharge will not be cheap. Since this is going to be a monumental cost, who
is paying the bill? I have looked for the blank check to handle our cost, and have tried
to analyze who is going to honor it. I conclude it makes no difference if it is paid for by
industry or by government — in the end, I am paying for it, and I am starting to wonder
how much non-pollution I can afford. Have you looked at it under this light?
Our industry sponsors a national organization known as the American Petroleum
Institute. Through this membership, data are collected recapping the industry's efforts,
costs, and benefits received, relative to waste water management. The API summary for
1975 indicates the U. S. petroleum industry spent 629 million dollars for waste water
82
-------
treatment plants, and 356 million dollars for operating costs on these plants. This is
about 1 million dollars a day for operating a system that does not bring revenue to the
industry. I don't know where this operating money comes from, but I feel if the industry
is to survive, this fund will come from products sold. Currently the industry refines about
16 million barrels per day of crude oil. On this basis, each barrel of petroleum products
sold is increased in value by 6$ just to cover the water treatment cost, and this is only a
starter. CRA has spent several million dollars in the last two years to construct waste
water treatment facilities, and I am sure all of you have had comparable expenditures.
Many of these facilities will meet and slightly exceed the 1977 standards, but they do
not, and let me stress, do not meet the 1985 standards. It is risky to predict what con-
struction and operating costs will be by 1985, and the degree and amount of equipment
that must be purchased if the industry meets the pollutant-free effluent standards.
Information developed by Batelle in Columbus, Ohio predicts the U. S. petroleum
industry will be spending at least 10 billion dollars a year — possibly as much as 17
billion dollars annually by 1985 to meet environmental regulations, and these are 1975
dollars. If the industry is processing 20,000 barrels per day of crude in 1985, the
consumer cost for environmental improvement can be as much as 5%$ per gallon on
today's energy requirements, and again, this is on today's dollar. This could be inflated
to 10 to 12$ per gallon or more by 1985. The cost exposure of municipal water treating
systems and other heavy industries can be equally as great, especially to the user-
consumer. The Brook ings Institution estimates that compliance with proposed amendments
to the Clean Air and Clean Water Acts by 1985 could cost this nation at least 500 billion
dollars. This number is so astronomical it is hard to comprehend, but it works out on the
average to nearly $2,500 per man, woman and child in this country.
I am not proposing that we cancel our efforts to improve the environment, but
let's look at it with common sense. We should continue to improve our water quality —
that is a must! but do it in such a manner that we reserve capital funds for productive
industrial growth, maintaining adequate employment opportunities, and preserving our
quality of lifestyles, and still have an acceptable, healthful atmosphere in which to live.
This is possible if Congress and others who are directing our National commitments
redirect their efforts to programs that are not wasteful, but are directionally correct, and
lead us to where we really want to go. We in industry are morally responsible to provide
this guidance and leadership. We need to talk to our employees — no, not the ones in
the plants, but the ones in Washington that we are supporting ~ the ones whom we have
elected and the ones who are working for us. I encourage you to redouble your efforts
along these lines. Our responsibilities are great and it will take lots of hard work on our
pant to correct the situation at hand.
You have a busy week scheduled. I wish you success in your meetings. I hope
you will continue to think and direct your efforts toward the amount of pollution abate-
ment that we can afford and really need.
83
-------
BIOGRAPHY DaleL. Kingsley
Dale L. Kingsley, Vice President of Refining,
CRA, Inc., holds a B.S. in Chemical Engineering from
the University of Missouri at Rolla. He started working
at CRA, Inc., a subsidiary of Farmland Industries, in
1952. He currently is responsible for the operation of
CRA refineries located in Coffeyville and Phillipsburg,
Kansas, and Scottsbluff, Nebraska as well as five gas
processing plants.
84
-------
SESSION IV
ORIGIN AND INTERPRETATION
Chairman
Marion Buercklin
Sun Oil Company, Inc.
Tulsa, Oklahoma
Speakers
Ridgway M. Hall, Jr.
"Regulation of Problem Pollutants Under the Federal
Water Pollution Control Decree"
W.M. Shackelford
"Evolution of the Priority Pollutant List
from the Consent Decree"
R.W. Dellinger
"Incorporation of the Priority Pollutants into Petroleum
Refining"
Leon Myers
"Generating Problem Pollutants Data for the EGD
Document: Refining"
A. Karim Ahmed
"Considerations for Defining Substances Hazardous to
the Environment"
85
-------
BIOGRAPHY Marion A. Buercklfn
Marion A. Buercklin Is Southwest Region
Coordinator of Environmental Affairs for Sun
Company, Inc. He has a B.S. degree in
Chemistry from the University of Arkansas. He
is currently chairman of the Oklahoma Refinery
Waste Control Council and is active on several
API task forces.
86
-------
"REGULATION OF PROBLEM POLLUTANTS UNDER THE FEDERAL
WATER POLLUTION CONTROL ACT: THE 1976 CONSENT DECREE"
Ridgway M. Hall, Jr.
Associate General Counsel for Water
U.S. Environmental Protection Agency
Washington, D.C.
I. INTRODUCTION
The U.S. Environmental Protection Agency is presently engaged in a major
regulatory program under the Federal Water Pollution Control Act to limit
the discharges of harmful or toxic pollutants to the Nation's waters. The
strategy which the Agency devised to implement this program is reflected in
a consent decree entered on June 9, 1976, by the United States District
Court for the District of Columbia, NRDC v. Train. 8 E.R.C. 2120 (D.D.C.
1976), which settled four lawsuits against the Agency relating to the regu-
lation of harmful pollutants. By the terms of this decree, EPA has promul-
gated toxic pollutant effluent standards for six highly toxic pollutants,
and is developing effluent limitation guidelines, pretreatment standards,
and new source performance standards for 21 major industries, covering 65
harmful pollutants.
The 65 pollutants were selected by an EPA task force,* assisted by
outside consultants. The selection criteria emphasized primarily the car-
cinogenic, teratogenic, mutagenic, or other toxic properties of the com-
pounds to humans or important aquatic organisms, as well as their presence
in industrial discharges. The 21 industries were selected based upon data
gathered through the National Pollutant Discharge Elimination System (NPDES)
permit program, contractor reports, and other surveys indicating that each
of these industries discharges substantial amounts of at least some of the
pollutants on this list.
By gathering data and developing the regulations on an industry-by-in-
dustry basis, EPA expects to provide each industry with a complete package
of regulations covering both existing plants and new sources, with discharges
either directly to the Nation's waters or to publicly-owned treatment sys-
tems. While most of the regulations will be primarily technology based,
consideration will also be given to toxicity data and potential human health
effects. Where the technology-based regulations are inadequate to protect
against adverse effects to human health, drinking water supplies, or the
ecosystem, more stringent limitations will be set. Finally, in this con-
nection the regulatory scheme, like the statute itself, provides for con-
sideration of State water quality standards.
It will be the function of this presentation to describe the process by
which the Agency selected the 65 problem pollutants for which it is com-
mencing the regulatory process under the Act, as well as the 21 major in-
dustries for which these regulations are being developed.
2. STATUTORY FRAMEWORK
87
-------
The starting point for any discussion of EPA's regulatory programs is
the statutory framework. EPA is required under the Federal Water Pollution
Control Act to regulate the discharge of pollutants into the Nation's waters.
The goals of this Act include the elimination of the "discharge of toxic pol-
lutants in toxic amounts", and, by 1985, the elimination of all discharge of
pollutants iato the navigable waters.
There are numerous sections of this complex 89 page statute which are
designed in various ways to achieve these goals. Two of the most important
sections are 301 and 304, under which the Agency is directed to develop
technology-based effluent limitations guidelines by which all industrial
point source dischargers are to achieve limitations by July 1, 1977, which
reflect the best practicable control technology currently available or "BPT".
Similar regulations must require achievement by July 1, 1983, of effluent
limitations based upon best available technology economically achievable, or
"BAT". For new sources, the Agency is required under Section 306 to develop
standards of performance based upon best available demonstrated control
technology.
These regulations are in turn applied to specific plants through permits
issued under the NPDES permit program established under Section 402 of the
Act. The permits are issued by the EPA Regional Administrator, or the State
Director of an approved State program, of which there are at this time 27.
For "indirect dischargers", who discharge to publicly owned treatment
works, the Agency is to develop pretreatment standards under Sections 307(b)
and (c) for existing and new sources, respectively.
Finally, for a limited number of highly toxic and persistent compounds,
the Agency may set stringent toxic pollutant effluent standards under Sec-
tion 307(a) with no express requirement that technology or costs be con-
sidered. These must be met by the affected industry within one year of
promulgation.
Under these sections the Agency has established effluent limitations
guidelines and new source performance standards for 43 major industries,
covering some 418 industrial subcategories. For many of these industries
we have also issued pretreatment standards for indirect dischargers.
During the first few years of our implementation of the Act, the Agency
concentrated primarily on regulating the traditional sanitary parameters:
biochemical oxygen demand, suspended solids, and pH. Although certain pol-
lutants of demonstrated toxicity, including certain organic compounds and
heavy metals, were also limited, these received secondary emphasis. More
recently, however, we have all become increasingly aware of the need to
shift our primary emphasis to the regulation of specific toxic pollutants.
This concern has been sparked not only by widespread reports of environ-
mental damage from toxic pollutants, but also by their presence in drinking
water and fish, often at disturbingly high levels, traceable in many cases
to specific industrial discharges.
88
-------
The Agency has also made use, albeit limited, of Section 307(a). Be-
cause that section embodies a pollutant-by-pollutant approach rather'than an
industry-by-industry approach, widespread use of this section would confront
corporate decision-makers with a moving target. It can be disruptive and
costly to industry to invest substantial sums in treatment technology for
one toxic pollutant, only to be told a few months later that another toxic
pollutant is to be regulated, forcing the installation of other technology
which might or might not be consistent with what was installed for the first
pollutant. The short one-year compliance time also imposes limitations on
the use of Section 307(a), given the length of time which it takes to install
sophisticated control technology.
Not surprisingly therefore, the Agency has promulgated toxic pollutant
effluent standards under Section 307(a) for only six pollutants: Aldrin/
Dieldrin, DDT, Endrin, Toxaphene, Benzidine, and Polychlorinated Biphenyls
(PCB's). At the present time, it is expected that 307(a) will be held in
reserve for those situations where the toxicity or public health threat is
such that prompt and stringent control is required regardless of the avail-
ability of control technology.
In light of this experience, the Agency concluded in October, 1975,
that if broad and effective regulation of problem pollutants is to be
achieved, it must be accomplished on an industry-by-industry basis. This
meant that the Agency would expand its use of Sections 301, 304, and 306,
as well as the pretreatment authorities in Section 307 (b) and (c), in con-
trolling these pollutants.
Incidentally, this approach of using technology-based regulations to
control the discharge of toxic pollutants by 1983 recently received strong
support from President Carter. In his environmental message to Congress on
May 23, he said:
I have instructed the Environmental Protection
Agency to give its highest priority to develop-
ing 1983-best-available-technology industrial
effluent standards which will control toxic
pollutants under the Federal Water Pollution
Control Act, and to incorporate these standards
into discharge permits.
3. DEVELOPMENT OF PRIORITIZED LIST OF POLLUTANTS
As the first step in developing a strategy for regulating these "toxic"
or "problem" pollutants, the Agency decided in the Fall of 1975, to assemble
a work group to develop a prioritized list of toxic pollutants which should
receive primary regulatory attention under the Water Act. The members of
the group included staff scientists from virtually all of the Agency's
offices and divisions having responsibilities or interests in this area,
with the Office of Water Planning and Standards taking the lead. Among the
members were public health specialists with expertise in areas of toxicology
including carcinogenicity, chemists, biochemists, biologists, engineers, and
a statistician. Two outside scientific consultants were engaged to assist
89
-------
in the project, and additional review and consultation was provided from such
sources as the National Cancer Institute and the Oak Ridge National Labora-
tory.
The objectives of the group were, first, to develop criteria for the
selection and prioritization of toxic pollutants, and second, to relate those
to the discharges of major industrial sources, and thereby derive a list of
pollutants appropriate for regulation under the Act.
The group began by an exhaustive literature search, including lists of
pollutants prepared by other organizations with expertise in this area. We
focused our attention on both the hazard of the pollutant and the degree of
actual or likely exposure to humans or wildlife as a result of industrial
discharges.
The basic criteria which this task force used to evaluate the potential
hazard of a toxic pollutant were: (1) evidence that the compound, or its
degradation products and metabolites, pose an actual or potential health
hazard, based on laboratory or human evidence that the substance produces
carcinogenic, mutagenic, or teratogenic effects, adverse effects on repro-
duction or behavior, or adverse effects on any organ system, (2) persistence,
(3) ability to bioaccumulate in organisms, and (4) evidence of synergistic
propensities for these toxic effects. In addition, the task force took into
"consideration all available evidence that the toxic pollutants might cause
lethal or sublethal effects on wildlife, particularly aquatic organisms and
those species dependant upon aquatic organisms for their food supply.
With respect to exposure, we considered the following factors, wherever
such information was available: (1) total production of the substance, (2)
use patterns, (3) estimated extent, both qualitative and quantitive, of
actual or likely point source water discharges, (4) the consequent exposure
of man and wildlife to the substance or its breakdown products and meta-
bolites, and (5) analytical methodology capabilities.
Among the more significant reports which we reviewed were the follow-
ing:
"Water Quality Criteria, A Report of the National
Technical Advisory Committee to the Secretary of
the Interior", U.S. Government Printing Office,
Washington, D. C. (1968). ("Green Book")
"Water Quality Criteria, 1972". National Academy
of Sciences, and National Academy of Engineering,
U.S. Government Printing Office, Washington, B.C.
(1974). ("Blue Book")
FWPCA Section 311 Supplement to Development Docu-
ment on Hazardous Substances (EPA November 1975).
EPA 440/9-75-009.
90
-------
"A Report Assessment of Health Risk From
Organics in Drinking Water by an Ad Hoc
Study Group to the Hazardous Materials
Advisory Committee," Science Advisory
Board, EPA, April 30, 1975. ("Drinking
Water Report")
"Identification of Organic Compounds in
Effluents From Industrial Sources" Versar,
Inc., EPA-560/3-75-002, April 1975.
"Final Report of National Science Founda-
tion Workshop Panel to Select Organic Com-
pounds Hazardous to the Environment."
(Developed in conjunction with Stanford
Research Institute) Dr. Norton Nelson et al.,
NSF, October 1975.
"Preliminary Assessment of Suspected Car-
cinogens in Drinking Water", EPA Office of
Toxic Substances, June 1975.
"Organics in Drinking Water: Listing of
Identified Chemicals Part I". Junk, et al.,
Ames Lab, ERDA, Iowa State University,
July 1975. ("Iowa Report")
"Registry of Toxic Effects of Chemical Sub-
stances" National Institute for Occupational
Safety and Health (NIOSH) (1975 ed.) and
"Suspected Carcinogens" A Subfile of the NIOSH
Toxic Substance List" (NIOSH, 1975 Ed.).
Oak Ridge National Laboratory Environmental
Mutagen Information Center Study dated Nov.
7, 1975.
"International Agency for Research on Cancer
Monographs on the Evaluation of Carcinogenic
Risk of Chemicals to Man", World Health Or-
ganization (1972-1973).
"Survey of Compounds Which Have Been Tested
for Carcinogenic Activity", P.H.S. Publica-
tion No. 149, USHEW.
After reviewing the work of others in this area, several things became
apparent. First, while these studies provided much helpful information,
none provided a direct answer to the difficult question of how to select and
prioritize toxic pollutants solely from an industrial point source effluent
discharge perspective. Second, our initial selection criteria mentioned
above, would almost certainly undergo an evolution, or refinement, as we
91
-------
began to apply them to raw data and evaluate the results.
The work group began its selection process by examining as large a num-
ber of compounds with potentially hazardous effects in water as would be
practical and feasible. The National Science Foundation, for example,
initially looked at 337 organic compounds, later narrowing this list to 80
deemed to have the highest interest for further research. This was somewhat
more helpful than the NIOSH toxic substance list, which indicates that the
number of unique substances for which toxic effects information may be avail-
able is approximately 100,000.
As a starting point the work group assembled a list of some 232 com-
pounds believed to be actually or potentially hazardous, and present in
water.
After having gathered as much information as possible with respect to
the initial working list of compounds, the work group determined that an
initial screening procedure was necessary in order to reduce the list to
those compounds requiring highest priority for in-depth review. The ini-
tial screening process employed the following selection criteria:
(1) Evidence of actual presence in effluent;
(2) Evidence of carcinogenic, mutagenic, or
teratogenic effects in laboratory test
systems, or human epidemiological studies;
or evidence of a high degree of toxicity
to aquatic organisms or systems.
Application of these criteria reduced the initial list of 232 to approxi-
mately 75. This screening procedure produced a workable group of compounds
for which there was both some evidence of point source discharge and for
which there was evidence of serious potential hazard to man or the environ-
ment.
This screened list of compounds and underlying data, as well as the
criteria, were then subjected to further analysis and quality control
evaluation. This led to a further refinement of both the list and the
criteria, with some compounds dropped and others added.
In applying the criteria to our initial list of compounds little ef-
fort was actually made to incorporate human epidemiological information
dealing with possible acute toxic effects. This was due primarily to the
fact that the main purpose of the regulations for which the list was being
developed was to protect against serious adverse effects which might be
posed by long term, or chronic exposure to these pollutants.
With regard to exposure, in order to determine whether or not a
particular substance is present in industrial effluent discharges, the work
group investigated all data sources available to the Agency. Only those
compounds for which there was information indicating actual presence in
point source effluent, or strong likelihood that the compound is present,
92
-------
were considered candidates for priority consideration. In determining which
compounds are present in point source discharges, the work group undertook
the following steps.
First, EPA's Effluent Guidelines Division compiled a "direct discharge
emissions inventory", which in turn was followed by a "discharge parameter
matrix" in which the pollutants known or strongly believed to be discharged
were listed by industry, coupled with such information as was available on
quantification. The principal sources of the raw data were permits and per-
mit applications, monitoring data, and information gathered by EPA contrac-
tors in the course of their developing the required data base to support
existing effluent limitations guidelines.
Second, we prepared an "effluent load reduction study", based on a sam-
ple of 1800 national pollutant discharge elimination system permits, which
list the maximum allowable discharge for various pollutants. An estimate
was made of the quantity of each pollutant discharged per day for each of
the regulated 43 industrial point source categories. By summing these
estimates for each pollutant, a rough estimate of the total discharge for
point sources could then be made for many of the pollutants. Even when sup-
plemented by monitoring reports, there are statistical and other limitations
on this type of data, and therefore it was used with caution.
In addition, the work group reviewed the several studies mentioned
earlier on the presence of many of the selected compounds in surface water
as well as drinking and well water.
With respect to food particularly fish and fish products) and humans,
additional monitoring sources were examined. These included specifically
the FDA Market Basket Survey and the EPA Human Monitoring Survey, which were
notably helpful regarding pesticides and heavy metals.
The prime source for data on annual production was the Stanford Research
Economic Handbook, which apparently derives a good deal of its information
from Federal Trade Commission data as well as from industry surveys. Rela-
tively little data were obtained on use patterns. For the purposes of this
selection process, however, it was less important to know the ultimate use
than it was to know the source and amount of the industrial discharge of
these compounds.
With respect to the final "exposure" criterion, "analytical methodology
capabilities", the work group ascertained that there were and are currently
available analytical methods capable of detecting all of the 65 high priority
compounds and/or generic categories on our final lists, although the degree
of confidence in the analytical methods used to detect these pollutants
varies somewhat especially at low concentrations.
The group completed its work in February, 1976. Its final work product
was four lists, consisting of 29, 18, 18, and 11 substances respectively.
The 29 pollutants on the first list which has the highest priority, meet the
following criteria:
93
-------
A. The pollutants are known to occur in point source effluents, in aquatic
environments, in fish, or in drinking water.
B. There is substantial evidence of carcinogenicity, mutagenicity and/or
teratogenicity in human epidemiological studies or in animal bioassay
systems.
C. It is likely that point source effluents contribute substantially to
human hazard, at least locally. This latter judgment is based on con-
sideration of the quantities emitted, the persistence of the compounds
in aquatic systems, their tendency to be stored in organisms used for
human food, and available information on effective doses in animal tests.
This first list contained the following substances:
LIST 1
1. Acenaphthene
2. Aldrin/Dieldrin
3. Arsenic compounds
4. Asbestos
5. Benzene
6. Benzidine
7. Beryllium compounds
8. Cadmium compounds
9- Carbon tetrachloride
10. Chlordane (technical mixture and metabolites)
11. Chloroalkyl ethers (chloromethyl, chloroethyl, and mixed ethers)
12. Chloroform
13. Chromium compounds
14. DDT and metabolites
15. Dichlorobenzenes (1, 2-, 1, 3-, and 1, 4-dichlorobenzenes)
16. Dichlorobenzidine
17. Diphenylhydrazine
18. Heptachlor and metabolites
19. Hexachlorocyclohexane (all isomers)
20. Lead compounds
21. Mercury compounds
22. Nickel compounds
23. Nitrosamines
24. Polychlorinated biphenyls (PCBs)
25. Polynuclear aromatic hydrocarbons (including benzanthracenes,
benzopyrenes, benzofluoranthene,
chrysenes, dibenzanthracenes,
and indenopyrenes)
26. 2,3,7,8 - Tetrachlorodibenzo-p-dioxin (TCDD)
27. Thallium compounds
28. Trichloroethylene
29. Vinyl chloride
94
-------
The 18 pollutants on the second highest priority list satisfy criterion
A, above, i.e., occurrence in point source effluents, in aquatic environ-
ments, in fish, or drinking water. However, in some cases the evidence of
carcinogenicity, mutagenicity, and teratogenicity is based primarily upon
structual similarity to compounds in List 1, or upon mutagenic activity in
bacterial screening systems; or testing has shown some evidence of carcino-
genicity, mutagenicity, or teratogenicity, but these results presently ap-
pear to be incomplete or equivocal. In addition, a few compounds are in-
cluded in List II on the basis of serious toxic effects other than carcino-
genicity, mutagenicity, or teratogenicity. These other serious toxic effects
are, for the most part, on aquatic organisms. Finally, the possibility of
significant human exposure attributable to point source effluents was judged
to be somewhat less than that for compounds in List I. (This judgment was
based on relatively small volume of discharges, or relatively low propen-
sity to persist in water or to accumulate in organisms.)
The substances on List II are as follows:
LIST II
1. Chlorinated benzenes (other than dichlorobenzenes)
2. Chlorinated ethanes (including 1,2-dichloroethane, 1,1,1-tri-
chloroethane, and hexachloroethane)
3. 2-chlorophenol
4. Dichloroethylenes (1,1-and 1,2-dichloroethylene)
5. 2,4-dichlorophenol
6. 2,4-dimethylphenol
7. Dichloropropane and dichloropropene
8. Endosulfan and metabolites
9. Endrin and metabolites
10. Fluoranthene
11. Haloethers (other than those in List I; includes chlorophenyl
ethers, bromophenylphenyl ether, bis(dischloroiso-
prophyl) ether, bis-(chloroethoxy) methane and poly-
chlorinated diphenyl ethers)
12. Halomethanes (other than those in List I; includes methylene
chloride, methylbromide, bromoform, dichlorobro-
momethane, trichlorofluoromethane, dichlorodifluoro-
methane, methylchloride)
13. Hexachlorobutadiene
14. Naphthalene
15. Pentachlorophenol
16. Phthalate esters
17. Tetrachloroethylene
18. Toxaphene
With respect to the 18 compounds on priority List III, all satisfy
criterion A: occurrence in aquatic environments, in fish, or drinking water.
In addition, all are known to have toxic effects on human or aquatic organ-
isms at relatively low concentrations. However, there is no substantial
evidence that these compounds have primary carcinogenic, mutagenic, or tera-
togenic effects. List III contains the following compounds:
95
-------
LIST III
1. Acrolein
2. Acrylonitrile
3. Antimony compounds
4. Chlorinated naphthalene
5. Chlorinated phenols (other than those in List II; including
trichlorophenols and chlorinated cresols)
6. Copper compounds
7. Cyanides
8. Dinitrotoluene
9. Ethylbenzene
10. Hexachlorocyclopentadiene
11. Isophorone
12. Nitrobenzene
13. Nitrophenols (including 2,4-dinitrophenol, dinitrocresol)
14. Phenol
15. Selenium compounds
16. Silver compounds
17. Toluene
18. Zinc compounds
The 65 compounds on these 3 lists are receiving the highest priority in
the Agency's regulatory programs currently being implemented under the Fed-
eral Water Pollution Control Act. They are listed together as Appendix A to
the consent decree issued in June, 1976, by the U. S. District Court for the
District of Columbia which was mentioned earlier.
In addition to these three lists, the work group developed a fourth
list of compounds which is receiving secondary regulatory attention by the
Agency. Compounds on this list are also known to occur in effluents or
drinking water. On the basis of existing evidence, they are judged to pre-
sent a less substantial direct hazard than the chemicals on Lists I-III.
However, there is reason to believe that they may be converted in the en-
vironment into derivatives or breakdown products which may be regarded as
precursors of hazardous compounds rather than major hazards in themselves.
List IV originally consisted of 11 items, and chlorine has since been added.
Under the consent decree, EPA is to gather data on these 12 pollutants. But
is not obligated to issue regulations for them. This list is as follows:
96
-------
4. SELECTION OF 21 INDUSTRIES
For the 43 industrial categories for which EPA has already issued efflu-
ent limitations guidelines in connection with the first round of permits
directed towards the July 1, 1977, date, considerable data were obtained on
which industries were discharging which pollutants, and in what volume. The
principal sources of this data include the NPDES permit applications, moni-
toring reports, "Development Documents" and other surveys and analyses pre-
pared by EPA's contractors in connection with the development of the first
group of regulations, sampling and analysis data, and other reports. The
staff report of the National Commission on Water Quality was also considered.
The 21 industries which were selected for the primary regulatory effort
have all been found to be discharging a substantial number of pollutants on
the list of 65, and in most instances the volumes of such discharges are
also substantial.
There was nothing magic in the number 21. The Agency did prepare a
rough prioritized ranking of the 43 industries, in the order of discharges
of number and volume of pollutants. The selection of the top 21 of these
industries reflects nothing more sophisticated than hard negotiations be-
tween the Agency and the plaintiff environmental groups in the four lawsuits
mentioned above. The environmental groups sought a high number to assure
maximum coverage. The Agency sought a lower number to preserve maximum
flexibility. In fact there is no question but that regulations ought to be
developed for these 21 industries. Moreover, to the extent that resources
permit, the Agency may well go beyond the 21 industries and develop compar-
able regulations for other industries as well.
The order in which the industries appear on the Agency's present work-
ing list is not the same as the original priority order. The present order
reflects to a large degree the availability of contractors and EPA project
officers, and similar administrative considerations. This list of indus-
tries is set forth below.
LIST IV
1. Acetone
2. n-alkanes (Cin - C
3. Biphenyl
4. Dialkyl ethers
5. Dibenzofuran
6. Diphenyl ether
7. Methylethyl ketone
8. Nitrites
9. Secondary amines
10. Styrene
11. Terpenes
12. Chlorine
97
-------
INDUSTRIES FOR WHICH REGULATIONS
WILL BE DEVELOPED UNDER THE
FEDERAL WATER POLLUTION CONTROL
ACT TO CONTROL PROBLEM POLLUTANTS
1. Timber Products Processing
2. Steam Electric Power Plants
3. Leather Tanning & Finishing
4. Iron & Steel Manufacturing
5. Petroleum Refining
6. Nonferrous Metals Manufacturing
7. Paving & Roofing Materials
8. Paint & Ink Formulation & Printing
9. Ore Mining & Dressing
10. Coal Mining
11. Organic Chemicals Manufacturing
12. Inorganic Chemicals Manufacturing
13. Textile Mills
14. Plastic & Synthetic Material Manufacturing
15. Pulp & Paperboard Mills
16. Rubber Processing
17. Soap & Detergent Manufacturing
18. Auto & Other Laundries
19. Miscellaneous Chemicals
20. Machinery & Mechanical Products
21. Electroplating
5. SUBSEQUENT REFINEMENT OF LISTS
Each of these lists is subject to refinement, and the consent decree pro-
vides the necessary flexibility to do so. Industries or industrial subcate-
gories may be added or deleted. Scientific knowledge never stands still.
It may be that as the program is implemented, some substances will be added
and others deleted. While great efforts and talents were brought to bear in
selecting the original 65, no one suggests that this list is perfect.
It should be noted that some of the items on this list are actually
families of compounds. As of this date, in implementing the program, EPA
has broken out the families into individual substances, and has further
prioritized within these families. This has resulted in a refined working
list of 129 substances.
The process by which this further refinement has been-accomplished will
be discussed next by Walter M. Shackelford of EPA's Athens, Georgia, Labora-
tory.
This concludes my presentation. I will be glad to answer any questions
you may have.
98
-------
REFERENCES
The author wishes to express appreciation to the Chairman of this task
force, John C. Kolojeski, who was formerly with EPA's Office of General
Counsel and Office of Water and Hazardous Materials, and is now President
of Clement Associates, Inc., scientific regulatory consultants in Washington,
D.C., for his assistance in documenting various activities of the task force
as discussed in this paper.
BIOGRAPHY
Ridgway M. Hall, Jr. is Associate
General Counsel, Water Quality Divi-
sion, of the Environmental Protection
Agency. He is responsible for ad-
vising and representing the Agency on
all legal matters relating to the
quality of the nation's waters, which
includes primarily issues arising un-
der the Federal Water Pollution Con-
trol Act and the Safe Drinking Water
Act. He received his B.A. degree from
Yale University in 1963 and his L.L.B.
degree in 1966 from Harvard Law School.
He is a former partner in the law firm
of Cummings and Lockwood of Stamford,
Connecticut, where he practiced law
from 1966-1975. Mr. Hall is a member
of the American Bar Association Sec-
tions on Litigation and Natural Re-
sources Law.
99
-------
DISCUSSION
Richard B. Costa: (CEC Thompson Engineering, Inc. Houston, TX) How many
substances on the "list of priority pollutants" are common to both industrial
and municipal effluents?
Ridgway Hall: We do have some information in that area but not a lot. We
know that a lot of them are coming out of POTW's. The major effort in
developing these lists was on direct discharges. We then sought to deter-
mine how much of what goes into the public owned treatment works in turn
passes through. We have only the roughest kind of data in that area. We
know that a number of them do pass through some plants to some degree, but
there is still more data gathering to do in that area, and frankly the dis-
charges from the POTW's were of secondary priority to the work group. We
will be gathering data in that area as we go about implementing this settle-
ment decree in connection with the development of pretreatment standards.
The principal thrust of the pretreatment standards is to require whatever
steps are necessary to prevent either upset of the POTW, or pass-through
by harmful pollutants in harmful amounts. Where we find these problems, it
will indicate to us that those pollutants should be controlled by appro-
priate pretreatment standards.
Mr. Costa; When would you expect that to be available?
Mr. Hall: It will vary from industry to industry and follow to some extent
the schedule which is outlined in the consent decree for the contracts. As
you probably know there are five series of dates for the issuance of con-
tracts, the development of proposed regulations, and promulgation. The
earlier industries on the administrative timetable will probably see that
data start falling out sooner. Most of the contracts have already been let,
and people are in the process of gathering data. I would suspect that some
of that data will be available in a good deal less than a year and perhaps
for some of those early industries as soon as 7 or 8 months from now. The
thing to do is contact the project officers on those if you're interested
in that kind of data.
A. Karim Ahmed; Will you please explain the list of twelve compounds which
you showed on your list No. 4 on the screen? I'm a little unclear about its
relation to the implementation of the Consent Decree.
Ridgway Hall; That list survives now as Appendix C to the Settlement Agree-
ment and it's governed specifically under paragraph 4(b) of the Settlement
Agreement for those of you who follow those sort of things with a fine
tooth comb. Under the Decree, as the Agency goes about its industry-by-
industry data-gathering and sampling - analysis efforts, wherever we find
any of those 12 compounds present in effluents we will gather data. We
will record it so that we will know who is putting out which of the com-
pounds on that list and in what amounts. It will therefore give the Agency
a good data base so if we later find out that the compounds are indeed
causing problems either through synergistic reactions, or breakdown or
metabolite reactions, we'll have the data in hand to go out and promptly
develop regulations to deal with those problems rather than having to come
100
-------
back a second time and bother these 21 Industries again with another massive
data gathering effort.
Peter Foley; The President's message also had a recommendation for changes
to Section 307. Could you give us some insight what problems the Agency
has with that Section and what amendments they will be seeking?
Ridgway Hall; Yes, I'd be glad to. The principal problem that we have with
307 is the fact that it basically is a pollutant-by-pollutant approach rather
than an industry-industry approach. Another problem is that both the stan-
dard-setting time frame and the compliance time are very short. But the
biggest limitation on its usefulness is that it proceeds on a pollutant-by-
pollutant basis. It is inherently disruptive when you are trying to develop
a program which will give major coverage to industries for a lot of pol-
lutants. You're trying to regulate and get controls in place for a large
number of toxic pollutants. As those of you who are corporate managers and
decision makers know, if the Agency can come to you and say, based on data
which together we jointly assembled, "The following controls appear sensi-
ble and necessary to control everything in your effluent that looks as though
it's going to be harmful", you can then go and plan on that. You can buy
the technology, you can install it, and you've got a reasonable installation
time. You don't have the moving target problem of EPA coming down the road
3 or 6 months later with "Oh, by the way, we just found mercury in your
effluent, you've got to put the following technology in for that."
I think the other troublesome features with it are as follows. First,
the Agency was told to develop a list of pollutants for regulation under
the section within 90 days after passage of the Act. It was then told to
develop toxic effluent standards within 180 days after that, and then run
formal rulemaking hearings, and within 6 months after proposing the stan-
dards to promulgate them. Then industry has only 1 year to comply with
those standards. That's a very rigid timetable for the regulatory program
in that we have to list the pollutants practically before we go out and
gather the data to find out if the pollutants are really troublesome or not.
Furthermore, one year is an awfully short time for industry to install
meaningful control technology=
I think therefore that the kinds of amendments which you might see are,
for one thing, language that would alter the regulatory trigger mechanism
by disconnecting the section from 90 days following passage of the Act,
which is long since past. While I cannot speak for the Administration at
this point, in my personal view the time for listing a pollutant under
Section 307 (a) should be when information comes to the attention of the
Administrator on the basis of which he deems it appropriate to regulate it.
This way you gather the data first and then make the regulatory decision.
There is also discussion about switching the kind of rulemaking from the
formal adversary trial-type proceedings to the more informal rulemaking
such as you see under Section 307(b) for pretreatment standards. This would
be notice and comment rulemaking with the option of a hearing, if you want
it. This was the procedure essentially set out by Congress last year when
it passed the Toxic Substances Control Act.
101
-------
There was a bill which was passed by the House earlier this year which
would have allowed some flexibility to extend the one-year compliance time
up to three years if the Administrator determines that a further extension
beyond the one year is necessary because of technological reasons. That
bill died in conference, but I suspect it will come up again. The adminis-
tration might want to include a provision on that amendment that would direct
the Administrator to consider any adverse effects on human health or the en-
vironment in determining whether or not to extend the period beyond the one
year. One can envision the situation where technology might be installable
let's say 2 years down the road but the human health hazard might be so bad
you would want to shorten the time notwithstanding the temporary unavail-
ability of technology. Those are the kind of amendments that I would fore-
see.
102
-------
EVOLUTION OF THE PRIORITY POLLUTANT LIST
FROM THE CONSENT DECREE
W. M. Shackelford and L. H. Keith*
U. S. Environmental Protection Agency
Environmental Research Laboratory, Athens, Georgia
INTRODUCTION
The Consent Decree of 7 June 1976 (1) required the Environmental Protection
Agency to assess industrial wastewater with respect to 65 compounds and compound
classes. To provide chemical analysis data at a reasonable cost, however, the compound
classes listed in the Consent Decree had to be resolved into representative individual
compounds. The factors to be considered for such a resolution procedure include the
known frequency of occurrence of the compound in water, the number of manufacturing
sites and quantity of material manufactured, and the availability of a reference standard.
Once the compound classes have been resolved by EPA, generation of a protocol for
screening analysis of a variety of effluent types becomes a problem that may be
quantified in terms of required analysis procedures such as the number of extractions,
instrument runs, etc.
CONSENT DECREE COMPOUNDS
For chemical analysis, the compounds listed in the Consent Decree can be
divided into three broad groups. The largest group is composed of those compounds and
classes of organic chemicals that are amenable to analysis by gas chromatography-mass
spectrometry. Of the 50 listed in this group, 13 are classes of compounds containing as
few as two and as many as several hundred known members. The second group consists of
13 metals and their compounds. As with the first group, each of the 13 compound
classes in this group could be construed to contain hundreds of constituents. The
elements of this group may be analyzed individually by plasma emission spectroscopy,
spark source mass spectrometry, or some other method of multielement analysis. The
final grouping consists of cyanide and asbestos, two substances that require specialized
analysis procedures not applicable to groups of materials.
Obviously, the challenge of chemical analysis of a sample for literally hundreds
of components is staggering — especially when it is considered that these components
may be only trace substftuents (parts per billion) in the overall sample matrix. The
*Present Location: Organic Chemistry Dept., Radian Corporation, Austin, Texas 78760
103
-------
drain on resources in governmental laboratories as well as laboratories in the private
sector in terms of manpower and money would be overwhelming if analysis were attempted
of effluent samples for the compounds and compound classes of the Consent Decree when
construed to their fullest extent. For these reasons it was deemed necessary to resolve
the list of 65 Consent Decree compounds and classes into a list of individual compounds
that represents faithfully each multicomponent class in the original list of 65. The
specific compounds mentioned in the Consent Decree, however, were automatically
passed through to the resolved list without further question. In addition, because
technology for identification of individual metal compounds, particularly organometallic
compounds, is not well developed, the metals were to be analyzed without regard to
oxidation state. This segregation of compounds leaves the problem of defining criteria
to properly choose representatives of the remaining broad chemical classes.
CRITERIA USED FOR RESOLVING LIST
To resolve the remaining Consent Decree compounds into a list that presents an
analytical problem of finite limits requires that decisions be made that address the spirit
and intent of the Consent Decree. For instance, it seems plausible that in deciding
which members of a compound class adequately represent the whole class, those
compounds that have little chance of finding their way into industrial effluents should be
of low priority. Those compounds, however, that are manufactured in quantity, are
used in manufacturing processes, or are by-products of other processes would get high
priority as group representatives.
Another aspect that may be used to prioritize the compounds that represent a
group is the previous occurrence of a given compound in an analysis of water samples.
If a candidate for group representation has not been identified in water in some previous
study, whereas another candidate has been previously identified, the priority for group
representation should be assigned to the latter candidate. The fact that a compound has
been identified in water previously is an indication that it is apt to be present in
additional similar studies.
A final aspect that must be considered in prioritizing chemical group represen-
tatives is the availability of analytical reference standards. Identification and quantifi-
cation of chemical compounds depends upon the use of reliable standards. If a reliable
standard of a candidate compound is not obtainable, quantitative analysis is impossible.
DATA USED IN RESOLVING LIST
The data used to resolve the Consent Decree list of compound classes falls into
three sets that are consistent with the above rationales. First, an EPA study of organic
compounds identified in water was used to provide the data on the compounds previously
detected in surveys of water (2). Second, data concerning the manufacture of compounds
was gathered fronrv previously conducted surveys of the chemical industry (3,4). Finally,
catalogues from various chemical supply houses were consulted to determine whether
analytical standards were available (5-14).
104
-------
Data previously gathered through analyses of various water types throughout the
world provided the most useful basis for identifying the candidates for group represen-
tation on the basis of their known existence and known amenability to state-of-the-art
analysis techniques. Of the original 65 compounds and classes, 5 had never been
detected in water samples. Moreover, 2 of the 13 groups of organics had no represen-
tatives that had been previously found in water. In the remaining 11 groups, however,
priorities could be set according to the past occurrence of a given group member in
water. Table 1 lists the 13 groups and their known frequency of occurrence in water.
Of the individual members of a group that had been found in water previously,
those that represented less than 5% of the total occurrence were removed from con-
sideration in the interest of reducing analysis costs. Table 2 shows the 13 groups and the
number of compounds that had been found in water within each group. Also listed is the
number of compounds within each group that met the greater than 5% occurrence
criterion.
Manufacturing data help prioritize candidate compounds as representatives of
chemical groups by indicating the amount of a given chemical that the environment may
have to absorb. Of the 13 groups of organics that were listed in the Consent Decree,
only 9 contained compounds that were manufactured in quantity. Compounds in these 9
groups that had no known manufacturers were given lower priority than those for which
production data existed. Table 3 lists the 13 groups of compounds and the number of
compounds for which production data existed. The production data overlapped with the
frequency data in most cases except that two chlorinated naphthalenes and one nitro-
samine are produced but have never been found in water. N-nitrosodiphenyl amine, the
rubber chemical, and 2-chloronaphthalene were added to the list on the basis of
production. The other materials had already been included because they met the
frequency criterion.
The availability of analytical reference standards for the candidate compounds
was determined by searching chemical supplier catalogues. As expected, standards for
most of the compounds that are manufactured or have been found in water previously
were readily available from a number of chemical supply houses. Where possible, non-
commercial sources were found for compounds mentioned specifically in the Consent
Decree. Those compounds that were only candidates for group representation, however,
were dropped from consideration. Table 4 lists the 13 groups and the total number of
considered candidates in each group along with the number eliminated because of lack of
a standard.
GENERATION OF AN ANALYTICAL PROTOCOL
The reason for resolving the Consent Decree list was to make possible the analysis
of industrial effluents for these compounds and classes of compounds within the constraints
of available time and money. EPA's Environmental Monitoring and Support Laboratory
in Cincinnati was given the responsibility of developing an analysis protrocol for
screening the effluents for the compounds on the resolved Priority Pollutant List. The
105
-------
Athens Environmental Research Laboratory was asked to provide input for the screening
analysis protocol for those organic compounds not amenable to the purge and trap
(Bellar-Lichtenberg) technique. Once the Priority Pollutant List had been evolved,
establishment of the analysis protocol for screening became a problem of applying known
methods and techniques to the specific set of compounds on the Priority Pollutant List.
In this case, the development of an analysis protocol also involves the consider-
ation of two aspects that are many times not considered in survey analysis. First, since
the actual laboratory work is more than any one contractor could handle, provisions must
be made to insure the compatibility of data from one laboratory to the next. Second,
this protocol must be limited to looking for only the priority pollutants. The method
used (computerized GC-MS), however, would also see—were there time to interpret the
data—some of those compounds eliminated as class representatives and others that at
some later date may be of interest.
Even though it is recognized that each effluent will present its own set of unique
problems, the protocol provides a measure of compatibility by giving a general approach
to follow and specifying techniques within the general approach. For instance, gas
chromatographic column packing materials are specified and a common lot is to be used
by all laboratories involved in the screening analysis.
To save data that will allow the future determination of compounds not on the
Priority Pollutant List involves two factors. First, the analysis protrocol must be general
enough to allow the detection of compounds other than those on the list. This require-
ment is satisfied by using extraction, concentration and detection techniques that do not
discriminate against any compounds that can be gas chromatographed—essentially state-
of-the-art survey analysis with mass spectrometric detection. Second, a method must be
developed for long-term storage of the raw data. This requirement is satisfied by
coupling a computer system directly to the mass spectrometer so that raw data may be
transferred to magnetic tape.
CONCLUSIONS
In listing classes of compounds rather than specific compounds, the Consent
Decree presented a problem that to the analytical chemist seemed without practical
limit. By using available data concerning previous detection in water, manufactured
quantities, and availability of standards, however, the ambiguous compound classes can
be replaced with specific compounds of high priority. Once the specific compound list
has been resolved, however, the necessity for interlaboratory comparability of data and
effective ways to save all raw data becomes even more acute in the development of the
final protocol.
REFERENCES
1. Consent Decree, U.S. District Court for the District of Columbia, 7 June 1976.
2. Shackelford, W.M. and L.M. Keith. Frequency of Organic Compounds
106
-------
Identified in Water, U.S. Environmental Protection Agency, Athens, GA
Publication No. EPA-600/4-76-062. December 1976.
3. 1976 Directory of Chemical Producers, USA. Chemical Information Service,
Stanford Research Institute, Menlo Park, CA.
4. Organic Chemical Producers' Data Base Program, Volume II. Radian Corp.,
Austin, TX. Prepared for U.S. Environmental Protection Agency, Athens, GA,
under Contract No. 68-02-1319. August 1976. (Unpublished report).
5. 1977-1978 Catalog. Aldrich Chemical Co., Milwaukee, Wl.
6. Catalog 18. Analabs, Inc., North Haven, CT. June 1976.
7. Catalog 750. J.T. Baker Chemical Co., Phillipsberg, NJ. July 1975.
8. Bulletin CS-100-8. Chem-Service, West Chester, PA. 1975.
9. 1975 Catalog. Chemical Procurement Laboratories, College Point, NY.
10. Catalog 48. Eastman Kodak Co., Rochester, NY. 1976.
11. Catalog No. 10. K & K Fine Chemical, Plainsville, NY. 1975.
12. Nanogen Index. Nanogens International, Freedom, CA. 1975.
13. 1976 Catalog. Pfaltz and Bauer Chemical Co., Stamford, CT.
14. Chemical Standards for Air-Water-Foods. RFR Corp., Hope, Rl. 1975.
DISCUSSION
J. D. Hallett, Shell Oil Co.: I have two questions. I have heard the words "screening
analysis" used several times both by you and Mr. Hall. Can you define this term and
would you describe your efforts on the "lab-to-lab" comparability program?
W. M. Shackelford: Screening analysis as intended in the analysis protocol is the quali-
tative and semi-quantitative analysis of effluents for the 129 priority pollutants. This
differs from a survey analysis in that a survey would be concerned with all components
in a sample.
J. D. Hallett; Define semi-quantitative.
W. M. Shackelford: Semi-quantitative as we mean it here is quantification by a method
that has not been standardized—in this case mass spectrometry.
J. D. Hallett: Are you talking about orders of magnitude?
W. M. Shackelford: As indicated in the protocol, concentration values are to be
reported in ranges—10-100 parts per billion (ppb) and greater than 100 ppb. To answer
your second question, the program to establish some measure of lab to lab comparability
includes supplying contractors with the same lots of GC column packings, analytical
standards, and requiring that all data be saved on 9-track magnetic tape in a specified
format.
J. D. Hallett; Will the analytical standards be available to industrial labs?
W. M. Shackelford: As far as I know, they will be available commercially.
107
-------
Don Rosebrook, Radian Corp.; Radian CorporaHon is the EPA contractor for preparation
of the Consent Decree standards. When this order is filled, the standards will be made
available to the general public.
Leo Duffy, Standard Oil Co. of Indiana; Many analytical chemists regard the GC-MS
to be a totally qualitative instrument. Would you remark on how one establishes a limit
of detection with a qualitative GC-MS instrument? Doesn't this limit of detection
float and depend on instrument response?
W. M. Shackelford; The term limit of detection as used in the analysis protocol is mis-
leading. Actually, the numbers referred to as limits of detection are required concen-
trations that must be detected. The 40 ng figure that is given was derived from the
amount of material injected after the standard extraction and concentration procedure
was performed on a 10 ppb solution. For some compounds the figure is 200 ng. So these
are practical working levels—not actually limits of detection. I agree that the mass
spectrometer is not quantitative by present methods. However, with the use of our
internal standard, d]Q anthracene, and calculation of response factors for each compound,
mass spectrometry appears adequate to establish the ranges of concentrations we have
proscribed.
Leo Duffy: In the semi-volatile compounds you list three key fragment ions along with
the GC retention time of the compound. Would you remark on the certainty of identifi-
cation of specific compounds, restricting yourself only to those three key fragment ions?
W. M. Shackelford: If the GC retention time is considered, the confidence in identifi-
cation by three fragment ions is increased. A window of + 1 minute, which corresponds
to about 10 MS scans on each side of the peak, has proved to be adequate. Of course,
identification with absolute certainty can only occur if a standard and unknown are run
and the complete spectra compared. For the purposes of the protocol, however, the
present method is sufficient.
F. L. Robertaccio, E. I. DuPont de Nemours Inc.; How much of your experience has
been on applying these methods to complex wastewaters?
W. M. Shackelford: The analysis protocol essentially involved the use of techniques
used at the Athens laboratory for the past three to five years. The Analytical Chemistry
Branch has been concerned with characterization of effluents and essentially all of our
experience has been using methods similar to those in the protocol. Although some work
has been done on drinking water, much has been done with actual effluents. The
analysis protocol has also been run on effluents in our lab.
N. F. Seppi, Marathon Oil Co.; You mentioned chlorination of waters. Would you
elaborate on whether or not this affects the concentration of chlorinated compounds
found in water?
W. M. Shackelford: The only elaboration I can make is to quote papers and reports that
108
-------
indicate that after chlorination of water containing organic materials occurs, chlorinated
organics are found. Whether or not increases in the concentration of chlorinated com-
pounds present before chlorination occurred, I do not know. Studies have shown that
chlorinated compounds not found before chlorination do show up after chlorination.
BIOGRAPHIES
Walter M. Shackelford
Walter M. Shackelford is a Research Chemist
in the Analytical Chemistry Branch of EPA's Athens
Environmental Research Laboratory. He has a B.S.
degree in Chemistry from the University of Mississippi
and a Ph.D. degree in Analytical Chemistry from
Georgia Tech. Prior to his work with EPA, Dr.
Shackelford was a post-doctoral research associate
with G.G. Guilbault at the University of New
Orleans and spent two years in research for the U.S.
Army at Edgewood Arsenal, Md.
Lawrence H. Keith
Lawrence H. Keith received his Ph.D. from
the University of Georgia in 1966. He worked at
the Athens Environmental Research Laboratory in
Athens, Georgia for over 10 years under the U. S.
EPA and its predecessors. Dr. Keith is the editor
of the book, Identification and Analysis of Organic
Pollutants in Water, and author or co-author of 39
technical publications including 10 chapters in five
books. At Radian Corporation he is Head of the
Organic Chemistry Department and is responsible
for the direction of a technical staff of 24. He is
the Secretary of the ACS Division of Environmental
Chemistry, Chairman of the Division Steering
Committee for the North American Council on
Organic Pollutants (NACOP), and Vice-Chairman
of the next Gordon Research Conference on Environ-
mental Sciences: Water.
109
-------
TABLE 1 TOTAL FREQUENCY OF OCCURRENCE FOR 13 CHEMICAL GROUPS
Consent Decree Group Number of Observations*
Chlorinated Benzenes 247
Chlorinated Ethanes 138
Chloroalkyl Ethers 67
Chlorinated Naphthalenes 0
Chlorinated Phenols 92
Dinitrotoluenes 14
Haloethers (other than above) 15
Halomethanes 360
Nitrophenols 13
Nitrosamines 0
Phthalate Esters 183
Poly nuclear Aromatic Hydrocarbons 49
DDT Metabolites 23
*5700 Total Observations for Study
TABLE 2 FREQUENCY OF PRIOR OCCURRENCE
OF COMPOUNDS WITHIN 13 CHEMICAL GROUPS
Number of Number of Compounds
Consent Decree Group Compounds Meeting 5% Criterion
Chlorinated Benzenes 28 9
Chlorinated Ethanes 19 9
Chloroalkyl Ethers 15 4
Chlorinated Naphthalenes 0
Chlorinated Phenols 16 5
Dinitrotoluenes 4 4
Haloethers (other than above) 4 4
Halomethanes 17 9
Nitrophenols 4 4
Nitrosamines 0
Phthalate Esters 19 7
Polynuclear Aromatic Hydrocarbons 12 9
DDT Metabolites 2 2
110
-------
TABLE 3 MANUFACTURING DATA FOR 13 CHEMICAL GROUPS
Number of Compounds
Consent Decree Group With Production Data
Chlorinated Benzenes - 4
Chlorinated Ethanes 5
Chloroalkyl Ethers 0
Chlorinated Naphthalenes 2
Chlorinated Phenols 3
Dinitrotoluenes 2
Haloethers (other than above) 0
Halomethanes 4
Nitrophenols 4
Nitrosamines 1
Phthalate Esters 2
Pol/nuclear Aromatic Hydrocarbons 0
DDT Metabolites 0
TABLE 4 NUMBER OF COMPOUNDS ELIMINATED
DUE TO LACK OF ANALYTICAL STANDARDS
Number of
Considered Number of
Consent Decree Group Compounds Compounds Eliminated
Chlorinated Benzenes 9 2
Chlorinated Ethanes 9 2
Chloroalkyl Ethers 8 2
Chlorinated Naphthalenes 2 0
Chlorinated Phenols 6 0
Dinitrotoluenes 4 1
Halothers (other than above) 4 1
Halomethanes 10 0
Nitrophenols 6 0
Nitrosamines 3 0
Phthalate Esters 7 2
Polynuclear Aromatic Hydrocarbons 9 2
DDT Metabolites 2 0
111
-------
"INCORPORATION OF THE PRIORITY POLLUTANTS
INTO EFFLUENT GUIDELINES DIVISION DOCUMENTS"
Robert W. Dellinger
Chemical Engineer, Effluent Guidelines Division
U. S. Environmental Protection Agency
The purpose of this paper is to explain the role of the Effluent Guidelines
Division relative to the 1976 Settlement Agreement (1). It will also provide
some incite into the overall relationship between the Effluent Guidelines
Division and ether responsible offices regarding future regulations which
will be affected by the Settlement Agreement.
BACKGROUND
The Effluent Guidelines Division (EGD) is a division of the Office of Water
Planning and Standards within the Office of Water and Hazardous Materials.
The EGD was founded in August of 1972 with the primary function of
contributing to the establishment of effluent limitations and guidelines and
standards of performance for new sources pursuant to Sections 301, 304, and
306 of the Federal Water Pollution Control Act as amended in 1972. This Act
is commonly referred to as Public Law 92-500. The division was originally
headed by Allen Cywin and is now headed by Robert Schaffer. Over the past
several years, the EGD has been involved in the publication of some 1700
regulations covering over 250 subcategories within 43 industrial point source
categories.
As of June 7, 1976, the function of the EGD has taken a new slant. On June
7, 1976, the Settlement Agreement with the Natural Resources Defense Council,
et al., came into effect and settled several cases in the District Court for
the District of Columbia (1). This Settlement Agreement requires that EPA
give consideration to 65 specific chemicals or classes of chemicals when
establishing effluent limitations reflecting the best available technology
economically achievable, also referred to as BAT or 1983 limitations. This
requirement for consideration of the list of chemicals carries over to the
establishment of New Source Performance Standards and to the establishment of
pretreatment standards for new and for existing sources. The authority by
which EPA has been directed to perform this task derives from Sections 301,
304 (b), 306, 307 (b), and 307 (c) of Public Law 92-500 (2). The Settlement
Agreement also establishes a schedule within which contracts are to be let,
regulations proposed, and regulations promulgated for a prioritized list of
21 industrial categories.
PRELIMINARY EFFORTS
Several activities had to be completed before the role of the Effluent
Guidelines Division relative to the Settlement Agreement could be
definitized. These activities included further work with the list of 65
chemicals or classes of chemicals, development of sampling and analysis
procedures, and the letting of contracts in order that work could begin on a
timely schedule.
112
-------
In order to enable the Agency to conduct a proper scientific study of
industrial discharges, the various pollutant parameters were further.defined.
Without this information as a point of reference, the analytical portion of
the BAT review would be unmanageable given the limited time schedules and
resources of the Agency. This list of 65 chemicals or classes of chemicals
was defined in November of 1976 to include 123 specific unambiguous
compounds. This list of 123 specific compounds appeared in Portfolio B of
the Petroleum Refining Industry Survey distributed to the industry in
February of 1977. Since that time, the list has been revised. This occurred
in April of 1977 with the addition of 6 compounds; now the priority pollutant
list totals 129 compounds (3). The six additional compounds include 5
additional PCB's and di-n-octyl phthalate. We believe that this list
fulfills the requirements of the court approved agreement and can be
evaluated analytically.
In addition, the Agency has established procedures for sampling and analysis
of industrial effluents for the priority pollutants. These procedures are
defined in the "Sampling and Analysis Procedures for Screening of Industrial
Effluents for Priority Pollutants" published in March of 1977 and revised in
April of 1977 (4).
While these tasks were being completed by personnel within various program
areas of EPA, the Effluent Guidelines Division set about to fulfill its
requirements relative to the Settlement Agreement. Due to a relatively small
staff, the EGD has historically depended on outside contractors to aid in the
fulfillment of its tasks. EGD technical personnel serve as project officers.
Their function is to monitor, assist, direct, and provide various inputs to
the contractor. The project officer works directly with the contracting
firm, usually on a daily basis. One of his main duties is to provide the
contractor with information regarding EPA policy.
The Effluent Guidelines Division should meet the schedule for executing
contracts that was established in the Settlement Agreement. The EGD let ten
contracts prior to January 1, 1977 (These included the timber products, steam
and electric, leather tanning, iron and steel, petroleum, paint and ink, coal
mining, ore mining, nonferrous metals, and paving and roofing industries.).
In addition, contractors have been selected for ten of the remaining eleven
industrial categories and negotiations are currently going on between the
selected contracting firms and EPA procurement. The miscellaneous chemicals
request for proposal will be readvertized and will involve multiple awards
rather than a single award. Six of the eleven contracts are scheduled for
award by June 30, 1977. The additional five will be awarded by October 31,
1977. The Scope of Work to be fulfilled by each of the contracting firms is
defined in "Request for Proposal No. WA 77-B074" dated February 7, 1977 (5).
The contractor currently involved in the petroleum refining industry study is
Burns and Roe of Paramus, N.J, EPA personnel from the Robert S. Kerr
Environmental Research Laboratory (RSKERL) are participating in the sampling
and analysis phase of the study. Another contracting firm, Ryckman,
Edgerley, Tomlinson & Assoc., Inc. (RETA) of St. Louis, Missouri, is also
assisting in sampling and analysis.
113
-------
DUTIES OF THE EFFLUENT GUIDELINES DIVISION
The Effluent Guidelines Division with the help of contractors will perform
the function of gathering and analyzing data. Its tasks include the
determination of presence or absence of the priority pollutants within waste
water generated by industrial point source categories or subcategories, an
evaluation of the quantities of priority pollutants present in these waste
waters, a determination of the extent to which these priority pollutants
exist subsequent to treatment by various control technologies, and an
estimate of the cost of implementing these various control and treatment
technologies. This information will be used during the decision-making
processes established within the Agency in fulfilling the requirements of the
Settlement Agreement.
The first step involved in completion of these tasks involves the development
of a profile of the industrial point source category. This involves an
identification of the plants which comprise the point source category in
terms of size, location, and number. It also includes a determination of
such information as the numbers of direct versus indirect dischargers, an
identification of the treatment systems employed at each plant, both end-of-
pipe and in-plant, and an identification of products produced or unit
operations employed at plants in the industry. This information forms the
foundation on which other information is based and will be usually obtained
by submission of a questionnaire to all or a statistically representative
portion of the industry. Portfolio A of the Petroleum Refining Industry
Survey provides the Agency the means to profile the petroleum refining
industry.
An important part of the study is the determination of presence or absence of
the priority pollutants. This will be accomplished in part by literature
searches, but the major thrust of_this exercise involves the implementation
of a sampling program. Presence or absence will be determined by a
"screening study" (4). This screening study involves the gathering of raw
water, raw waste water, and final treated effluent samples for analyses for
the priority pollutants. Analyses involve gas chromatography/mass
spectrometry (GC/MS) techniques for volatile and semi-volatile organics as
well as analysis for metals and pesticides (4). Upon obtaining the results
of this screening survey, decisions will be made regarding the necessity to
further investigate certain of the priority pollutants. These decisions will
be based on the guidance provided in paragraph 8 of the Settlement Agreement
(1). Exclusion of parameters can be justified if the specific pollutant is
present soley as a result of its presence in intake waters, or if the
specific pollutant is either not present in the discharge or is present in
insignificant quantities and not likely to cause toxic effects.
Upon making the decision of which parameters to be concerned with relative to
a point source category or subcategory, the second phase of the sampling
program is implemented. This involves verification or quantification of the
amount of priority pollutants present. It also, ideally, involves a
determination of the unit operation or operations which result in the
discharge of the specific priority pollutants of concern. In the
114
-------
quantification phase the degree to which the priority pollutant is removed by
various existing treatment systems within the point source category will be
evaluated.
As more of these BAT studies get underway and data become available, the
potential exists for supplementing information relative to one point source
category with data from another point source category. This could involve a
transfer of technology from a point source category where a more advanced
technology is practiced to one in which a lesser technology is practiced. It
is very early in the program to predict to what degree, if any, direct
transfer of technology could or would be applicable or whether or not it is,
in fact, necessary or desirable.
The EGD will also develop cost data to correspond to the various technologies
thought to be potential BAT technologies. This may include in-plant control
measures, recycling techniques to reduce wastewaster discharges, source
control prior to biological treatment, or end-of-pipe technologies. Efforts
will be made to provide good capital and operating cost estimates for the
identified potential BAT technologies.
EGD Efforts tŁ Date Relative ^o the Petroleum Refining Industry
Regarding the petroleum refining industry specifically, this overall EGD work
plan has been altered slightly. Due to time constraints and delays, a
decision was made to combine the screening and quantification phases into a
single program. Initially, twelve plants were to be sampled. At six of the
plants, pilot activated carbon studies will be conducted. This sampling
study is now being conducted by RSKERL. A rationale was developed for
selection of twelve individual refineries to be sampled (6). This involved
establishing the following criteria for selection:
1. The selected plants must be BPT plants; i.e., they must be attaining BPT
limitations or fulfilling the requirements of their July 1, 1977, permit
requirements.
2. The selected plants would have a single outfall. This criteria was
established strictly due to economic considerations, namely the large cost of
analysis of individual samples.
3. The selected plants would cover as many of the unit operations found in
the petroleum refining industry as possible, and
4. The selected plants would represent various crude sources. Because no
completed questionnaires had been received at the time of selection, it was
agreed that this criteria would be met through a selection of plants with a
wide geographic distribution.
A secondary consideration was that in the selection of the six refineries
where pilot carbon studies would be performed, at least four would involve
activated sludge treatment. The other two would also involve some other form
of biological treatment (i.e., aerated lagOOM, oxidation ponds, etc.).
115
-------
The EPA contractor, Burns and Roe, went to work on developing the list of 12
refineries to be sampled. They visited EPA regional offices and State
offices. They gathered Discharge Monitoring Report data while speaking to
Regional and State personnel. Forty-five refineries were identified which
met the first two criteria (attainment of BPT limitations while having a
single outfall) (6). The Oil and Gas Journal was searched to determine the
unit operations employed at the various refineries. The fourth criteria of
geographic mix was added to the selection process.
A matrix was developed which resulted in the compilation of a list of 12
refineries to be sampled with 2 alternates corresponding to each selected
refinery. The current petroleum refining industry sampling study is based on
this list of 36 plants.
Various contacts were made with the American Petroleum Institute and the
National Petroleum Refiners' Assocation to inform them of the refineries
selected for sampling. API representatives contacted individual plants to
determine the mode of operation for the duration of the sampling study, March
to July. Based on this additional information, several first choices dropped
out due to various reasons (such as scheduled turnarounds and strikes).
Representatives of alternate refineries were contacted. A sampling schedule
was then developed around the availability of the selected refineries. The
decision was made that due to the time constraints imposed by the Court and
due to the tremendous workload on RSKERL personnel, this sampling schedule
would be strictly adhered to.
With the unavailability of certain refineries and a shift to alternates, the
final list of 12 refineries to be sampled omitted several unit operations.
Additional funding became available relative to the petroleum refining
industry study. The decision was made to supplement the initial sampling
program by sampling additional plants. Ryckman, Edgerley, Tomlinson, and
Associates, Inc. is conducting the supplemental sampling program which, at
the present time, includes 5 refineries (7). Table I gives a listing of, the
petroleum refineries to be sampled in the BAT review and the scheduled date
of initiation of the sampling program at each individual refinery.
THE RELATIONSHIP OF RESPONSIBLE PROGRAM AREAS
The total of the information gathered by the Effluent Guidelines Division
forms a part of the technical information on which final effluent limitations
will be established. Other studies are being conducted by other divisions
within the Office of Water-Planning and Standards. Major inputs will be made
by the Criteria and Standards Division headed by Dr. Kenneth Mackenthun, by
the Monitoring and Data Support Division headed by Dr. Edmund M. Notzen, and
by the Office of Analysis and Evaluation headed by Mr. Swep Davis.
The Criteria and Standards Division (CSD) is responsible for studies relating
to health and environmental effects, will provide physical and background
data relating to the priority pollutants, and will interface with other
government agencies to obtain health and environmental effects information.
116
-------
They will also develop specific criteria that will provide justification for
regulation of the priority pollutants.
The Monitoring and Data Support Division (MDSD) is responsible for general
studies of each priority pollutant including geographical and quantitative
profiles of users and manufacturers, and for a summary of losses of priority
pollutants to the environment through discharges to water by industrial
producers and users, municipalities, agricultural sources, and natural
sources. They will develop a data base associated with each industrial
category and gather information relating to ambient water quality to provide
an overview of the magnitude and geographical extent of potential problem
areas. This information will be used to provide an overall risk assessment
based on the quantity of priority pollutants released and their environmental
effects.
The Office of Analysis and Evaluation (OA&E) is responsible for economic
impact studies which will include an increased emphasis toward benefit
analysis. Economic studies will include a clear description and evaluation
of load reductions, of health and water quality risk avoidance, and of
correlations between industry location, water quality, and potential health
problems.
The inputs of all responsible program areas merge at the working group level.
Working groups involve personnel representing the responsible program areas
but include personnel representing such groups as Research and Development,
Toxic Substances, Pesticide Programs, Air Programs, General Counsel,
Enforcement, Planning and Evaluation, and EPA Regional Offices. Decision-
making will begin at the Working Group level with recommendations regarding
regulations passed along to EPA management for evaluation.
At the present time, working group activity relative to the petroleum
refining study has been limited to plant selection for the sampling program
and other inputs relating to the technical (EGD) and economic (OA & E)
studies. As more information is collected, the Working Group will take a
more active role in the decision-making process.
MAJOR MILESTONES
The following is a schedule of the major milestones which are important
relative to the petroleum refining BAT study:
EPA Draft Development Document Available for Public Comment - March 7, 1978.
Public Meeting to Discuss Draft Development Document - April 7, 1978.
Proposal of Regulations in the Federal Register - July 15, 1978.
Public Meeting to Discuss Proposed Regulations - September 22, 1978.
Promulgation of Regulations in the Federal Register - January, 22, 1979.
117
-------
SUMMARY
The Agency has a considerable amount of work to do relative to the Settlement
Agreement. The EGD has a major area of responsibility, that of gathering the
technical information on which standards will be established. The petroleum
refining study is in the early stages of development; we are awaiting the
results of GC/MS analysis for non-volatile organics and the results of metals
and pesticides analyses. Upon receipt of this data, the Agency will make the
decisions, at the working group level, which will determine the immediate
direction which the petroleum refining study will take.
REFERENCES
(1) Natural Resources Defense Council, et al. v.. Train, 8 E.R.C. 2120
(D.D.C. 1976).
(2) Public Law 92-500, 92nd Congress, S. 2770, October 18, 1972.
(3) "Rationale for the Development of BAT Priority Pollutant Parameters,"
Effluent Guidelines Division, Office of Water and Hazardous Materials,
U.S. Environmental Protection Agency, Washington, D.C., June, 1977.
(4) "Sampling and Analysis Procedures for Screening of Industrial Effluents
for Priority Pollutants," U.S. Environmental Protection Agency,
Environmental Monitoring and Support Laboratory, Cincinnati, Ohio, April,
1977.
(5) "Request for Proposal No. WA 77-B074," U.S. Environmental Protection
Agency, Washington, D.C., February 7, 1977.
(6) "Selection of Refineries for RSKERL Sampling Program," Burns and Roe
Industrial Services Corporation, Paramus, NJ, February 22, 1977.
(7) "Selection of Refineries for B&R Supplemental Sampling Program," Burns
and Roe Industrial Services Corporation, Paramus, NJ, May 4, 1977.
DISCUSSION
Paul Mikolaj. Lion Oil Co.; When will the results of EPA's preliminary
screening studies be available?
R^ jj^ Dellinger; ' As of next week, 14 of the 17 plants which have been
icheduled for sampling will have been visited. Results of the first 6
refineries screen-sampled by RSKERL are anticipated to be received by EGD
around mid-June. It was originally anticipated that this data would be
received by the end of May. Data relating to the other refineries sampled by
RSKERL and RETA will be available at a much later date. It is too early to
predict the actual date of availability at this time; we are pretty much at
the mercy of the analytical labs with regard to turnaround time.
Paul Mikolaj; Will this information be publicly available?
118
-------
R! Ht Dellinger; Eventually this information will be publicly available and
published in an EPA development document. Between now and that time I would
anticipate that there will be an exchange of data between *EPA and
representatives of the individual refineries where split samples were taken
In any event, the EPA data will be submitted to the refineries as soon as
data become available.
BIOGRAPHY
Robert W. Dellinger is a chemical
engineer with the Effluent Guidelines
Division of EPA. He has B.S. and M.S.
degrees in Chemical Engineering from
Virginia Polytechnic Institute and the
University of Maryland, respectively.
He has been with the Effluent Guide-
lines Division since November of 1972.
TABLE I
PETROLEUM REFINERIES SCHEDULED FOR SAMPLING AS OF JUNE 3, 1977
Group
Conducting
Carbon? Sampling (1)
Scheduled
Refinery
Gulf
Exxon
Clark
Hunt
Texaco
Mobil
Getty
Phillips
Shell
Conoco
Asamera
Exxon
Exxon
Quaker State
Sun
Coastal States
Arco
Location
Philadelphia, PA
Bay town, TX
Hartford, IL
Tuscaloosa, AL
Lockport, IL
Augusta, KS
El Dorado, KA
Sweeny, TX
Anacortes, WA
Ponca City, OK
Commerce City, CO
Billings, MT
Benicia, CA
Newell, WV
Toledo, OH
Corpus Christi, TX
Philadelphia, PA
Date
3/21
3/21
3/28
3/28
4/4
4/4
4/18
5/2
5/2
5/16
5/23
5/30
5/30
6/6
6/13
6/27
6/27
No
No
No
No
No
No
No
Yes
Yes
No
No
Yes
Yes
No
No
Yes
Yes
RSKERL
.RSKERL
RSKERL
RSKERL
RSKERL
RSKERL
RETA
RSKERL
RSKERL
RETA
RETA
RSKERL
RSKERL
RETA
RETA
RSKERL
RSKERL
(1) All sampling teams are assisted by Burns & Roe of Paramus,
NJ
119
-------
GENERATING DATA ON THE PRIORITY POLLUTANTS
FOR THE EFFLUENT GUIDELINES DIVISION
Leon H. Myers
Chief, Industrial Sources Section
Source Management Branch
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma
During the past decade, the Citizens of the United States have become
vitally interested and involved with the environmental conditions of the
Nation's water supplies. Our Congressional leaders recognized this concern,
and in October 1972, enacted legislation "to restore and maintain the chemical,
physical, and biological integrity of the Nation's waters."
Within the framework of the Federal Water Pollution Control Act Amend-
ments of 1972, the newly created Environmental Protection Agency was directed,
as National Policy, to prohibit the discharge of toxic pollutants in toxic
amounts (Section 101.(a) (3). Also, in Section 101 of PL 92-500, sub-section
(a) (6) declares that "it is the National Policy that a major research and
demonstration effort be made to develop technology necessary to eliminate the
discharge of pollutants into the navigable waters of the contiguous zone and
the ocean."
Section 307 (a) of the act requires EPA to publish a list of toxic pol-
lutants and to promulgate effluent standards for such toxic pollutants. The
following characteristics shall be taken into account; "the toxicity of the
pollutant, its persistance, degradability, the usual or potential presence of
the affected organisms, and the nature and extent of the effect of the toxic
pollutant on such organisms."
In September 1973, EPA promulgated a list of nine toxic pollutants and in
December 1973, proposed standards for these nine toxic pollutants.
On June 7, 1976, a settlement agreement was signed between Russell Train,
EPA Administrator, and four concerned citizen organizations to investigate
the presence of 65 parent chemical constituents reported to be present in the
Nation's water supplies. The 65 chemicals are listed in Table 1. This set-
tlement agreement also listed 21 point source categories by Standard Indus-
trial Classification (SIC)including 2911 Petroleum Refining.
After the agreement was signed, the EPA's Effluent Guidelines Division
was delegated the responsibility to satisfy the legal conditions of the
document. After preliminary meetings regarding the sampling, analytical, and
data storage protocols related to the study, the Robert S. Kerr Environmental
Research Laboratory (RSKERL) in Ada, Oklahoma, was contacted to provide tech-
nical assistance in sampling and analysis for the petroleum refining industry.
A meeting was held in Ada in November with representatives of EPA's
Effluent Guidelines Division, the Office of Research and Development, an
120
-------
American Petroleum Institute Task Force, and key RSKERL personnel. The meet-
ing provided an insight regarding potential sampling and analytical problems
associated with the general protocols presented by Effluent Guidelines Divis-
ion's Quality Control Branch.
From the results of this meeting, a preliminary plan was proposed and
accepted. Phase I was designed as a screening study to determine the presence
or absence of the priority pollutants, and Phase II was a screening carbon
study. Sampling and analytical procedures presented two major problem areas
which needed immediate attention; these problem areas were:
Sampling; The original protocol dated November 1976, required automatic
samplers be used to collect composite samples. Most automatic samplers use
petroleum-derived tubing which has the ability to adsorb or desorb organics
from the tubing walls into the water sample. Another problem in using auto-
matic samplers is the lack of electrical safety devices required when sampl-
ing in an explosive atmosphere area such as the API separator. A variance
was requested to allow the sampling be conducted by EPA personnel; this was
granted due to the aforementioned problems.
Analytical
The list of 65 chemical compounds, presented problems to the analysts be-
cause of the ambiguity and colossal analytical effort in separating, identify-
ing, and quantifying most of the compounds, metabolites, and isomers which
appeared on the list. A list of "Unambiguous Compounds" was prepared to
specifically name each pollutant which is to be identified.
This list was prepared by scientists from EPA's Athens, Georgia, labor-
atory and personnel of the Effluent Guidelines Division. The list of un-
ambiguous compounds was derived from a compendium of information gathered by
these scientists from an earlier study on organic compounds in potable water
supplies. The list of compounds shown in Table 2 is identical to the specific
compounds listed in the settlement agreement. The classes of compounds are
represented by carefully selected individual material. The EPA list was dis-
cussed with the plaintiffs of the settlement agreement, and there was no
dissension.
Another analytical problem encountered was the protocol prepared in
November by EPA scientists from the Cincinnati and Athens Laboratories. The
November protocol was a preliminary analytical exercise and was distributed
for review comments by industrial, society, and EPA scientists. In November,
a meeting was held at Atlanta, Georgia, with EPA participants from Effluent
Guidelines, Athens, and Ada. A request was made to prepare a specific pro-
tocol for non-volatile organics and volatile organics analytical procedure.
Athens scientists prepared the protocol for non-volatile organics, and
Cincinnati scientists prepared the protocol for the volatile organic compounds.
The procedures were merged into one protocol and presented in March 1977.
"These were the major problems encountered in making preparations for this
study, and they were corrected prior to the study date.
121
-------
In November, RSKERL requested a list of 12 refineries be selected
Effluent Guidelines for the field study. The selected refineries would be
representative of the five refinery categories, various crude oil sources,
geographic locations, and meet Best Practicable Control Technology Currently
Available (BPC1C), or EPA regional permit conditions. An alternate list of
six refineries meeting the aforementioned conditions was to be furnished in
the event that one or more of the primary selections become unavailable.
Effluent Guidelines would then contact the American Petroleum Institute
who would contact the refinery to determine if there were any planned turn-
arounds or other problems which could interfere with the purpose of the study.
After the selected refineries had been contacted by API and the results re-
layed to Effluent Guidelines, a final selection of refineries was prepared.
After confirming the selected refineries, the Effluent Guidelines' Project
Officer notified each refinery they would be visited by the RSKERL team for
the priority pollutant study. The 12 refineries to be visited by RSKERL
personnel for sampling are shown in Table 3.
This study was divided into two phases: Phase I is considered a screen-
ing study to determine the presence or absence of priority pollutants, and
Phase II is a screening study on the effectiveness of carbon treatment to
remove the priority pollutants. These two phases will be discussed separately.
Phase I
Six refineries were selected for this phase, and those refineries were:
1. Gulf, Philadelphia 4. Clark, Hartford
2. Exxon, Baytown 5. Texaco, Lockport
3. Hunt, Tuscaloosa 6. Mobil, Augusta
At each refinery, a minimum of three points were selected for sampling.
The intake water to the refinery was selected to provide information on the
background quality as related to the priority pollutants. The second sampl-
ing point represented the API separator or dissolved air flotation effluent,
which would indicate the quality of water that would be expected without
biological treatment. The third sampling point selected was representative
of the NPDES sampling point; this sampling point represented biologically
treated wastewater.
Sampling times were established over a three-day study period, and the
composite samples were prepared from aliquots obtained at three-hour inter-
vals. On the first sampling day, the composite sampling began at hour 1300
and was completed at hour 1000, the following morning. Twenty-four hour
composite sampling was started at 1200 on the second day and 1100 on the
third day and completed at 0900 and 0800, respectively. Grab samples were
obtained at the final sampling time for each day. Table 4 indicates the
parameter, type of samples, and preservative used for the samples. In
addition to the priority pollutant samples, Effluent Guidelines contracted
with Ryckman, Edgerly, Thomlinson, and Associates Laboratory to furnish con-
tainers and analysis for "classical parameters" for each sampling point.
Table 6 is a list of the "classical parameters," type of sample, container,
122
-------
and preservative furnished by the contract laboratory. Equipment used to
sample the water was glass, stainless steel, or aluminum to prevent sample
con t ainina t ion.
Figure 1 is a photograph of the screen at the Gulf-Philadelphia Refinery,
where their intake water is obtained from the Schuylkill River. Figure 2 is
representative of the intake water, this particular sample point is the dis-
charge side of the pump at the Gulf-Philadelphia Refinery. Figure 3 is a
photograph of the corrugated plate interceptor used at Hunt Oil in Tuscaloosa,
Alabama; this sampling point is also shown in Figure 4 which is a bypass line
to the pH meter. Figure 5 is a photograph of the DAF unit the Hunt Refinery
uses for a final clarifier; and Figure 6 is the sampling point which is prior
to the holding pond.
The six refineries selected for "screening only" have been visited and
sampled. At this point in time, a draft has been prepared to describe the
refinery's wastewater treatment system, and analyses have been concluded and
recorded for the parameters with the exception of volatile organics and non-
volatile organics.
The American Petroleum Institute contracted with Exxon Research to
accompany the EPA sampling team and obtain "replicate samples" at Exxon-
Baytown. In addition, Gulf-Philadelphia; Texaco-Lockport; Mobil-Augusta;^ and
Exxon - Baytown provided company personnel to accompany the EPA team to
obtain replicate samples. A request has been made to each refinery which
obtained "replicate samples" to furnish their data for inclusion- in the
report.
Activated Carbon Screening Study
EPA's Effluent Guidelines Division also selected six petroleum refineries
for activated carbon pilot scale screening studies. The petroleum refineries
selected are:
COMPANY LOCATION CLASS TREATMENT SYSTEM
1. Shell Anacortes, Wash. B activated sludge
2. Phillips Sweeney, Tex. C aerated lagoon
3. Exxon Benecia, Cal. B activated sludge
4. Exxon Billings, Mont. C aerated lagoon
5. Coastal States Corpus Christi, Tex. C activated sludge
6. ARCO Philadelphia, Pa. B activated sludge
The purpose of this pilot scale screening study is to determine if gran-
ular activated carbon possesses the capability to adsorb any of the priority
pollutants and the treatment effectiveness encountered when powdered activated
carbon is used to supplement a biological treatment system which is treating
petroleum refinery wastewaters.
There are two general types of activated carbon treatment which are
being investigated during this screening study. The granular carbon system
consists of placing a bed of granular carbon in an enclosed container and
123
-------
percolating filtered biotreated water through the bed in the downflow mode.
The second system is a powdered activated carbon system, where the powdered
carbon is mixed with the mixed liquor of an activated sludge system.
A. Granular Activated Carbon System
During this screening, three virgin activated carbons from three carbon
manufacturers and one activated carbon, which had been previously used at a
refinery's activated carbon system and regenerated for recycle, are being
investigated.
The refinery's bio-effluent is pumped through a multi-media filter
column to remove suspended solids, insoluble oils, and pin point floe and
scum which may be in the final effluent. The multi-media thick wall glass
column is 5 ft. in length and 6 in. in diameter. On the bottom of the filter
column is a stainless steel screen. Three inches of limestone rock are
placed on the screen followed by 6 in. of washed sand and 18 in. of anthra-
filt media for the top layer. Figure 7 is a diagram of the filter system
used during the carbon screening study. When the pressure in the column
exceeds 15 psig, the column is backwashed with either potable or carbon
treated water, and an alternate multi-media filter is used to supply filtered
bio-effluent to the carbon columns.
RSKERL has been supplied with virgin granular activated carbon from the
following manufacturers who recommended the specific carbon to be used to
treat petroleum refinery effluent:
1. Calgon Filtrasorb 300, 8 x 30 mesh
2. ICI Hydrodarco 3000, 8 x 30 mesh
3. Wesvaco WVG, 12 x 40 mesh
In addition to these three carbon sources, the Atlantic Richfield Co. (ARCO)
supplied a source of Calgon1s Filtrasorb 300 carbon which had been used and
regenerated at ARCO's Watson Refinery's activated carbon treatment system.
Figure 8 is a schematic of the granular carbon pilot scale system designed
for this study.
Each column has a stainless steel sieve screen in the bottom of the col-
umn with 3 in. of limestone serving as a support bed for the specific carbon.
Fifteen pounds of activated carbon was placed on top of the limestone rock
for a carbon bed depth of 36 to 40 in. Multi-media filtered effluent is
pressured into the top of each of the columns and the water percolates
through the column into the discharge pipe which contains a valve limiting
the flow to 0.25 g/m for each column. Figure 9 is a schematic of the granular
carbon pilot system.
An alternate carbon system employing powdered activated carbon treatment
is being evaluated at the four refineries where activated sludge treatment is
employed. In this system, powdered activated carbon is added directly into
the mixed liquor basin, and.the carbon residual maintained at 4000 mg/1.
124
-------
RSKERL designed and fabricated a 0.25 g/m complete mix activated sludge
system from carbon steel. A schematic of the powdered carbon pilot scale
system is shown in Figure 10. To acclimate the pilot system, mixed liquor
from the refinery's aeration basin is pumped into the pilot plant's aeration
basin, and the refinery's return sludge is added to the clarifier. Concen-
trations of suspended solids in the mixed liquor tank are maintained to
approximate the full-scale plant's operation. During the initial period of
acclimation in the pilot plant, the biological plant influent will be fed
to the pilot scale aeration basin. Total organic carbon analyses are used to
determine if the pilot scale effluent quality is essentially the same as the
full-scale system.
On Day 5 of the study, five pounds (2,250 grams) of the selected powdered
carbon, which was predetermined by an isotherm selection procedure, will be
added to the inlet of the bio-plant in a slurry over a 10-hour time period.
Effluent from the final clarifier is collected in a sump; the water from
the sump flows through glass wool to catch any powdered carbon which might be
discharged. The glass wool and sump are emptied back into the aeration basin,
thereby maintaining a near constant powdered carbon concentration in the
pilot system.
CARBON SCREENING SAMPLING
Priority Pollutant Program
Screening samples are obtained from the intake, API separator, and NPDES
sources on three consecutive days; the protocol, preservation, and procedures
are the same as accomplished with the first six refineries. These same
samples will be obtained on the sixth day, ninth day, twelfth day, and
fifteenth day. A schedule of sampling periods for the priority pollutants
is shown in Table 6.
Beginning with Day 6, the carbon screening pilot plant will be sampled
and again on Days 9, 12, and 15. The sampling frequency and sampling loca-
tion program is shown in Table 7.
The collected samples are air-freighted to Ada in ice chests. Upon
receipt of the samples, phenolics and cyanides are analyzed to meet 24-hour
preservation criteria. The three 24-hour composite non-volatile organics
samples are composited into one sample and extracted in accordance with the
March protocol. The extracts are forwarded to a contract laboratory for
GC/MS analysis. Metals samples are digested and analyzed by the March proto-
col procedures and forwarded to EPA's Region V Laboratory where they will be
analyzed by plasma emission spectrometer for comparative purposes. The
three 24-hour volatile organics samples are composited and the composite
forwarded to a contract laboratory for analysis. Split extracts for the non-
volatiles are sealed in glass ampules and maintained a 4°C, and will be
analyzed by an EPA laboratory to provide quality control information on the
study.
125
-------
During the carbon study, total organic carbon analysis is obtained every
eight hours on the influent to the mixed-media filter and the effluents from
the mixed-media filter activated carbon columns 1, 2, 3, 4, plus the pilot
scale activated sludge influent and final effluent.
In addition to the priority pollutants, Effluent Guidelines Division
contracted laboratory analyses for the same "classic parameters" that were
analyzed in the Phase I screening study.
A final report will be prepared which will, in essence, report all
analytical data obtained at each refinery, the refinery's data, analytical
quality control, and description of the refinery, crude oil sources, etc.
REFERENCES
(1) Public Law 92-500, 92nd Congress, S. 2770, October 18, 1972, Page 1.
DISCUSSION
F. L. Robertaccio; The methodology used for the "carbon-biological" system
being screened in this study is insufficient to be an adequate representation.
I would like your comments on this remark and your opinion on how the aspect
of your study might effect selection of models for establishment of effluent
guidelines.
Leon H. Myers; First, I don't know how it will affect guidelines, that is
out of my baliwick; I am in the research department. I mentioned, and I
mentioned on purpose, that we only had three constraints. Time, money, and
manpower. We are at each refinery 15 days, about four of those days are
taken up in the initial screening. That leaves about 10 days that we are
there running the carbon systems. I can't disagree with you, we can't get
one good sludge age, much less two or three sludge ages to look at the
powdered system at all. In my opinion, to use any of this data, you would
have to use it very carefully. You have to remember that we are operating
at a quarter GPM over a small time period. Our inclination is that we need
at least a 10 GPM pilot scale study conducted parallel with the full-scale
pilot scale study on powdered activated carbon and about 10 GPM granular
carbon study conducted at the same plant, over about a year's period of time
because of seasonal variations. We have discussed this with the effluent
guidelines personnel and with research headquarters personnel. We haven't
found anyone to disagree with us, we also haven't found any funding response
Does that answer your comments?
126
-------
BIOGRAPHY
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 Sources Section,
of the Source Management Branch at the Robert S.
Kerr Environmental Research Laboratory, Ada,
Oklahoma.
127
-------
TABLE 1. SETTLEMENT AGREEMENT LIST OF 65 POLLUTANTS
Acenaphthene
Acrolein
Acrylonitrile
Aldrin/Dieldrin
Antimony and compounds
Arsenic and compounds
Asbestos
Benzene
Benzidine
Beryllium and compounds
Cadmium and compounds
Carbon tetrachloride
Chlordane (technical mixture and metabolites)
Chlorinated benzenes (other than dichlorobenzenes)
Chlorinated ethanes (including 1, 2-dichloroethane, 1,1,1-trichloroethane,
hexachloroethane)
Chloroalkyl ethers (chloramethyl, chloroethyl, and mixed ethers)
Chlorinated naphthalene
Chlorinated phenols (other than those listed elsewhere; includes trichloro-
phenols and chlorinated cresols)
Chloroform
2-chlorophenol
Chromium and compounds
Copper and compounds
Cyanides
DDT and metabolites
128
-------
Dichlorobenzenes (1,2-,1,3-, and 1,4-dichlorobenzenes)
Dichlorobenzidine
Dichloroethylene (1,1-and 1,2-dichloroethylene)
2,4-dichlorophenol
Dichloropropane and dichloropropene
2,4-dimethyl phenol
Dinitrotoluene
Diphenyihydrazine
Endosulfan and metabolites
Endrin and metabolites
Ethyl benzene
Fluoranthene
Haloethers (other than those listed elsewhere; includes chlorophenyphenyl
ethers, bromophenylphenyl ether, bis (dischloroisopropyl)ether,
ether, bis-(chloroethoxyl) methane and polychlorinated diphenyl
ethers)
Halomethanes (other than those listed elsewhere; includes methylene chloride,
methyl chloride, methyl bromide, bromoform, dichlorobromomethane,
trichlorof1uoromethamne, dichlorodif1uoromethane)
Heptachlor and metabolites
Hexachlorobutadi ene
Hexachlorocyclohexane (all isomers)
Hexachl orocycl opentadi ene
Isophorone
Lead and compounds
Mercury and compounds
Naphthalene
Nickel and compounds
Nitrobenzene
129
-------
Nitrophenols (Including 2,4-dinitrophenol, dinitrocresol)
Nitrosamines
Pentachlorophenol
Phenol
Phthalate esters
Polychlorinated biphenyls (PCBs)
Polynuclear aromatic hydrocarbons (including benzanthracenes, benzopyrenes,
benzofluoranthene, chrysenes, diben-
zanthracenes, and indenopyrenes)
Selenium and compounds
Silver and compounds
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
Tetrachloroethy1ene
Thallium and compounds
Toluene
Toxaphene
Trichloroethylene
Vinyl chloride
Zinc and compounds
130
-------
TABLE 2. SPECIFIC COMPOUNDS
Compound
1. *acenaphthene
2. *acrolein
3. *acrylonitrile
4. *aldrin
5. *dieldrin
6. *benzene
7. *benzid1ne
8. *carbon tetrachloride (tetrachloromethane)
9 *chlordane (technical mixture & metabolites)
Chlorinated benzenes (other than dichlorobenzenes)
10. chlorobenzene
11. 1,2,4-trichlorobenzene
12. hexachlorobenzene
Chlorinated ethanes (including 1,2-dichloroethane, 1,1,1-
trichloroethane and hexachloroethane)
13. *l,2-dichloroethane
14. *l,l,l-trichloroethane
15. *hexachloroethane
16. 1,1-dichloroethane
17. 1,1,2-trichloroethane
18. 1,1,2,2-tetrachloroethane
19. chloroethane
Oiloroalkyl ethers (chloromethyl, chloroethyl and mixed ethers)
20. *bis(chloromethyl) ether
21. *bis(2-chloroethyl) ether
22. 2-chloroethyl vinyl ether
Chlorinated naphthalene
23. 2-chloronaphthalene
1-bromodecane Std.
1-bromododecane Std.
Chlorinated phenols (other than those listed elsewhere; includes
trichlorophenols and chlorinated cresols)
24. 2,4,6-trichlorophenol
25. p-chloro-m-cresol
26. *chloroform (trichloromethane)
131
-------
27. *2-chlorophenol
DDT and netabol ites
28. *4,4'-DDT
29. 4,4'-DDE
30. 4,4'-DDD (p.p'-TDE)
Dichlorobenzenes (1,2-; 1,3-; and 1,4-dichlorobenzenes)
31. *l,2-dichlorobenzene
32. *l,3-dichlorobenzene
33. *l,4-dichlorobenzene
Di chl orobenzi di ne
34. 3,3'-dichlorobenzidine
achloroethylenes (1,1-dichloroethylene and 1,2-dichloroethylene)
35. *1,1-di chloroethy1ene
36. *1,2-trans-dichloroethylene
37. *2,4-dichlorophenol
Di chl oropropane and di chl oropropene
38. 1,2-dichloropropane
39. 1,3-dichloropropylene (i,3-dichloropropene)
40. *2,4-dimethylphenol
Dini trotoluene
41. 2,4-dinitrotoluene
42. 2,6-dinitrotoluene
43. *1,2-di pheny1hydrazi ne
Endosulfan and metabolites
44. *orendosulfan
45. *3i-endosulfan
46. endosulfan sulfate
Endrin and metabolites
47. *endrin
48. endrin aldehyde
49. endrin ketone
132
-------
50. *ethylbenzene
51. *fluoranthene
Haloethers (other than those listed elsewhere)
52. *4-chlorophenyl phenyl ether (p-chlorodlphenyl ether)
53. *4-bromophenyl phenyl ether
54. *bis (2-chloroisopropyl) ether
55. *bis (2-chloroethoxy) methane
Halo?rethanes (other than those listed elsewhere)
56. *methylene chloride (dichloromethane)
57. *methyl chloride (chloromethane)
58. *methyl bromide (bromomethane)
59. *bromoform (tribromomethane)
60. *dichlorobromomethane
61. *trichlorofluoromethane
62. *dichlorodifluoromethane
63. chlorodibromomethane
Heptachlor and metabolites
64. *heptachlor
65. heptachlor epoxide
66. *hexachlorobutadiene
Hsxachlorocycloharane (all isomers)
67. *a-BHC
68. *|3-BHC
69. *Y-BHC (lindane)
70. *<5-BHC
71. *hexachlorocyclopentadiene
72. *isophorone
73. *naphthalene
74. *n1trobenzene
MtrophenoZs (including 2,4-dinitrophenol and dinitrocresol)
75. 2-nitrophenol
76. 4-nitrophenol
77. *2,4-dinitrophenol
78. 4,6-dinitro-o-cresol
Mtrosani nes
79. N-nitrosodimethylamine
80. N-nitrosodi-n-propylamine
81. N-nitrosodiphenylamine
133
-------
82. *pentach1orophenol
83. *phenol
Phthalate esters
84. bis (2-ethylhexyl) phthalate
85. butyl benzyl phthalate
86. di-n-butyl phthalate
87. diethyl phthalate
88. dimethyl phthalate
Pol ychl orz not ed bi phenyl s (PCff s)
89. PCB-1242 (ArocUor 1242)
90. PCB-1254 (Arochlor 1254)
Polynuclear aronntic hydrocarbons (including benzanthracenes,
benzopyrenes, benzofluoranthene, chrysenes, dibenzanthracenes,
and indenopyrenes)
91. 1,2-benzanthracene
92. benzo[a]pyrene (3,4-benzopyrene)
93. 3,4-benzofluoranthene
94. 11,12-benzofluoranthene
95. *chrysene
96. acenaphthylene
97. anthracene
98. 1,12-benzoperylene
99. fluorene
100. phenanthrene
101. l,2:5,6-dibenzanthracene
102. indeno (l,2,3-C,D)pyrene
103. pyrene
104. *253,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
105. *tetrachloroethylene
106. *toluene
107. *toxaphene
108. *trichloroethylene
109. *vinyl chloride (chloroethylene)
1-brompdecane (possible internal standard)
1-bromododecane (possible internal standard)
TOTAL METALS
1. Arsenic 5. Lead 9. Antimony 13. Zinc
2. Beryllium 6. Mercury 10. Copper 14. Asbestos
3. Cadmium 7. Nickel 11. Selenium 15. Cyanides
4. Chromium 8. Thallium 12. Silver
134
-------
Table 3
REFINERIES SELECTED FOR THE PRIORITY POLLUTANT STUDY
Refinery
Gulf
Exxon
Hunt
Clark
Texaco
Mobil
Phillips
Shell
Exxon
Exxon
Coastal States
ARCO
Location
Philadelphia, Pa.
Bay town, Tex.
Tuscaloosa
Hartford, 111.
Lockport , 111 .
Augusta, Kan.
Sweeney, Tex.
Anacortes, Wash.
Benecia, Calif.
Billings, Mont.
Corpus Christi, Tex.
Philadelphia, Pa.
Refinery
Class
C
E
A
B
B
B
C
B
B
C
C
B
Treatment System
Trickling Filter
Activated Sludge
Aerated Lagoon
_ Activated Sludge
Activated Sludge
Filtration
Activated Sludge
Oxidation
Aerated Lagoon
Activated Sludge
Activated Sludge
Aerated Lagoon
Activated Sludge
Activated Sludge
135
-------
Table 4
PRIORITY POLLUTANTS SAMPLING INFORMATION
Parameter
Non Volatile
Metals
Mercury
Cyanide
Phenolics
Asbestos
Volatile
24-hour
Composite
X
X
Grab
Sample
X
X
X
X
X
Sample
Container
1 Gallon Glass
1 Gallon Plastic
1 Quart Plastic
1 Quart Plastic
1 Quart Glass
1 Quart Plastic
40 ml Vial
Preservative
Ice
HN03
HN03
NaOH
H3P04
Ice
Ice
Table 5
CLASSICAL PARAMETERS SAMPLING INFORMATION
Parameter
BOD5
TSS
COD
TOC
NH3-N
24-hour
Composite
X
X
X
X
X
Grab Sample
Sample Container
Plastic
Plastic
Plastic
Plastic
Plastic
Preservative
Cool 4°C
Cool 4°C
H2SOA. to pH<2
H2S04 to pH<2
Plus Cool, 4°C
HoSO^ to pH<2
Cr° X
Sulfide X
Oil & Grease
Plus Cool, 4°C
Plastic HN03 to pH<2
Plastic 2 ml Zinc Acetate
Glass H2SOA. to pH<2
Plus Cool, 4°C
136
-------
Table 6
SCHEDULE OF SAMPLING PERIODS
Day 1 234 5678
W Th F S Su M T W
S S S S S
C C C C C
R R R R R
E E E E E
E E E E E
N N N N N
Sample locations are depicted in
9 10 11 12 13 14 15 1617
Th F S Su M T W Th F
S S S
C C C
R R R
E E E
E E E
N N N
Figure 11.
Table 7
SAMPLING FREQUENCY
Sample Site
Intake
Separator
Bio-Effluent
Mixed-Media Filter Eff.
Activated Carbon Eff.
Reaenerat.pH farhnn Fff.
AND LOCATION PROGRAM
Day
1 2 3 6 9 12 15
XXX
XXX
X X X X X X X
X X X X
X X X X
X X X X
Activated Sludge/Activated
Carbon Influent
Activated Sludge/Activated
Carbon Eff.
137
-------
i
Figure 1. INTAKE SCREEN HOUSE
Figure 2. INTAKE WATER SAMPLE POINT
138
-------
Figure 3. Corrugated Plate Interceptor-Hunt/Tuscaloosa
Figure 4. Bypass Line From Corrugated Plate Interceptor
Hunt/Tuscaloosa
139
-------
Figure 5. DAF Unit-Hunt/Tuscaloosa
Figure 6. Final Effluent-Hunt/Tuscaloosa
140
-------
Bio-Tree ted
Effluent
Pressure
© GouQe
„
:v.«/-*•* <""'-'
!'4 // ' . ' ' * '.'•••'
'
'.-" *• /V**' v«
-' * X / , *•'
To Corbon
Columns
3 Limestone
Rock
Stainless Steel Screen
Teflon Gasket
Figure 7. MIXED-MEDIA FILTER DIAGRAM
141
-------
Multi-Media
Filter Effluent
Pressure Gauge
Teflon Gasket
-•—Glass Column
36 Carbon
3 Limestone Rock
Teflon Gasket
Carbon Treated
Water
Figure 8. CARBON COLUMN
142
-------
Bio- Effluent
i
/ /
/ t
n '
t
>
f
/ i
•.
t i
t
t /
.
K
,, Ul /
, ' 1
y "
/ 4
m*f J
_J
~* / .
«J X
'' */
/ Ul'
/. • 2 ' '
'i ,
4
* /
4
y x '
//
•J /
H y
5
/ /
^'x
'/"/'
,
1
4
r 1
^
o
a:
«•
^
•i
^
>
M>
0
z
CARBON
2.
' 1
f
5
n»
••
^^
•
>
2>
CARBON
| 1
t
o
fl:
>
10
o
^
CARBOI
P
a
ui
i_
oc
UJ
^
UI
0
111
^
CARBO
V T T T
Figure 9. PICOT CARBON SYSTEM
-------
Mixer.
-Air Manifold
m
g/in Bio-
System
Effluent
Final Clarifier
Weir
Sludge Blanket
Effluent
GlassWool
Packed
Sump
Figure 10. PILOT SCALE ACTIVATED SLUDGE SYSTEM
-------
"CONSIDERATIONS FOR DEFINING SUBSTANCES
HAZARDOUS TO THE ENVIRONMENT"
A. Karim Ahmed
Staff Scientist, Natural Resources Defense Council
I am pleased to be here today to address this conference
about one of the most important programs embarked upon by the
federal government to regulate the proliferation of toxic sub-
stances in our environment. This morning, Mr. Ridgeway Hall of
the Environmental Protection Agency (EPA) gave us an excellent
presentation of the events that led to the settlement agreement
- known to some of us as The Consent Decree - between the Natural
Resources Defense Council (NRDC) and other environmental groups
and the EPA. This out of court settlement, which was approved by
the U.S. District Court for the District of Columbia on June 9,
1976,! is a wide-ranging regulatory strategy, with a clearly de-
fined timetable, for controlling the discharge of toxic sub-
stances into our nation's waterways, as required under the
Federal Water Pollution Control Act (FWPCA).2 since we have
heard a good deal about EPA's present and proposed implementation
of the Consent Decree from Mr. Hall and other speakers, I will
attempt to give you a somewhat different overview of the same
subject matter - namely the problem we face when we wish to de-
fine substances that are hazardous to the environment. I will do
so in the context of reviewing with you some of the major federal
statutes that deal with the hazardous or toxic substances issue.
I would like to conceptually divide major federal statutes
that have dealt with the problems of toxic substances in the past
and do so currently into two main divisions: (1) statutes that
primarily address toxic substances in terms of their effects on
human health, and (2) statutes that were intended to regulate the
impacts of toxic substances on the environment. Now I have pur-
posely divided this into black-and-white terms and I will try to
explain to you why. First, let's look at the statutes that deal
with problems of human health. These generally come under the
purview of the Food and Drug Administration, under the Food, Drug
and Cosmetic Act;3 the Department of Labor, under the Occupa-
tional Safety and Health Act;4 and the Consumer Product Safety
Commission, under the Consumer Product Safety Act,5 and the
Federal Hazardous Substances Act.6 These statutes, which I will
explain as we go along, primarily deal with effects on human
health, and are regulated by the three federal agencies I have
just mentioned.
On the other hand, the EPA, which was created in the early
70fs through the reorganization of the federal bureaucracy, was
given jurisdiction to regulate impacts on the environment. How-
ever, as we have now discovered, the environment also includes
man. The EPA has a particularly difficult task in trying to de-
fine toxic substances which have, first of all, an impact on the
145
-------
environment, and secondly, an impact on human health, or both.
The various statutes that EPA has authority under are the Federal
Water Pollution Control Act,7 the Clean Air Act,8 the Toxic Sub-
stances Control Act,9 the Safe Drinking Water Act,lu the Resource
Conservation and Recovery Act,11 and the Ocean Dumping Act.1-'
EPA also regulates pesticides under the Federal Insecticide, Fun-
gicide and Rodenticide Act.1^
The problem that we're faced with here is essentially one
of overlapping jurisdiction. Let me give you examples of some of
the problems we have had of late. When the issue of fluorocar-
bon's impact on stratospheric ozone was first brought to the at-
tention of the federal agencies, we had complete chaos- In fact,
the first thing that the agencies did was to run to the Depart-
ment of Justice to ask for a memorandum that would sort out the
issue of legal jurisdiction. The Natural Resources Defense Coun-
cil had petitioned the Consumer Product Safety Commission (CPSC)
who, in spite of having jurisdiction over only about five to ten
percent of the total aerosol products on the market, had strong
regulatory authority under the Consumer Product Safety Act.14
When we confronted them with our petition, the Commission argued
at first that the problem was one of environmental concern and
only indirectly would affect human health - i.e. the increase of
ultra-violet radiation, caused by environmental loss of ozone
would indirectly lead to an increase in skin cancer incidence.
So, the CPSC felt that it really didn't fit under their jurisdic-
tion. Similarly, the FDA said that they had statutory authority
with foods, drugs, and cosmetics as they affect human health
directly, but had not been given authority to deal with environ-
mental issues. Lastly, the EPA put a final ironic twist to this
regulatory comedy by claiming that under the Clean Air Act it had
jurisdiction on the lower atmosphere only. They were not sure
about the stratosphere; perhaps some other agency dealt with the
upper atmosphere. So this classic "passing-the-buck" game began
back in 1974. At that time, this issue was too new and unfamil-
iar, and somewhat controversial, for any agency to want to han-
dle. We had a really troublesome situation where none of the
agencies knew where their jurisdiction lay and how they should go
about regulating products under statutes that defined their au-
thority narrowly, or at best, ambiguously.
Another example, which is just beginning to emerge, is the
problem of genetic engineering. In genetic engineering, we have
a naturally occurring substance called DNA which, as you know, is
the subject of biological manipulation. One would surely think
that the Department of Health, Education and Welfare, under the
U.S. Public Health Service Act,15 would have clear jurisdiction
on this issue since, under credible experimental conditions, new
life forms may be produced, such as a bacterial system that could
be potentially infectious. But the EPA now claims that, under
the Toxic Substances Control Act (TSCA), DNA may be defined as a
chemical substance. Consequently, under certain provisions of
146
-------
TSCA, it may have jurisdiction. Here, they're claiming jurisdic-
tion, rather than avoiding it. In a sense, we have another po-
tential conflict here - one of asserting jurisdiction.
With respect to the fluorocarbon-ozone issue that I men-
tioned earlier, what subsequently happened was a formation of an
inter-agency committee under the auspices of the Council on Envi-
ronmental Qulaity (CEQ) leadina to a thorough study by the Na-
tional Academy of Sciences.ib'17 A few months ago, as you know,
all three agencies (the CPSC, and FDA and the EPA) jointly pro-
posed regulations governing the ban of fluorocarbons as propel-
lants in aerosol products.18 For the first time, a kind of coop-
erative regulatory venture is being experimented with by the
federal agencies on a hazardous substance that does not fit under
neatly defined jurisdictions.
There are, on the other hand, some hazardous chemical sub-
stances that appear to be clearly defined under different stat-
utes. Let me give you an example - asbestos is one of the best.
If one has a problem of asbestos in water, one would go to the
EPA, since under the Federal Water Pollution Control Act or the
Safe Drinking Water Act they have jurisdiction.19 If it's an
air related problem, then one would also go to the EPA, for under
Section 112 of the Clean Air Act, they have promulgated regula-
tions with respect to asbestos emissions.20 if it's a problem in
consumer products, for example, in spackling compounds, artifi-
cial fireplace logs, or asbestos-containing ceiling tile, one
would go to the Consumer Product Safety Commission which has ju-
risdiction either under the Consumer Product Safety Act or the
Federal Hazardous Substance Act.21 If it's a matter of asbestos
in talcum powder or toiletries, one would go to the FDA who have
clear authority under the Food, Drug and Cosmetic Act.22 It it's
a labeling question - a label to warn consumers - one would go to
one of the line agencies or directly to the Federal Trade Commis-
sion (FTC) which could issue regulations under the Federal Fair
Labeling and Packaging Act.23 If it's a workplace hazard, one
would go to the Department of Labor's Occupational Safety and
Health Administration for setting up occupational standards.z-
And finally, if it's a universal problem, one would go back to
the EPA. Why? Because they now have the Toxic Substances Con-
trol Act,25 enacted into law last year, which is supposed to take
care of everything generically and completely. So, the EPA has
the biggest task again.
Let's get down to specifics in defining substances that are
hazardous. One way of defining substances that are hazardous to
the environment is to look at it in a narrow "legal" sense. It
is reasonable to expect that the statutes were written at^differ-
ent times for different reasons, since the congressional intent
for enacting such laws tend to be rather different. But when one
examines the various statutes on th« books, one find%that ^orae
of the laws do not have an explicit definition of what is meant
147
-------
by the term "toxic." They might, for example, address the issue
of toxicity by alluding to questions of safety and unsafety in
very general terms. This is particularly true with the Federal
Food, Drug and Cosmetic Act (FDCA). The FDCA has lengthy sec-
tions on procedural matters, about burden of proof, etc., but the
actual definition of toxic substances is never made clear. This
also appears to be the case under the Federal Insecticide, Fungi-
cide and Rodenticide Act (FIFRA), which has no definition of what
is a hazardous or harmful pesticide. However, a clearer state-
ment about hazardous substances can be found in the Federal Haz-
ardous Substances Act (FHSA). This particular Act is interesting
because it defines "hazardous" in very narrow terms and yet has
broad application. Under Section 2(f) of the FHSA, "hazardous
substance" means:
"Any substance or mixture of substances which
is (1) toxic, (2) corrosive, (3) ah irritant,
(4) a strong sensitizer, (5) flammable or com-
bustible, and (6) generates pressure through
decomposition, heat or other means, if such
substances or mixtures may cause substantial
injury or substantial illness...."
Notice here that the definition deals with either health or
safety of human beings. It does not deal with environmental is-
sues. (This is an important section of the Act because later I
will go back to it when I talk about a proposed EPA regulation.)
Under Section 2(g) of the same Act, we have a definition of what
is "toxic:"
"Any substance (other than a radioactive sub-
stance) which has the capacity to produce in-
jury or illness to man through ingestion, inha-
lation or absorption through any body surface."
Here, again, toxic is defined with respect to human beings. In
Section 2(h), we have a definition of "highly toxic," which is
defined in terms of three explicit guidelines that are employed
in animal tests: (1) for ingested substances, an "LD^Q value which
is equal to or less than 50 milligrams per kilogram; (2) for in-
haled substances, an LCso figure which is equal to or less than
200 parts per million, or 2 milligrams per liter; and (3) dermal
absorption, an LDso figure equal to or less than 200 milligrams
per kilogram. We have here a definition of "toxic" or "highly
toxic" in a very conventional sense of the term - meaning acute
animal or human toxicity. There is no reference to chronic tox-
icity, nor does it refer to environmental harm of any kind. Ob-
viously, the intent of this Act was to regulate products or sub-
stances that would acutely injure human beings.
We should now examine statutes that simultaneously deal
with human and environmental effects. The best place to start,
148
-------
I think, is the Toxic Substances Control Act which, even thouqh
it is the most recently passed law, is causing a great deal of
confusion in the minds of a lot of people who have to deal with
this statute. Let us examine whether we have any definition of
a hazardous or toxic substance in this Act. It first occurs un-
der Section 4, which is the section that deals with testing of
chemical substances and mixtures, and here it is basically de-
fined in terms of substance or mixture which "may present an un-
reasonable risk of injury to health or the environment." That
is, the definition is very broad. In subsection (2) (A) of Sec-
tion 4, mention is made of the kinds of standards or testing that
would be required to assess what is "unreasonable" - that is,
what are the unreasonable risks to health and the environment -
and mentioned are tests for carcinogenesis, mutagenesis, terato-
genesis, behavioral disorders, cumulative or synergistic dis-
orders, etc., in addition to requiring testing for acute toxic-
ity, subacute toxicity, and other chronic toxicity. The same de-
finition occurs again under Section 6 of the Act, wnich is the
section that deals with the actual regulation of hazardous sub-
stances, that provides the agency with the authority to remove a
toxic substance from the marketplace:
"If the Administrator finds that there is a
reasonable basis to conclude that the manufac-
ture, processing, distribution in commerce,
use, or disposal of a chemical substance or
mixture, or that any combination of such ac-
tivities , presents or will present an unrea-
sonable risk of injury to health or the envi-
ronment , the Administrator shall by rule apply
one or more of the following requirements to
such a substance or mixture...."
Finally, in Section 7 of the Toxic Substances Control Act,
we have the Imminent Hazards section which, I think, is an impor-
tant provision to recognize. Here, the Administrator is given
certain additional powers to commence action when substances or
mixtures possess an "imminent or unreasonable risk of serious or
widespread injury to health or the environment." And, once
again, the theme is both health and the environment.
We shall now examine the Federal Water Pollution Control
Act and see how clearly the term "hazardous substance" is de-
fined. We find that we really do not have an unequivocal answer.
In a sense, it is a bit of a mess because when one examines the
FWPCA even an experienced environmental attorney has a difficult
time in figuring out what the correct definitions are. Appar-
ently the only place in the FWPCA where the term "toxic pollu-
tant" is defined is under Section 502, paragraph 13:
"The term 'toxic pollutant1 means those pollu-
tants or combination of pollutants including
149
-------
disease causing agents which...on the basis
of information available to the Administrator,
cause death, disease, behavioral abnormalities,
cancer, genetic mutations, physiological mal-
functions (including malfunctions in reproduc-
tion) or physical deformities, in such organ-
isms or their offspring."
It would appear to be all inclusive. One, would not use
such a broad definition to select a list of highly toxic pollu-
tants. However, this is the manner by which "toxic pollutant"
has been defined under Section 307(a), which deals with the de-
velopment of effluent standards of toxic pollutants in the FWPCA.
The prohibition of toxic pollutants in toxic amounts is clearly
stated in Section 101(a)(3) of the Act which, when combined with
Section 502(13), gives us the "correct" intent of Section 307(a).
This is apparently the way the concept of "toxic pollutant" is
defined under the Act.
In Section 311 of the FWPCA, on the other hand, we see a
completely different approach since it deals with the liability
of an accidental spill. But it has a broad definition of "dis-
charge" in this section, for it "includes, but is not limited to
spilling, leaking, pumping, pouring, emitting, emptying, or
dumping." One would think that Section 311 would be even more
all-inclusive in terms of point discharges than accidental
spills; moreover, it goes on to define "hazardous substances" as
those that may "present an imminent and substantial danger to the
public health or welfare including, but not limited to, fish,
shell-fish, wildlife, shoreline and beaches." Interestingly
enough, there is also a provision in the FWPCA giving the Admin-
istrator certain emergency powers under Section 504, where a si-
milar language is repeated, i.e. "presenting an iminent and sub-
stantial endangerment to the health of persons and to the welfare
of persons...." What we have here is a provision that, on the
one hand, appears to be quite restrictive in defining emergency
powers of the Administrator and, at the same time, defines haz-
ardous substances quite broadly. We must therefore ask: which of
the two statutory "sets" is greater? Is Section 307(a) a "sub-
set" of Section 311; are they two separate, disjunctive "sets;"
or do they overlap? We don't really have clear answers to
this/ in part because we don't have a Section 311 regulation pro-
mulgated as yet. This only compounds the present regulatory con-
fusion.
Let us now examine the problem of defining "hazardous sub-
stance" from a more technical or scientific point of view. In
the past, promulgated regulations, reflecting the intent of fed-
eral statutes, emphasized acute toxicity - for example, the
Federal Hazardous Substances Act which, as I mentioned earlier,
only dealt with short-term acute toxicity and defined toxicity in
'-erms of LDso's and LCso's, etc. More recently, there has been a
150
-------
growing concern about long-term chronic effects and conseauentlv
about carcinogenic and mutagenic effects. These toxic substances
exert their effects over a long time period when ingested in
small amounts by man or animal and may cause cancer with latency
periods of 20 to 40 years. As Mr. Hall mentioned this morning
that was one of the chief criteria used in selecting the priority
chemical substances under the settlement agreement. We are also
concerned with certain key environmental factors when defining
hazardous substances. These are generally questions of: (1) che-
mical persistence - biological and chemical degradibility, (2)
movement in the environment, i.e., how does it get transported -
by air, by water, or by soil, (3) bioconcentration in the envi-
ronment, including biomagnification in the food chain, and (4)
synergistic and cumulative effects of the chemical substance.
These are all now becoming part and parcel of defining hazardous
substances from a scientific point of view. In selecting sub-
stances to be regulated under the Consent Decree, all the above
factors were given appropriate weight, and we hope led to a bet-
ter selection process than had been used by the agency in the
past.
Lastly, I will briefly mention our views on the Consent
Decree. As you may recall, under Section 307(a) of the FWPCA,
the EPA had ninety days, to publish a list of toxic substances for
which effluent standards were to be established. We interpreted
this section of the Act to apply to those highly toxic or persis-
tent substances for which there would be neither a technology-
based nor an economically-based effluent standard. After initial
litigation by NRDC, EPA in July 1973 proposed a list of nine sub-
stances to be regulated under Section 307(a) of the FWPCA. Most
of the substances were pesticides, a few metal ions, and PCB. At
administrative hearings held by the EPA, the agency was literally
swamped with technical information from the affected industries
that claimed that the agency did not have an adequate scientific
basis for setting proposed new effluent standards. Consequently,
the EPA abandoned their proposal and did not promulgate effluent
standards under Section 307(a). NRDC filed additional suits
against EPA, claiming that there were a large number of chemical
substances (certainly greater than nine) that could be termed
toxic or hazardous and ought to-be regulated under Section
307(a).
To make a long story short, we finally ended in an out-of-
court settlement which is now known as the Consent Decree. In
arriving at the Consent Decree, it is very interesting to note
some of the issues that were agreed to by both sides. There was
a recognition that there were different provisions in the FWPCA
that could regulate toxic substances, and that Section 307(a) was
not necessarily the only handle that should be used. These addi-
tional provisions include Sections 301, 304 and 306. The basis
for developing a priority list under this concept was that the
agency was given more flexibility in dealing with toxic
151
-------
substances, some of which have only recently been shown to have
potential long term problems of carcinogenicity and mutagenicity.
Thus, instead of using Section 307(a) as the only basis for regu-
lating toxic substances, with its limited timetable schedules and
larger burden of proof on the agency, EPA would use a combination
of different provisions under the FWPCA, using a technology-based
and New Source Performance Standard approach as provided for in
Section 306. What we have now is a list of some 129 chemical
compounds that have been identified as priority substances. Ef-
fluent standards for six substances have already been promulgated
under 307(a) provisions (as required under the Consent Decree)
and the standards for the remaining substances will be issued un-
der BAT (best available technology) provisions of the Act.
The big puzzle that now remains is the development of regu-
lations under Section 311 of the FWPCA. The agency has proposed
a list of substances to be regulated under Section 311, which, I
think, is totally inadequate since most of the information on
hazardous substances in the proposed regulations is based upon
aquatic toxicity, and the agency has chosen to define Section 311
in those terms. And to make matters even more puzzling, EPA used
the same criteria of defining toxicity as the Hazardous Sub-
stances Control Act which, as I mentioned earlier, only defined
acute toxicity. The agency has essentially used the same LD$Q,
LCso figures in arriving at their conception of what is acutely
toxic to human beings. They have explicitly ignored questions of
chronic effects as is clear from the preamble to the proposed re-
gulations. 26 jn fact, EPA has made a point of not being con-
cerned with issues of chronic effects, for reasons which are to-
tally unclear. Consequently, these proposed regulations are
still pending and we do not know when they will be promulgated.
It is clear that whatever regulations will now be adopted
under the FWPCA, we will have to be concerned both with the im-
plementation of the Consent Decree and with the implementation of
the Toxic Substances Control Act. For example, a major issue
right now is the inventory reporting provision under TSCA. This
provision requires EPA to obtain from the affected industries a
complete knowledge of what chemical substances are being used, in
what amounts, and for what purpose. Under the Consent Decree,
EPA has been gathering similar information on substances on the
priority list. These same questions have not been addressed by
the agency under the Toxic Substances Control Act, where they are
given authority to seek such information. EPA has been subjected
to a lot of criticism by NRDC and others about the way they've
gone about collecting this critical bit of information. In the
final analysis, if we are to have meaningful control of the dis-
charge of toxic substances into our environment, we will have to
see a dovetailing of the effort between the Water Program staff
of the EPA, who are implementing the Consent Decree program, and
the staff of the Office of Toxic Substances, who are trying to
implement the Toxic Substances Control Act. Only then can we
152
-------
have confidence in the ultimate success of EPA's program to con-
trol the ever growing impact of hazardous substances on the envi-
ronment. Thank you very much.
REFERENCES
2.
3.
4.
5.
6.
7.
8.
9.
10.
11
12
13
14
15
16
17
18.
19.
20.
21.
22.
23.
24.
25.
26
Natural Resources Defense Council, et al. v. Train 8 E R c
2120 (D.D.C. 1976). ~ J* " '
Public Law 92-500; 33 U.S.C. §§1251 et seq.
21 U.S.C. §§301 et seq.
1590.
1207; 90 Stat.
Public Law 91-596; 84 Stat.
Public Laws 92-573, 94-273, 94-284; 86 Stat,
503-510, 514; also 15 U.S.C. §§2052 et seq.
15 U.S.C. §§1261 et seq.
Supra, Reference 2.
Public Law 91-604; 84 Stat. 1676.
Public Law 94-469; 15 U.S.C.
Public Law 93-523; 88 Stat.
Public Law 94-580; 42 U.S.C. §§6901 et seq.
Marine Protection, Research and Sanctuaries Act of 1972;
Public Laws 92-532, 93-254, 93-62, 94-326; 86 Stat. 1052; 88
Stat. 50; 88 Stat. 1430; 89 Stat. 303; 90 Stat. 725.
Public Laws 92-516, 94-51, 94-109, 94-140; 7 U.S.C. §§136 et
se
§§2601 et seq.
1660.
Petition of the Natural Resources Defense Council to the
Consumer Product Safety Commission, November 19, 1974.
42 U.S.C. §264; also petition of the Environmental Defense
Fund and the Natural Resources Defense Council to the De-
partment of Health, Education and Welfare, October 28, 1976.
Fluorocarbons and the Environment, report of the Federal
Task Force on Inadvertent Modification of the Stratosphere,
Council on Environmental Quality, 1975.
Halocarbons; Effects on Stratospheric Ozone, Panel on Atmos-
pheric Chemistry, National Academy of Sciences, Washington,
D.C., 1976.
42 Fed. Reg. 24536 et seq, May 13, 1977.
Supra, References 2 and 10.
Supra, Reference 18.
Supra, References 5 and 6.
Supra, Reference 3.
Public Law 93-608; 18 U.S.C. §1457.
Supra, Reference 4.
Supra, Reference 4.
^0 Fed. Reg. 59959-60017, December 30, 1975.
153
-------
BIOGRAPHY
A. Karim Ahmed is Staff Scientist
with Natural Resources Defense Coun-
cil, Inc. (NRDC), New York, New York,
and Adjunct Assistant Professor at
State University of New York, College
at Purchase. He holds a B.S. degree
in Physics from University of Karachi,
a M.S. degree in Chemistry and a Ph.D.
degree in Biochemistry from University
of Minnesota. Before assuming his
position with NRDC, Karim served as
Research Director for Minnesota Public
Interest Research Group, and Executive
Assistant to Director at Consumers
Union. He is a member of Air Pollu-
tion Control Association, American
Chemical Society, American Public
Health Association, New York Academy
of Sciences, and Scientist's Institute
for Public Information.
154
-------
SESSION V
PROBLEMS
Chairman
George J. Putnicki
Visiting Professor, Environmental Sciences
University of Texas at Dallas, Texas
Speakers
Dwight G. Ballinger
"EPA's Analytical Development Program for
Problem Pollutants"
Fred T. Weiss
"Fates, Effects and Transport Mechanism of
Pollutants in the Aquatic Environment"
Donald I. Mount
"Measuring Aquatic Impact of Toxic Contaminants"
Davis L. Ford
"An Overview of Advanced Treatment Systems"
155
-------
BIOGRAPHY George J. Putnicki
George J. Putnicki is currently a Visiting
Professor in the graduate programs in Environmental
Sciences at the University of Texas at Dallas. He
holds a B.S. in Civil Engineering from Marquette
University and a M.S. from Oregon State University.
Prior, Mr. Putnicki has had extensive experience
in the EPA Region Six and was Deputy Regional^
Administrator and Director of the Hazard Materials
Control and Surveillance Control Divisions.
Mr. Putnicki has received many awards and
commendations from the EPA and is a registered P.E.
in Oklahoma, Texas and Louisiana.
156
-------
ERA'S ANALYTICAL DEVELOPMENT PROGRAM FOR PRIORITY POLLUTANTS
Dwight G. Ballinger, Director
Environmental Monitoring and Support Laboratory - Cincinnati
INTRODUCTION
Previous speakers in this Open Forum have discussed the details
of the 1976 Consent Decree for priority pollutants and the development
of a list of elements and compounds for which standards are to be set.
I will focus my remarks on the analytical methods requirements involved
in the implementation of the Decree and describe briefly what EPA is
doing to meet these requirements.
Although the standards will be established by 1983, this will be
only the first step in reducing the volume of these hazardous materials
discharged to surface waters of the United States. The standards will
be incorporated in permits administered by the states and EPA and the
permit conditions will require the monitoring of waste discharges and
the reporting of the volume and concentration of these pollutants in
each discharge. Section 304(g) of the Federal Water Pollution Control
Act requires the Administrator of EPA to promulgate test procedures for
use in determining compliance with permit conditions. Under this section,
an analytical method must be selected for each parameter listed in the
permits and these procedures or an acceptable alternative, are to be
used in all waste monitoring. Test procedures were first published
in 1973 and were revised and expanded in the Federal Register of
December 1, 1976. In practice, the list of analytical methods are first
published as proposed, public comments are received and considered, and
then the test procedures are published as final regulations. The
methods in the original listing and later revisions were chosen in close
cooperation with the state agencies, other federal agencies, and method
standardization groups such as the committee for Standard Methods for
the Examination of Water and Wastewater and ASTM Committee D-19. The
petroleum industry is ably represented on Committee D-19 and the Committee
has provided significant input to the selection of analytical methods.
Basically, the analytical methods requirements for the priority
pollutants are the same as those for other pollutants a proven,
^resented at the Second Open Forum on Management of Petroleum Refinery
Wastewater, Tulsa, Oklahoma, June 7, 1977
157
-------
defensible test procedure for each of the elements and compounds for
which standards are set and monitoring is required.
PROBLEM DEFINITION
Such a simple statement, however, is deceptive. A better perspective
on the problem discloses a number of critical factors. The original
list of 65 toxic substances contained a number of groups of organic com-
pounds such as chlorinated benzenes, chlorinated phenols, haloethers, and
polynuclear aromatic hydrocarbons. When each of the individual compounds
in these groups are defined the list of specific compounds comes to about
120. If standards are set for each compound, a test procedure must be
available for that compound. To meet the requirements for monitoring, test
procedures for these compounds must be available by 1983. The majority of
these procedures will require analysis for complex organic structures.
These constituents originate in a wide variety of industrial wastes
including plastics manufacturing, the production of synthetic organic
chemicals, and, of course, petroleum refining. The priority pollutants
often occur as by-products rather than the principal product and are
therefore in relatively small amounts. While most of them have been
identified in waste streams, they have not often been measured in environ-
mental samples with adequate precision and accuracy. Preliminary liter-
ature searchs have failed to produce the methodology required and a sig-
nificant amount of new development will be necessary.
Among the list of compounds for which test procedures will be
required are a number of interest to the petroleum industry. In the
group of polynuclear aromatics are 16 individual compounds; there are
11 phenols and three nitrobenzenes on the list.
The anticipated maximum permissible concentrations in the permits
have yet to be established since they will be dependent on the best
available treatment, the relative toxicity, and other factors. These
concentrations, however, will be minimal and the analytical "target" for
methods development is 10 yg/1. Based upon previous work, even lower
working ranges may be necessary to detect and measure the specific
pollutants.
An appropriate test procedure must measure these concentrations of
the pollutant with positive qualitative identification, in wastewaters
containing many interferents of similar chemical structure and present
in amounts 100 to 1000 times the measured constituent.
In addition, the test procedure should not be tailored to a particular
wastewater, but should be applicable in all wastes where the pollutants
can occur. This variety of substrates will require extensive separations
and cleanup to achieve specificity in the test method.
Since the test procedure will be used by both the discharger and
the regulatory agencies, it must be within the technical and economic
capabilities of these laboratories. The method must be as simple as
158
-------
possible to minimize the workload, the instrumentation used must be
readily available on the open market at reasonable cost, and there must
be a sufficient number of analysts having the necessary skills and train-
ing to perform these analyses. These constraints are common to all test
procedures promulgated for environmental monitoring and must be considered
in the development and final selection of an analytical method to be used
nationwide. While a test procedure developed in an academic laboratory may
produce satisfactory results in the hands of the research scientist who
developed it, it may not meet the needs for routine monitoring.
The problem for EPA and the task assigned to the Environmental Moni-
toring and Support Laboratory in Cincinnati is to have all these analytical
methods available by June 30, 1983. By that time, under ideal conditions,
the methods will have been thoroughly tested on a large number of actual
wastes samples, subjected to round-robin testing by a group of environ-
mental laboratories, have been widely disseminated for familiarization,
and perhaps adopted by industry groups and the method standardization
organizations as standard procedures.
We fully recognize the magnitude of the task and EPA is committed
to applying its resources to see that methods are available to carry
out the monitoring requirements inherent in the Consent Decree.
CURRENT STATUS OF METHODOLOGY
A number of elements and compounds identified as priority pollutants
are now incorporated in discharge permits. Test procedures for these have
been published in accordance with Section 304(g) of PL 92-500. The most
recent amendments, published in December 1976, include approved methods
for 16 of the 65 pollutants in the Consent Decree. Thus test procedures
for all of the heavy metals and a few of the organics are already avail-
able and published. Methods development research at the Cincinnati labora-
tory has produced procedures for many of the common pesticides, chlorinated
compounds, and volatile halogenated compounds. These methods are widely
used and are being standardized and adopted by EPA, ASTM, and Standard
Methods for the Examination of Water and Wastewater. They will be incor-
porated in the monitoring program as permit conditions require. At the
time of the signing of the Consent Decree, methods had been developed
for 34 of the 65 elements and compounds listed, although not all of the
methods had been thoroughly tested in a variety of wastes and few of them
had been subjected to inter!aboratory study considered necessary for full
documentation of the procedures.
METHODS DEVELOPMENT CONTRACTS
The methods research described has been conducted with limited re-
sources. It is apparent that a crash program will be necessary to meet the
needs of priority pollutant standards. To meet this challenge, EPA is
developing a series of research contracts leading to analytical methods for
these pollutants. A Request for Proposal has been issued calling for as
many as 12 contracts covering 114 organic compounds. Each of the contracts
159
-------
is for test procedures based upon similarity in chemical structure and/or
predicted instrumental response for a group of compounds.
The contract effort is divided into two phases. Phase I requires
the development and evaluation of an interim procedure in the laboratory,
based on response to pure standards of the material in a matrix of simi-
lar compounds. After the successful completion of this task, Phase II
will require testing the method in a minimum of five industrial waste-
waters, covering at least five of the SIC codes specified in the Consent
Decree. In addition, the contractor will determine the stability of these
samples over a seven-day period of holding and prescribe adequate methods
for the preservation of the samples. Where possible gas chromatography is
to be the primary approach and specific columns and detectors are to be
used in the development of the interim methods. At least two dissimilar
chromatographic columns must be used.
The contract period is to include 12 months of laboratory and field
work, with three additional months for report preparation. The final
report must provide complete method descriptions in a standard format,
as well as statements of precision and accuracy obtained on the actual
wastewater samples examined. The minimum detection limit of the method
in the waste samples must also be determined and reported. Contracts
are to be awarded by October, 1977, and the final contract product is
to be a series of test procedures suitable for routine monitoring of
the priority pollutants. The total contract costs are projected to be
between $1 million and $2 million.
Experience indicates that the task of developing these methods is
formidable. Difficulties with separation of the measured constituent
from the interferring matrix can be expected. Instrumental conditions
will need to be modified to provide quantitative results. The detection
limits and working range desired may be difficult to obtain and the
precision and accuracy may be initially unsatisfactory. Additional refine-
ment of the methods will be needed as working experience identifies
problems. Since the final proof of success in methods development is the
widespread use of the procedure, efforts will be made to provide the
methods to interested groups for evaluation prior to any promulgation in
the regulations. The method descriptions and supporting data will be
forwarded to ASTM and Standard Methods for their consideration as stand-
ardized procedures.
In keeping with the policies of EPA, a parallel effort in quality
assurance will be carried out. The Quality Assurance Branch of the
Environmental Monitoring and Technical Support Laboratory routinely
provides reference samples for many of the contaminants included in
water supply and wastewater regulations. Standard reference samples
for the priority pollutants will be developed to the extent that avail-
able resources permit. A series of quality control samples, of known,
stabilized concentration, will be prepared and made available to labora-
tories performing analyses on wastewaters containing the pollutants.
A parallel set of performance samples, in the working range of the methods
but in concentrations unknown to the analyst, will be developed to evaluate
160
-------
performance capabilities of laboratories conducting analyses on priority
pollutants. These samples will be developed through inhouse and contract
research and should be available for a significant number of the compounds
by 1983. Quality control and performance samples for most of the heavy
metals listed in the Consent Decree are already available from the Cincinnati
laboratory.
SUMMARY
The setting of standards for priority pollutants requires a major
effort in establishing adequate test procedures for the measurement of
these substances in environmental samples. The method requirements in-
clude approximately 120 elements and compounds many of which are complex
organic structures. The hazardous substances occur in a variety of wastes
from different industries representing significantly different substrates.
The test methods must be specific for the material in the microgram per
liter range and must be within the technical and economic capabilities
of industrial and governmental laboratories.
EPA is approaching the challenge by means of a series of research
contracts for 12 groups of similar compounds, which will provide test
procedures for 109 organic contaminants. The total contract effort is
expected to cost more than $1.5 million. The preliminary methods should
be available for field testing by 1980 and will be promulgated as regula-
tions by 1983.
BIOGRAPHY
Mr. Ballinger Is Director of the EPA Environmental
Monitoring and Support Laboratories in Cincinnati. Mr.
Bal linger holds a B.S. degree in Chemistry from the
University of Cincinnati, is the author of many publications
in the analytical chemistry of water and wastes. He is a
member of the American Chemical Society Water Pollution
Control Federation, American Water Works Association,
and Research Society of America.
161
-------
DISCUSSION
Leonard Crame, Texaco, Inc.: When the 1983 EPA toxic tests are developed, what
guarantee will we have that the final EPA accepted toxic pollutant test procedures will
be consistent with those tests used to develop guidelines?
Bellinger; Those that have been used to define the presence or concentration?
Crame: What assurance do we have that the numbers that will be calculated for guide-
line purposes will be consistent with what the new test procedures will be?
Bellinger: You'll have absolute assurance that that is the case. In the first place, the
contracts now defining BAT are based upon the qualitative aspect. We don't have the
problem of identification that we will have in the final. They are also contracts that
define the presence of the material and relative concentration. There is no assurance
that the specific test procedures will be the same but given the technical constraints of
both there are only so many approaches to the determination of an organic compound and
they very likely will be consistent, but not exactly. I think, for example, that our
final test procedure for monitoring will be perhaps at lower concentrations but require
cleaner, better separations and more specific. I think there will be a difference; I hope
it is not a significant difference. I don't anticipate any changes in it.
Crame: If it does change anything, will there be a mechanism for changing guidelines?
Ballinger: I don't see how there would be a difference. If for example the standard is
based on the concentration in effluents, that standard was developed irrespective of how
it is measured. So a measurement technique would not change that if both are specific
for the pollutant. It may well be that the analytical method is more sensitive than
required. If the standard is not restrictive, that's great; I hope that the standard is not
below this sensitivity, then we are in trouble. I don't see the problem if these are not
just exactly the same procedures.
Crame; If you use one analytical method now and come up with another in 2 or 3 years
from now, will this new method actually be checked against the old method?
Bal linger.- They will be checked against the same wastes.
Crame; Right, from our own experience with certain refinery wastewaters even the
slightest change in an analytical test will give you significantly higher numbers.
Bal linger; That is true in some measurements, say oil. That's because some measure-
ments are what we call our empirical that the answer you get depends on the method.
That's not the case in general with organic structures based on GC or GC/MS. I think
in one case we are dealing with what amounts to a definition of the result depending on
how you do it. Certainly that's true with what we call oil and grease. I don't think
that same case applies when looking at GC/MS.
162
-------
Ben Buchanan, Phillips Petroleum Company: Will EPA issue or compile the methods
after they are developed?
Ballinger: If you mean by an independent method description, yes, EPA intends to
publish these methods in our format.
Buchanan: They haven't always been the same method as put out by ASTM and APHA in
the past. Are they going to be the same?
Ballinger; We hope so. We will make every effort to get ASTM D19 to adopt the
adopt the methods that we have been gathering supporting data on, but they are totally
independent groups having a number of independent opinions and we have no control
over what D19 does. So it's possible that they will come out with a different method.
Buchanan: The official will be the EPA method?
Ballinger: The official will be the EPA method.
Buchanan: I just wanted to comment that the level of skill required when you get down
to parts per billion hasn't been required in the past and that laboratories that are not
used to operating down in this area may find it more difficult to get close answers and
maybe it might even come to the fact that people with greater skill would have to be
employed to do analysis down in this region.
Ballinger: I don't think there's any doubt that when you lower a detection limit or a
working range you do require a greater efficiency on the part of the operator, better
control over instrument conditions and so on.
Buchanan: In chromatography it's very difficult dealing with interferences as you
probably know, and higher molecular weight compounds are more difficult to separate
so there may be some real problems with these.
Ballinger: There will be some real problems.
Buchanan; Thank you, sir.
Leo Duffy, Standard Oil Company of Indiana; I took by your comments that perhaps you
intend not to use GC/MS in the forthcoming analytical contracts. Is this right?
Ballinger; That is correct. We do not intend to use GC/MS in these contracts if we can
get away with it and just use GC, because of the technical and economic factors. A
good GC/MS as you know will put you in the $150,000-$200,000 bracked We con-
sider that not a good approach to routine analyses. So our contracts specifically call
bor GC and specifically for certain column and detectors.
F. L. Robertaccio, DuPont; This question does not relate to your topic, but I wanted
163
-------
to ask it anyway. What is your personal opinion of the possibility of laboratory certifi-
cation?
Bellinger: There are basically no requirements in 92-500 for certification of
laboratories. EPA cannot require that data come only from certified laboratories
because it is not a part of the legislation. We do feel that the agency has a responsi-
bility for the data that it uses and in decision making therefore anticipate that we will
evaluate the laboratories, not certify; judge their data accordingly. Certifying means
that we go in and inspect and say the results coming from this laboratory are certified
and correct. We do intend to evaluate the laboratories. It is a fine distinction but it
is a legal one.
John Hallett, Shell Oil Co: What is being done to evaluate the performance of
analytical contractors being used in the refining industry priority pollutant screening
and validation surveys?
Bellinger; Right now we do not have the performance samples completely developed for
those contracts, but we are working on it right now.
Hallett: Thank you.
Bal linger: However, that is simply an evaluation of EPA's contracts.
164
-------
METHODS DEVELOPMENT CONTRACTS
LISTED POLLUTANT NUMBER OF COMPOUNDS IN GROUP
Phthalate Esters 6
Haloethers 7
Chlorinated Hydrocarbons 9
Nitrobenzenes and Isophorone 4
Nitrososamines 3
Dioxin 1
Benzidine 3
Phenols 11
Polynuclear Aromatics 16
Pesticides and PCBs 25
Purgeables (Volatiles) 26
Acrolein, Acrylonitrile, Dichlorofluoromethane 3
165
-------
"FATES, EFFECTS & TRANSPORT MECHANISMS
OF POLLUTANTS IN THE AQUATIC ENVIRONMENT"
Fred T. Weiss
Shell Development Company
Houston, Texas
ABSTRACT
Evaluation of environmental effects from exposure to low concentrations
of inorganic or organic compounds in the aquatic environment must include a
concern for the reactivities and fate of these compounds under natural
conditions. At the present time it is not possible to model the detailed
fate of many compounds due to lack of accurate information on chemical
reactivity and physical transport. However, during the past decade much
information has been obtained outlining the general chemical and physical
behavior of certain pollutants. Consequently it is possible to make genera-
lized predictions of the fate and extent of persistence of a number of
compounds and to point out areas of research to obtain more complete chemical
and physical data.
INTRODUCTION
The intent of this report will be to review, in summary form, the fate,
transport mechanisms and effects of the important classifications of pollu-
tants of concern for petroleum refinery wastewater. These can be listed
broadly as Hydrocarbons, Organics and Metals as indicated in Figure I. For
this discussion organics are considered as other than Hydrocarbons. Since
Dr. Mount will discuss toxicity testing in the paper immediately following I
do not plan to elaborate on the topic of toxicity. For each of the principal
classifications a great deal of environmental information has been obtained
during the last decade and reported in many diverse publications. However, in
only a limited number of these studies are any quantitative results available.
Consequently, the information obtained generally provides suggested pathways
and recommends research to establish more completely the mechanisms and rate
of the processes which exist for the ultimate removal or tranformation to
other species which may represent the end product.
HYDROCARBONS
Since the Forum has been called to review wastewaters from petroleum
refineries it is appropriate to deal first with the hydrocarbons present as
pollutants in aquatic systems. Because of the overall significance much work
has been done and reported on hydrocarbon properties and behavior. One report
to which reference will be made is from the National Academy of Sciences
entitled "Petroleum in the Marine Environment".21 Another is from the American
Institute of Biological Sciences and is the Proceedings of the 1976 Symposium
on "Sources, Effects and Sinks of Hydrocarbons in the Aquatic Environment".2
166
-------
A study of high degree of significance to our discussion is that funded
by the National Science Foundation - Research Applied to National Needs (RANN)
Program on the "Petroleum Industry in the Delaware Estuary" which is currently
active.24 This project is staffed with scientists from Rutgers University and
the Philadelphia Academy of Natural Sciences and is engaged in investigating
the relative effects of the petroleum industry and other sources in the
Delaware Estuary.
The general picture of fates of hydrocarbons in the aquatic system was
schematically illustrated by W. D. Garrett1 in his 1972 publication on the
impact of surface films on the properties of the air-sea interface. Figure II
taken from this publication is an illustration of the forces which modify
hydrocarbon oil slicks on water. Figure II shows, in simplified form, that
the forces which operate on oil in water include evaporation, solution,
emulsion formation, sedimentation, oxidation, biological assimilation and
others. It is important to note that natural, biogenic hydrocarbons, with
structures similar to petroleum hydrocarbons are often present in natural
systems.21 The analytical differentiation between biogenic and petroleum
hydrocarbons takes considerable effort in many cases.
Volatilization of the lower molecular weight hydrocarbons takes place
quite rapidly. This is important because of the toxicity of some of the
volatile fractions which are lost soon. Simulated weathering has shown that a
crude oil on water loses essentially all components boiling below Cj2 within
24 hours in the laboratory.3 It has been observed that the toxicity of a
crude oil mixed with water was greatly reduced even by short weathering.1*
Solution. Evaporation and solution both act most strongly on the low-
molecular weight hydrocarbons. The true solubility of hydrocarbons drops
exponentially as a function of their molecular volume.5 Consequently, although
toluene is soluble in water to 515 ppm, anthracene is soluble to less than
0.1 ppm. The higher polynuclear aromatics are considerably less soluble.
Oxidation. This is a selective process and can be subdivided into chem-
ical oxidation and photooxidation. Absorption of photons by the heavier
aromatics with suitable spectral properties will initiate the reaction chain
leading to conversion of these materials. Products of the reaction will
contain oxygen and-could consist of phenols, carboxylic acids, alcohols and
the like. Since these are more water soluble than the hydrocarbons, they will
be more completely lost to the water column leading to greater dispersion.
Microbial degradation. This subject has been studied over many years by
many competent investigators. They conclude7'8»9»10 that microorganisms which
oxidize various hydrocarbons are widely distributed in soil and water, espe-
cially in estuaries and shorelines. Although normal paraffins are most
susceptible, virtually all hydrocarbons are degraded including aromatic
hydrocarbons of condensed structures. These bacteria, which use hydrocarbons
for an energy source, are effective in converting hydrocarbons to carbon
dioxide and water. Probably one of the major ways in which hydrocarbons are
removed in estuaries and coastal areas is by microbiological degradation. For
example, Zobell7 reported that in coastal areas bacteria can oxidize from 0.02
to 2 grams of hydrocarbon per square meter per day depending on several factors
167
-------
including rate of oxygen diffusion. In shallow waters this approaches conver-
sion rates on the order of a part per million of hydrocarbon per day. It is
important to realize that bacteria are active even in very cold arctic
water.1 1 >12
Biochemical uptake metabolisms and discharge of hydrocarbons. It has
been shown that marine organisms, maintained in water containing oil, take up
hydrocarbons in their tissues. Following a spill of No. 2 fuel oil near West
Falmouth, Massachusetts in September, 1969, Blumer13'llf»15 analyzed oysters,
scallops, and other marine organisms and found that they had taken up oil
fractions. He kept three oysters in flowing sea water in his laboratory. One
oyster was analyzed after it had been kept in flowing sea water for 72 days
and the other two after 180 days. His publication in 1970ltf stated that none
of these three oysters had purged themselves of the oil they contained prior
to the beginning of the experiment. He concluded "Thus, once contaminated,
shellfish cannot cleanse themselves of oil pollution".14
A number of scientists have reinvestigated this matter and have shown
that marine animals do indeed take up hydrocarbons but actually do cleanse
themselves. Experiments have been done with, literally, hundreds of aquatic
animals. Data now in hand clearly show that aquatic animals do take up hydro-
carbons but that, when the animals are placed in clean water, the hydrocarbons
are purged or metabolized. For example, Lee and his co-workers16'17 described
the uptake, metabolism and discharge of radio-labeled aromatic hydrocarbons by
mussels and by fish. They found that these compounds did indeed find their
way from sea water into the aquatic animal tissues, but that when the animals
were placed in clean water, the hydrocarbons were lost. The mussels purged
the hydrocarbons unchanged whereas the fish metabolized these products. In
another study, Anderson18 carried out exposure tests on oysters and clams,
illustrated in Figure III taken from his work. In these tests, oysters were
initially exposed to a highly aromatic fuel oil. The animals were then placed
in clean sea water. Specific analyses for individual hydrocarbons were made
and purging was found to the level of analytical sensitivity. Very similar
data have now been found in studies at other locations in the United States
including Battelle-Northwest in the State of Washington19 and more recent data
from Woods Hole20 using, in each case, local animals. In every case examined,
purging was found to occur.
The National Academy of Science Report21 states: "organisms such as
mussels and oysters have been shown to eliminate most absorbed petroleum
hydrocarbons when placed in clean water".
Accumulation in the food web. The additional question of accumulation of
petroleum hydrocarbons in the food chain should be considered. Lee's
results16*17 show a rapid metabolism of certain aromatic hydrocarbons in
fish. Data from Anderson and others show relatively rapid purging of hydro-
carbons from animals placed in a clean environment. Once the animal is no
longer subjected to water contaminated with oil, the affected organism cleanses
itself quickly of whatever oil contamination that it may have incurred.
Therefore, it is not likely that such contamination would become concentrated
by transfer from one trophic level to the next through the food chain. In
fact, the National Academy Report21 states "There is no evidence for food-web
magnification of petroleum hydrocarbons in marine organisms."
168
-------
Significance of Field Studies. Field studies have been conducted princi-
pally in regions where oil and gas production has taken place.^5,^6 xhese have
all shown that neither oil spills nor continued production have detrimental
effects on the environment. Grossling of the U. S. Geological Survey reported4**
that, even in areas of extremely massive oil spills of some years ago, there
are "no Death Valleys". Natural forces which degrade oil have taken over and
converted the spilled oil into the end products of nature, carbon dioxide and
water. Reviews of current producing areas have shown no detrimental
effects.1*5'46 Detailed chemical and biological surveys of producing area and
a control area in the Gulf of Mexico shows that production of petroleum had no
discernable effect on the environment.1*5 The underwater surveys1*6 of Platforms
"Hilda" and "Hazel" in the Santa Barbara Channel have demonstrated much
increased marine biota in and around the platforms. Studies of the petroleum
industry in the Delaware Estuary24'1*7 have shown that the city of Philadelphia
and its environs contribute much more "oil and grease" than does the refining
industry in the area. Contributions of "oil and grease" from the metropolitan
area are larger from drains and storm-water runoff than from the refining
sources.
POLYNUCLEAR AROMATICS HYDROCARBONS (PAH)
Because of the concern for the presence of some carcinogenic polynuclear
aromatic hydrocarbons (PAH) in the aquatic environmental this subject is
worthy of separate discussion. It must be pointed out that only a limited
number of PAHs are carcinogenic. Those that are carcinogenic will contain 4,
5 or 6 rings. Only certain isomers are carcinogenic; for instance benzo(a)-
pyrene is carcinogenic but not its isomers.1*2 Considerable data are now
available from a number of sources21'23 which show that polynuclear aromatics
are widely distributed in soils at very low concentrations and may have
occurred on the earth's surface during geologic time. Presumably a source of
PAHs has been from combustion such as forest fires or in more recent times,
from burning of coal.
Microbial degradation. Biological degradation of polynuclear hydrocarbons
has been discussed by Professor Gibson at the Symposium on Sources, Effects &
Sinks of Hydrocarbons in the Aquatic Environment.2 He pointed out that, since
aromatic hydrocarbons have been ubiquitously distributed throughout the
environment over geologic time, living organisms have evolved enzyme systems
which oxidize these compounds. Table I, taken from his publication25 shows the
types of aromatic hydrocarbons known to be susceptible to microbial oxidation.
One common feature of the mechanisms used by bacteria to degrade these types
of compounds is the introduction of hydroxyl groups. Generally, two hydroxyl
groups are introduced and the diol so produced is then subjected to further
degradation. A metabolic pathway for the complete degradation of naphthalene
is indicated in Figure IV. The compounds shown in brackets have been completely
identified. The end products of this mechanism are the non-toxic products,
carbon dioxide and water.
Less is known about the detailed chemistry of microbial degradation of
PAHs that contain three or more rings. However, evidence is being obtained
that bacteria are capable of oxidizing aromatic hydrocarbons that range in
size from benzene to benzo(a)pyrene. A mutant strain of bacteria, faexj&Unefex.a
169
-------
B836 has been shown to oxidize biphenyl,26 phenanthrene,27 and anthracene28
to dihydroxydiols. Benzo(a)pyrene was oxidized by this bacteria to dihydroxy
product as shown in Figure V, as well as to other products resulting from
further degradation.29
Biochemical uptake and discharge. Because of the specific concern for
the PAHs, investigators have repeated uptake and purging studies with indi-
vidual PAHs and find results identical to that found with other hydrocarbons.
That is, aquatic animals will take up PAHs in their gut and their tissues, and
that these are purged or metabolized when the animals are placed in clean
water. Recent work of this nature has been conducted by Neff39 and Neff and
Anderson.40 The conclusions of these studies were that the pattern of uptake
and depuration of benzo(a)pyrene in the clam is much the same as observed
earlier from the naphthalenes. Benzo(a)pyrene appears to be accumulated more
rapidly, however, and released more slowly than the naphthalenes. In their
experiments purging was a consistent process, declining to a level of only 1.4
percent of the accumulated material in 20 days. Benzo(a)pyrene could not be
detected in clams maintained in isotope-free sea water for 58 days (limit of
detection 0.01 ppm). It is interesting that the viscera contained most of the
activity at all sampling periods.
Photochemical oxidation. Since the polynuclear aromatic hydrocarbons
have high absorptivities in the ultraviolet, it is likely that photochemical
oxidation plays an appreciable role in their degradation. In fact a review
states that photooxidation is probably one of the most important processes in
removal of polycyclic hydrocarbons from the atmosphere. **2
ORGANICS
Many of the same factors which operate to dispose or degrade hydrocarbons
are also effective on many organics, depending of course on properties of the
individual compounds. Such factors as volatilization, solution, sedimentation
and oxidation are likely candidates for removal or transformation and should
be studied. It is well known that ethers are easily oxidized and it is likely
that this is the mechanism of their loss in a natural system. Oxidation is
also possibly the route for the loss of other relatively reactive organic
compounds. Because of the reactivity of organic functional groups other
reactions such as hydrolysis, reduction and dechlorination may be important.48
Photochemical. Photochemical energy reaches us from the sun in the form
of photons with wave lengths covering the spectra from infrared to the far
ultraviolet and including the visible. In fact, all the solar energy we
receive is in the cumulative energy of the photons which fall upon the earth.
Photosynthesis is our most important photochemical process, without which
there could be no life on earth since plant growth depends upon the photo-
chemical conversion of atmospheric carbon dioxide into organic substances.
The significance of photochemical processes is becoming more and more apparent
in the environmental chemistry of organic materials. The photochemical
oxidation and dechlorination of a number of the highly chlorinated pesticides
and similar compounds has been well established. A particularly significant
recent paper appeared in "Science" last March indicating that the toxic,
complex substance TCDD (2,3,7,8 tetrachlorodibenzo-p-dioxin) is photochemically
170
-------
degraded by sunlight6 in the natural environment. Presumably the degradation
reactions involve dechlorination and perhaps oxidation. Photo sensitived
oxidations may also play a significant role in the observed degradation of DDT
in the environment30 as well as with other pesticides.l+8
Microbial degradation. The microbiological metabolism of organics has
been the subject of detailed study for a number of years and has been exten-
sively reported.31'32 Investigations similar to those outlined above for
polynuclear aromatics have been reported on phenols (Dagley, University of
Minnesota),32 on surface-active agents (Huddleston and Allred, Continental Oil
Co.),32 on insecticides (Matsumura and Boush, University of Wisconsin)32 and
on other classes of commercial organic compounds that could be present in
waste waters. A number of bacteria widely distributed in nature, have the
ability to degrade phenolic compounds. The ability to degrade phenolics is
not confined to bacteria. These reactions take place also with fungi and in
species of (UpQAg^Wu, pWlicJJUIMm and OO&poHa, The end products of these
conversions are simple, essentially non-toxic organic structures.
Organic halogen compounds have been shown to undergo a number of micro-
biological reactions which result in loss of halogen and its replacement
either by hydroxyl or hydrogen. Microbiological hydrolytic dehalogenation is
the most common reaction. These reactions seem quite general in natural
soils. Some examples which have been documented119 include lindane which
undergoes microbial decomposition under anaerobic conditions to release
chloride ion and the brominated derivatives of ethane, propane and butane
which produce either the simple hydrocarbons or alcohols upon loss of halogen.
Aliphatic halides such as allyl chloride and 1,3-dichloropropene hydrolyze
readily in the soil.
Edwards50 points out that insecticides in soils are largely broken down
by microorganisms. Some of molds indicated above to be able to utilize
phenolics are also effective in converting chlordane and heptachlor into
hydrophilic degradation products. Several soil bacterial species have been
found able to dechlorinate lindane.
Biochemical uptake and metabolism. It is well known that aquatic species
concentrate organochlorine insecticides33 but is is perhaps not as well
recognized that there is also an appreciable amount of degradation by the
living species. Johnson, Saunders, Sanders and Campbell33 used radiochemicall}
labeled aldrin and DDT in studies with freshwater invertibrates and found, in
3-day exposures, that some degree of degradation of aldrin and DDT occurred
with all organisms examined in the limited time of the tests.
It is interesting that natural products containing bound halogen are
produced by aquatic biota. Dr. Faulkner of Scripps presented a paper at the
Symposium on Sources, Effects & Sinks of Hydrocarbons in the Aquatic
Environment2 which described a group of halogen-containing terpenes biogeni-
cally synthesized by aquatic organisms.
Significance of field observations. There are many natural reactions
Which degrade organic compounds, including organic pesticides, which have now
been documented from field and laboratory observations. As a result of these
171
-------
reactions it has been stated that "there are no permanent organic pesticides".1*8
Degradation apparently continues onto the non-toxic inorganic products with
rates depending upon many factors, such as climate, temperatures, soil type
and so on.
METALS
To put a perspective on the effects of metal ions in the aquatic environ-
ment it is useful to turn to an important concept introduced by Dr. Ketchum of
Woods Hole. This is the Relative Critical Index as described in Table II,
This Table lists toxic elements considered to be of critical importance to
aquatic pollution. The "toxicities" (concentrations in pg/1) assigned to the
metals in the table are in a decreasing order of toxicity, and each "toxicity"
concentration quoted is considerably smaller than the concentration acutely
toxic to man or the aquatic environment. These toxicity values are obtained
from "Water Quality Criteria".41
The concept of Relative Critical Index in Table II is derived by Ketchum3"*
by dividing the annual amount of a substance mobilized by human or natural
activities by the assumed "toxicity". This index helps to identify elements
as pollutants and assists in establishing the relative contributions of trace
elements from nature and from man-made sources. For instance, it can be seen
that the greatest input of most trace metals listed is from natural sources.
Where the natural contribution is a large portion of the potential supply,
control of individual sources whose concentration levels are low would have
little effect on water quality.
Chelation. 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 most natural waters much of the free
metals ions would probably be bound to organic substances naturally present in
the water. There is 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.35'36 This could be due either to the fact that the organo-
metallic complex is to bulky to enter a biological system or it could be due
to the lack of availability of the metal ion for reaction with enzymes within
the cells.
Reduction. Methylation of mercury in the aquatic environment can be
caused by the reductive action of anaerobic bacteria at the bottom-water
interface.51 These bacteria produce the highly toxic dimethylmercury (CHs^Hg
and methylmercury CH3Hg+. It was this type of mercury pollution which was the
cause of the notorious "Minamata disease" in Japan in 1953.52
Sedimentation. Detailed metal analysis of sediments and benthic marine
populations have been conducted near sewage outfalls in the Pacific Ocean off
the Southern California Coast.37 No significant differences were observed for
sole caught in polluted and unpolluted areas for the following trace elements,
mercury, cadmium, copper, zinc, iron and cobalt in the sediments. In the
reported discussion which followed the above presentation37 it was stated that
similar results were found in England. The suggestion was made that sediments
may be a useful final sink for many metals and that the adsorption of the
172
-------
metal ions to the surface of sediments precludes their uptake by biological
systems. I would suggest that an important research study would be that of
evaluating the biological availability or nonavailability of sediment-associated
metals in the aquatic environment.
Essential nutrients. Almost any element or substance can be toxic to
plants when present in abnormally high concentrations. Yet many toxic elements
have been shown to be essential nutrients at low concentrations, at least to
certain plant species.38'43 Drake of the University of Massachusetts lists1*3
as essential micronutrients the following: iron, manganese, copper, zinc,
boron, molybdenum and chlorine. It has been recognized especially by orchard
growers that essential elements must be introduced when soils are deficient.
Consequently "zinc nails" are sometimes driven into trees when soils are
deficient. Enzyme system in plants seem to require manganese, iron, and
molybdenum as well as magnesium in trace amounts. There is apparently some
evidence that selenium may be a necessary trace nutrient to some plants.38
The assistance of Dr. Paul Porter of the Shell Biological Sciences
Research Center, Modesto, California is gratefully acknowledged in the
preparation of several sections of this paper.
BIOGRAPHY
Fred T. Weiss is Senior Staff Research
Chemist, Analytical Department, Shell Develop-
ment Company, Houston, Texas. He holds
degrees in Chemistry from the 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 applica-
tion of methods to trace analysis. In recent
years much of his work has been devoted to en-
vironmental measurements.
173
-------
REFERENCES
1. Garrett, W. D. (1972), "Impact of Natural and Man-made Surface Films on
Properties of the Air-Sea Surface" in "The Changing Chemistry of the
Oceans," (D. Dryssen and D. Jagner, Eds.)» Wiley Interscience, pp. 75-91,
New York.
2. American Institute of Biological Sciences (1976), "Sources, Effects &
Sinks of Hydrocarbons in the Aqutic Environment", 1401 Wilson Boulevard,
Arlington, Virginia 22209.
3. Kreider, R. E. (1971), "Identification of Oil Leaks and Spills,"
Proceedings of Joint Conference on Prevention and Control of Oil Spills,"
pp. 119-124, Washington, D. C.
4. Templeton, W. L. (1974), "Summary Report on Effects of Oil Discharges,
Domestic and Industrial Wastewaters on the Fisheries of Lake Maracaibo,
Venezuela," Battelle-Northwest Laboratory, Richland, Washington.
5. McAuliffe, C. D. (1966), "Solubility in Water of Paraffin, Cycloparaffin,
Olefin, Acetylene, Cyclo-Olefin and Aromatic Hydrocarbons," J. Phys.
Chem., 70, 1267.
6. Crosby, D. G. and Wong, A. S., (1977) "Environmental Degradation of 2, 3,
7, 8 Tetrachlorodibenzo-p-dioxin (TCDD)", Science, 195, 1337.
7. Zobell, C. E. (1969), "Microbial Modification of Crude Oil in the Sea,"
Proceedings of Joint Conference on Prevention and Control of Oil Spills
pp. 317-326, New York.
8. Colwell, R. R., Walker, J. D., and Nelson, J. D., Jr., (1973) "Microbial
Ecology and the Problem of Petroleum Degradation in Chesapeake Bay," in
Microbial Degradation of Oil Pollutants, Louisiana State University,
Publication. No. LSU-SF-73-01.
9. Miget, R. J., Oppenheimer, C. H., Kator, H. I., and LaRock, P. A. (1969),
"Microbial Degradation of Normal Paraffin Hydrocarbons in Crude Oil,"
Proceedings of Joint Conference on Prevention and Control of Oil Spills,
pp. 327-331, New York.
10. Walker, J. B., Colwell, R. R., and Petrakis, L. (1975) "A Study of the
Biodegradation of Louisiana Crude Oil Employing Computerized Mass Spectro-
metry," Proceedings of Joint Conference on Prevention and Control of Oil
Pollution, San Francisco, California, pp. 601-605.
11. Kinney, P. J., Button, D. K., and Schell, D. M. (1969), "Kinetics of
Dissipation and Biodegradation of Crude Oil in Alaska's Cook Inlet,
"Proceedings of Joint Conference on Prevention and Control of Oil Spills",
pp. 333-340, New York.
174
-------
12. Robertson, B., Arhelger, S., Kinney, P. J., and Button D K (1973)
"Hydrocarbon Biodegradation in Alaska Waters," In microbial Degradation
of Oil Pollutions, Louisiana State University, NO. LSU-SG-73-01, pp.
171—184.
13. Blumer, M., Souza, G., Sass, J. (1970), "Hydrocarbon Pollution of Edible-
Shellfish by an Oil Spill," Marine Biology. _5, pp. 195-202.
14. Blumer, M., Sass, J., Souza, G., Sanders, H. L., Grassle, J. F. and
Hampson, G. R., (1970), "The West Falmouth Oil Spill," Woods Hole
Oceanographic Institution, Woods Hole, Massachusetts.
15. Blumer, M., and Sass, J. (1972), "West Falmouth Oil Spill, Data Available
in November 1971, II, Chemistry," WHOI 72-19, Woods Hole Oceanographic
Institution, Woods Hole, Massachusetts.
16. Lee, R. F., Sauerheber, R., and Benson, A. A. (1972), "Petroleum Hydro-
carbons: Uptake and Discharge by Marine Mussel Mytilus edulis," Science.
177. pp. 344-346.
17. Lee, R. F., Sauerheber, R., and Dobbs, G. H. (1972), "Uptake, Metabolism
and Discharge of Polycyclic Aromatic Hydrocarbons by Marine Fish," Marine
Biology, 177. pp. 344-346.
18. Anderson, J. W., Neff, J. M., Cox, B. A., Tatem, H. E., and Hightower,
G. M., (1974), "The Effects of Oil on Estuarine Animals: Toxicity,
Uptake and Depurations, Respiration," in symposium: Pollution and the
Physiological Ecology and Estuarine and Coastal Water Organisms, sponsored
by the Environmental Protection Agency and the Belle W. Baruch Coastal
Research Institute of the University of South Carolina, November 14-17,
1973.
19. Vaughan, B. E. (1973), "Effects of Oil and Chemically Dispersed Oil on
Selected Marine Biota - Laboratory Study," Battelle-Northwest Laboratories,
API Publication, No. 4191.
20. Teal, J. M., and Stegeman, J. J. (1973), "Accumulation, Release and
Retention of Petroleum Hydrocarbons by the Oyster, CtlO&AOA&Lea (AtAg.6u.ca,"
Marine Biology, 22, pp. 37-44.
21. National Academy of Sciences (1975), Ocean Affairs Board Workshop,
"Petroleum in the Marine Environment" Washington, D. C.
22. Pancirov, R. J. and Brown, R. A. (1975), "Analytical Methods for Poly-
nuclear Aromatic Hydrocarbons in Crude Oils, Heating Oils and Marine
Tissues," Proceedings of Joint Conference on Prevention and Control of
Oil Pollution, San Francisco, pp. 103-113.
23. Blumer, M. and Youngblood, W. W. (1975), "Polycyclic Aromatic Hydrocarbons
in Soils and Recent Sediments," Science, pp. 53-55. See also Blumer,
Science 134, 474 (1961).
175
-------
24. "Petroleum Industry in the Delaware Estuary" (1977) National Science
Foundation - RANN Program Grant Env. 74-14810 A03.
25. Gibson, D. T. (1976), "Microbial Degradation of Carcinogenic Hydrocarbons
and Related Compounds," presented at "Symposium on Sources, Effects and
Sinks of Hydrocarbons in the Aquatic Environment", American University
Washington, D. C., August 9-11, pages 224-238 (See Reference 2).
26. Gibson, D. T., Roberts, R. L., Wells, M. C. and Kobal, V. M. (1973),
"Oxidation of Biphenyl by a B/ce/eAoicfe/ta Species." Biochem. Biophys.
Res. Commun. 50, page 211.
27. Selander, H., Yagi, H., Jerina, D. M., Wells, M.C., Davey, J. F.,
Mahadevan, V., and Gibson, D. T. (1976), "Dihydrodiols from Anthracene
and Phenanthrene," J. Am. Chem. Soc.
28. Akhtar, M. N., Boyd, D. R., Thompson, N. J., Gibson, D. T., Mahadevan, V.,
and Jerina, D. M. (1975), "Absolute Sterochemistry of the Dihydroanthracene-
cci- and ;t/iaxi6-l,2-diols Produced from Anthracene by Mammals and Bacteria,"
J. Chem. Soc., 2506.
29. Gibson, D. T., Mahadevan, V., Jerina, D. M., Yagi, H., and Yeh, H. J. C.
(1975), "Oxidation of the Carcinogens Benzo(a)pyrene and Benzo(a)anthracene
to Dihydrodiols by a Bacterium," Science, 189, page 295.
30. Ivie, G. W., Casida, J. E. (1971), "Photosensitivers of the Accelerated
Degradation of Chlorinated Cyclodienes and other Insecticide Chemicals
Exposed to Sunlight on Bean Leaves", J. Agr. Food Chem., 19, 410-416.
31. Goring, C. A. I. and Hamaker, J. (Editors) (1972) "Organic Chemicals in
the Soil" Volumes I & II, Marcer Dekker, New York.
32. McLaren, A. D. and Skujins, J., (Editors) (1971) "Soil Biochemistry"
Marcer Dekker, New York.
33. Johnson, B. T., Saunders, C. R., Saunders, H. 0., and Campbell, R. S.
(1971) "Biological Magnification and Degradation of DDT and Aldrin by
Freshwater Invertebrates," J. Fish, Res. Bd. Can. 28: 705-709.
34. Ketchum, B. H. (1973), "Symposium on Ocean Pollution," Statement made
before Senate Commerce Committee, Subcommittee on Oceans and Atmosphere,
June 12.
35. Lerman, A. and Childs, C. W. (1973), "Metal-Organic Complexes in Natural
Waters: Control of Distribution by Thermodynamic, Kinetic and Physical
Factors," In "Trace Metals and Metal-Organic Interactions in Natural
Waters." Edited by Singer, P. C., Ann Arbor Publ., Inc., Ann Arbor,
Michigan.
176
-------
36. 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.
37. DeGoeij, J. J. M., Guinn, V. P., Young, D. R. and Mearns, A. J. (1974)
"Neutron Activation Analysis Trace Element Studies of Diver Sole Liver
and Marine Sediments," IAEA-SM-175/15, pages 189-200 (International
• Atomic Energy Agency, Vienna).
38. "Soil-The Yearbook of Agriculture" (1957) - U.S. Department of Agricul-
ture, pages 165-171.
39. Neff, J. M. (1975), "Accumulation and Release of Petroleum - Derived
Aromatic Hydrocarbons by Marine Animals," in Symposium on Chemistry,
Occurrence and Measurement of Polynuclear Aromatic Hydrocarbons, presented
at Division of Petroleum Chemistry, American Chemical Society, Chicago,
August 24-29.
40. Neff, J. M. and Anderson, J. W. (1975), "Accumulation, Release and
Distribution of Benzo(a)pyrene C~llt in The Clam Rang/to. Cunea&l," Joint
Conference on Prevention and Control of Oil Pollution, San Francisco,
pages 469-471.
41. National Academy of Sciences (1974), "Water Quality Criteria," U.S.
Government Printing Office, Washington, D. C.
42. National Academy of Sciences (1972), "Particulate Polycyclic Organic
Matter," Washington, D. C.
43. Drake, M. (1968) "Soil Chemistry and Plant Nutrition" in "Chemistry of
the Soil" Second Edition, Editor F. E. Bear, Reinhold Publishing Company,
New York, pages 398-400.
44. Reference 2, page 36.
45. El-Sayed, S. Z. (1974), "Effects of Oil Production on The Ecology of
Phytoplankton off the Louisiana Coast," Project OV-66-JHM, Gulf
Universities Research Consortium - Offshore Ecology Investigations.
46. Bascom, W., Mearns, A. J. and Moore, M. D. (1976) "A Biological Survey
of Oil Platforms in the Santa Barbara Channel". J. Petroleum Technology,
November 1976, pages 1280-1284.
47. Hunter, J. V., Yu, S. L. and Whipple, W., Jr. (1976), "Measurement of
Urban Runoff Petroleum in "Urbanization and Water Quality Control",
Published by American Water Resources Associates, Minneapolis, Minnesota.
48. Crosby, D. G. (1973) "The Fate of Pesticides in the Environment" in
Annual Reviews of Plant Physiology 24, 467-492.
177
-------
49. Meikle, R. W. (1972) "Decomposition: Qualitative Reactions", in Reference
31, pages 203-210.
50. Edwards, C. A. (1972) "Insecticides", in Reference 31, pages 531-537.
51. Wood, J. M., Kennedy, F. S. and Rosen, C. G. (1968), "Synthesis of Methyl-
mercury Compounds on Extracts of a Methanogenic Bacterium", Nature, 220,
pages 173-174.
52. Dugan, P. R., (1972) "Biochemical Ecology of Water Pollution", Plenum
Press, New York.
178
-------
Figure I
SCHEMATIC REVIEW OF PRINCIPAL FATES, EFFECTS AND TRANSPORT
MECHANISMS OF POLLUTANTS IN THE AQUATIC ENVIRONMENT
Hydrocarbons
Organics
Metals
Fates
Volatilization
Solution
Sedimentation
Oxidation
Volatilization
Solution
Sedimentation
Oxidation
Photochemical Dechlorination
Chelation
Oxidation/Reduction
Sedimentation
Transport
Mechanisms
Air-Water Interface
Sediment Transport
Air-Water Interface
Sediment Transport
Sediment
Transport
Effects
Oxygen Consumption
Biological
Assimilation
Toxicity
Oxygen Consumption
Biological
Assimilation
Toxicity
Biological
Assimilation
Essential Trace
Elements
Toxicity
-------
BURSTING
BUBBLES
ATMOSPHERIC
OXIDATION
RAIN AND
FALLOUT
SEA SURFACE
OIL SLICK
V/ATER-1 N -01L EMULSION
1 I .ss
OIL-IN-WATER //
EMULSION If
DISSOLUTION
\\
BUBBLED
TRANSPORT
Iff
SPREADING
CONVECTION
AND
UPWELLING
NONBUOYANT
OIL RESIDUES
MIXING AND
SINKING
CIRCULATIONS
ADSORPTION
ON NONBUOYANT
PARTICLES
BUOYANT
MATERIAL
BIOLOGICAL
ASSIMILATION
BOTTOM SEDIMENT
Fig. II - Natural Force* Which Diapers* and Modify Oil Slick* on Watar Aft«r
Garrett,1 180
-------
THE ACCUMULATION AND DISCHAPCTrv*
NO. 2 FUEL OIL COMPONENTS
BY OYSTERS AND CLAMS
10 15 20
TIME OF EXPERIMENT - DAYS
Figure III
C02+H20
I
t
COOH
rr°
CH2 COOH
COOH
CHO
COOH
Figure IV Metabolic Pathway for the Degradation of Naphthalene
by Certain ?*e.udomon0U> Species. From Gibson
181
-------
9- Hydro* ybenjo [o] pyrene
NONENZYMATIC
T
02
Beijerinckio wild lype
ACID
PRODUCTS
C/j-9,10- Dirtydro*y-9.IO-
dihydrobenzo [o]pyrene
Figure V Oxidation of Benzo (a)pyrene by B&ij &tu.nc.ksia B836.
From Gibson25
Table I
AROMATIC HYDROCARBONS KNOWN TO BE OXIDIZED BY MICROORGANISMS
a)
NONOCYCLIC
Benzene
Toluene
Xylenes
Tri and tetramethylbenzenes
Alkylbenzenes
Cycloalkylbenzenes
DICYCLIC
Naphthalene
Methylnaphthalenes (mono and di)
Ethylnaphthalenes
POLYCYCLIC
Pyrene
Benzo(a)pyrene
Benzo(a)anthracene
Dibenzo(a)anthracene
Benzperylene
Perylene
TRICYCLIC
Phenanthrene
Anthracene
a)From Gibson25
182
-------
Table II
METAL TOXICITY AND RELATIONSHIP TO INPUT
34
ELEMENT
MERCURY
CADMIUM
SILVER
NICKEL
SELENIUM
LEAD
COPPER
CHROMIUM
ARSENIC
ZINC
MANGANESE
TOXICITY
yg/i
0.1
0.2
1
2
5
10
10
10
10
20
20
RATE OF MOBILIZATION (109 G/YR)
A (MAN)
1.6
0.350
0.07
3.7
0.45
3.6
2.1
1.5
0.7
7
7.0
B (NATURAL)
2.5
2.65
11
160
7.2
110
250
200
72
720
250
C
TOTAL
4.1
3.0
11.1
164
7.7
113.6
252.1
201.5
72.7
727
257
RELATIVE CRITICAL INDEX,*
1012 1/YR
MAN
16,000
1,750
70
1,350
90
360
210
150
70
330
350
NATURE
25,000
13,250
11,000
80,000
1,440
11,000
25,000
20,000
7,200
36,000
12,500
00
*RELATIVE CRITICAL INDEX =
INPUT
TOXICITY
-------
WEISS PAPER DISCUSSION
Paul Mikolaj, Lion Oil Co^; Do you have any Idea of how the information you
have presented will be used in the setting of the standards for the priority
of pollutants?
Weiss; I would hope that the answer is yes. If anybody cares to answer that
from the floor I would like to hear an answer.
Ridgeway Hall, EPA: I think without a doubt if information of that type is
timely submitted to EPA in the course of the rule-making proceedings, we
certainly will consider it.
Ridgeway Hall, EPA; On your chart of Table II which you last had up there,
could you tell us what the significance of the right-hand column was and how
those numbers were derived and what they mean?
Weiss; That is actually the relative critical index itself. These columns
come from taking the toxicity and dividing the toxicity into the rate of
mobilization to develop a ratio. What it means is that the ratio indicates
the relative significance of the element to toxicity.
D. I. Mount, EPA, Duluth; In regard to Table II that you have on the board
I think it is a useful table but I think it is also important to point out
that it makes no account of the form in which those metals are being trans-
ported or converted by man and by nature. For example, in the case of mercury
most of the natural mercury is transported in the mineral form or in sulfide
forms that are insoluble. In the case of much of the man-transported mercury,
it is in vapor form in power plant stacks so that the biological significance
in a given amount of that metal will make it quite different, even the total
quantities show a different picture.
George J. Putnicki, UTD; Are there any taste and odor studies conducted
concurrently with the concentrations of the number of two fuel oil?
Weiss; We did not conduct any taste or odor studies. There are independent
studies which have been done on taste and they show depuration. The oyster
farmers in Louisiana find it is about the same order of time, three or four
weeks. When they observe contaminated oysters, they put them into a clean
bed, so the timing is the same but they were not done concurrently.
184
-------
MEASURING AQUATIC IMPACT OF TOXIC CONTAMINANTS
Donald I. Mount, Ph.D.
U. S. Environmental Protection Agency
Environmental Research Laboratory-Duluth
The impact of toxic pollutants on aquatic systems is often a
principal concern about the manufacture and use of chemicals, that may
be discharged or otherwise released into the environment. Our system of
streams and lakes seems to function in much the same way as the lymphatic
system in our bodies—collecting what "seeps through" as a result of our
industrialized society that liberally employs synthetic chemicals in its
day-to-day operations. Whether toxic chemicals are put in landfills,
discharged in liquid effluents, incineratedj or lost to the air through
vaporization, we should not be surprised to find them in our rivers and
lakes. If toxic and persistent enough, we can expect them to cause
problems, either from direct effects on the environment or through
residues in organisms. We have also learned, that pollutants may be
drastically altered in their chemical form or biological behavior once
released into the complex environment of streams and lakes.
The local area of release is not the only one of concern. We may
well find effects of pollutants occurring far downstream or even in our
coastal waters without any discernible effect in the immediate area of
discharge.
The responsibility of these chemicals after discharge, no matter
how geographically remote or how long after the release has been made,
must become a way of life and a part of doing business. The alternative
under which we have been living in the past places the burden of remedial
measures on those who did not cause the problem. It is encouraging to
see that many companies are as concerned about the consequences of the
chemicals in their wastes as are the regulatory agencies, and I am sure
that this concern will lead to a lessening of the crises that we have
faced in the past few years. Undoubtedly, with modern analytical methods
and our considerably improved knowledge of aquatic systems, the apparent
increase in problem chemicals may well be one of better identification
rather than an increase in problems.
For all of these reasons, and perhaps for many others, there is a
mounting concern and effort to develop more rapid, reliable and cheaper
methods for predicting the impact of toxic chemicals on the aquatic
environment and before they become problems. The passage of the Toxic
Substances Control Act undoubtedly was hastened by these same problems
185
-------
and concerns. Since the emphasis in that act is one of properly testing
chemicals before they are used and before they become a problem, the
need for rapid, predictive tests has been even further strengthened. If
indeed that is the thrust of the regulations which will be developed,
then it is obvious that most of the predictive toxicology work will have
to be done in the laboratory and field studies will be relegated by and
large to an assessment of the accuracy of decisions made during the pre-
market testing period.
The testing of single chemicals under the Toxic Substances Control
Act and the assessment of toxicity from petroleum refinery wastewater
have one thing in common. The variety of pollutants and mixtures in
both is so large that what can be done on any one chemical or any one
waste will have to be quick and relatively inexpensive.
Petroleum refinery wastewater has posed difficult problems to
those trying to develop acceptable waste treatment systems to adequately
remove the toxic chemicals contained in it. These wastes contain chemicals
that are water insoluble but fat soluble, and these are the very ones
which most often cause residue problems in organisms and are often of
the highest toxicity. Unfortunately, one cannot find "reagent grade"
petroleum refinery wastewater on which to do his experiments and make
predictions, so the problems associated with these wastes are compounded
as compared to the ones associated with pure, single chemicals.
At this point, it would be well to discuss the significance of
some specific toxic effects which have received much attention in the
past few years. I refer particularly to carcinogenicity, as well as
teratogenicity and mutagenicity. There are two problem areas in which
these effects should be considered in assessing aquatic impact. One
area is the induction of any one of these toxic effects in aquatic
populations with resultant population effects. In such cases, we must
recognize that these effects are no more significant than many other
effects such as reduction in growth rate, mortality, or reduced reproductive
rates, to the populations of concern. It makes no difference to society
whether aquatic organisms are killed by malignant tumors or by avoidance
to a particular material in a water body. In either case, it is a
decimating factor on the population, but no particular importance is
attached to the effect because it is due to cancer. On the contrary
more emotional importance is attached to human suffering from cancer
than to suffering from an automobile accident. Therefore, carcinogenic,
mutagenic, and teratogenic properties have no special significance for
aquatic organisms. If, however, aquatic organisms accumulate chemicals
with these properties and thereby increase the exposure of human populations
to such chemicals, then we must have special concern.
No useful purpose would be served by listing all of the tests that
are now available for assessing impact on aquatic systems in a paper
such as this one. Great progress has been made in the last fifteen
years in developing more refined and sophisticated tests to measure the
toxicity of chemicals to aquatic systems. Indeed, the state of the art
as it is practiced in aquatic laboratories probably approaches the
186
-------
quality of that used in mammalian toxicology laboratories and standard
protocols for conducting aquatic toxicity tests are easy to find and
comparable to those for mammals. The problems in mammalian toxicology
and aquatic toxicology are all generically the same and revolve around
appropriate test organisms for prediction of effects, quality of test
animals, length of exposure, effects measured, and laboratory quality
control. One who needs to test aquatic impact now has available to him
a selection of test organisms (U.S.EPA 1975) which have been successfully
employed in laboratory testing. There are at least fifteen or twenty
aquatic organisms ranging from fish to protozoans and algae that have
been successfully cultured in laboratories and are adaptable to test
conditions in aquatic testing systems. While some of these species may
not be particularly important in themselves in aquatic systems, such as
for example, Daphnia magna, the data base concerning the sensitivity of
such organisms to a variety of chemicals strongly suggests that they are
not overly sensitive and that they can be used as the "white rat" for
predicting effects on other organisms.
The data base on toxicity of chemicals to^ aquatic organisms is now
reaching proportions large enough so that some reasonable judgments can
be made in the selection of the itf&st appropriate test organisms. Especially
with single chemicals produced for particular purposes, often the objective
in producing the chemical leads one to the selection of the proper
organism. Obviously, if the chemical is produced for use as a herbicide,
it only makes good sense to test its effects on plants such as algae or
macrophytes. Likewise, chemicals designed to kill insects should be
tested on aquatic arthropods and preferably aquatic insects. This is
not to suggest that a variety of organisms should not be tested where
possible, but since time and funds are nearly always limiting, the bulk
of the effort available can be expended on what are likely to be the
more sensitive organisms.
For a variety of aquatic species, there are now acute and chronic
test methods available which enable one to measure effects ranging from
short term, LC50 measurements to very sophisticated and sensitive life
cycle tests (U.S. EPA, 1973; U.S. EPA, 1976; Woelke, 1972; APHA, 1975).
In chronic tests, one can examine the effects of a toxicant on all life
stages of the organism and measure effects on reproduction and progeny
growth as well as effects such as malformations. There are now approximately
100 chronic tests with several fish species on perhaps 50 to 75 different
toxic materials. An examination of this data base reveals that one can
be reasonably accurate for most chemicals by looking at segments of the
life history in toxicity tests and avoid the expensive chronic tests for
full life cycles (McKim, 1977 and Macek, 1977). These authors point out
that if one measures the toxic effects on the eggs and larvae of fish,
he will find as the no-effect concentration one that is not greatly
different from the one that would be found if the animals were exposed
throughout their life history beginning with eggs and ending with growth
data on the F, progeny. Other studies have revealed that the gill-
cleaning reflex commonly called the "cough response" is a highly reliable
indicator of the concentrations which will or will not have chronic
effects in a life history exposure for fish (Spoor, et al., 1971; Drummond,
187
-------
et_ al., 1973 and Drummond, 1974). Through the use of both models and in
a different approach by using resin columns (Neely, et^ al., 1974; Chiou,
e_t al., 1977 and Veith, 1976) bioconcentration properties of chemicals
can be rather accurately predicted from either very short tests with
organisms or through a strictly chemical, analytical approach. Indeed
sufficient correlations exist now to enable us to be reasonably certain
that the single chemical property of the partition coefficient (between
octanol and water) is a reliable predictor of whether or not one can
expect the chemical to bioaccumulate.
Species of fish are now available that make possible chronic life
history tests in a matter of two months or less (Smith, 1973) and tests
with Daphnia and some of the aquatic insects are now fully developed and
can be expected to produce reliable results in an acceptable period of
time. The cost of doing tests such as described above is certainly
acceptably competitive with other analyses that are now required on
chemicals and the insertion of biological tests into requirements for
evaluating chemicals is becoming a matter of routine. Physiological and
biochemical tests on fish in particular and other tests on invertebrates,
have yet to find a promim^^^»lace in assessing the effects of toxicants
on aquatic organisms. Perhaps this is^^»esult of the unregulated body
chemistry which is so typical of most of the aquatic poikilotherms.
It seems quite fair to say that the ability to measure aquatic
toxicity and expected toxic effects on aquatic populations is substantially
more advanced than is our ability to predict the behavior of chemicals
in the environment. The metabolic pathways, the transport of these
chemicals from area to area, and perhaps most important of all, the
permanency of the apparent sinks such as the sediment in lakes, seem the
most difficult judgments to make when assessing environmental impact.
Estimating what concentrations will occur as a result of expected production
and usage rates and how the chemicals will get there, where they will go
and how long they will stay before they are degraded into something
else, is not routine.
This leads me to express a word of caution about the confusion
which can exist regarding the distinction between toxicity and hazard.
In our rush to assure that highly toxic chemicals are not released into
the environment, there is a danger that we will reject a chemical because
it has a high toxicity and a low hazard in favor of another chemical
which has a much lower toxicity but a higher hazard. Our systems, both
aquatic and terrestrial, have lived, evolved and thrived in the presence
of some v«ry highly toxic substances. Ozone rates very high on the list
of toxic materials, and yet its persistence in the environment is so
short that it is not a problem. Too often, I'm afraid, demonstrated
high toxicity casts the dye before an assessment has been made as to
whether or not the chemical is going to be an environmental hazard.
Experience has taught us that the chemicals to be most concerned about
are those which are persistent and are nearly water insoluble but highly
fat soluble. These are the ones which last for a long time in the
environment and accumulate in the bodies of aquatic organisms. They too
are the ones that are likely to form unacceptable residues in those
188
-------
organisms used for human consumption. During the last five years, this
category includes most of the emergencies that have arisen on the'Great
Lakes, the James River, and elsewhere in the country.
With the widespread use of rapid tests, a danger exists that we
will stop developing better methods that are both quicker and more
accurate. An even greater danger is that we will discontinue the develop-
ment of chronic toxicity data and other information which is so vital to
understanding how chemicals affect aquatic systems. It has only been
through the development of an extensive chronic toxicity data base that
the validity of many of these short-cut methods, now available, could be
evaluated. Unless we continue to advance the basic science of aquatic
toxicology, the return on future efforts to develop better methods and
make better predictions will be lessened. We have an urgent need to see
that a proper balance occurs between funds expended for the development
of laundry lists of numbers for making regulatory decisions and, on the
other hand, the advancement of the basic science of aquatic toxicology.
We need better understanding of modes of action and further work on
selection of the most appropriate organisms on which to perform our
tests. Neither can we forget that our ultimate goal is to protect a
system which is not a random collection of individuals, but rather a
relatively intricate grouping of plant and animal populations which are
interdependent on each other and which will all be affected by a change
in any one. Sometimes one gets the impression that those working on
aquatic ecosystems think that there is a universal aquatic ecosystem
which, if understood, could explain all other systems. Certainly such
is not the case, but indeed most workers do expect that the general
functions that occur in aquatic systems are similar enough that once the
fundamental ones are understood, data from laboratory experiments such
as the toxicity tests described above can be more intelligently and
efficiently applied to the problems of the real world.
The increasing scarcity of natural resources and the attendant rise
in cost for these materials will probably force us to use a property of
ecosystems which we have tried not to use in the past decade. I refer,
to the assimilative capacity of waters which can be so useful to us, but
which has been so abused during the first three quarters of this century.
Because we so foolishly abused that valuable resource of aquatic systems
in the past does not in any way preclude the intelligent use of it in
the future. As man becomes smarter about the total ecological effect of
his activities which satisfy his seemingly infinite desire for contraptions,
we may recognize that the mountains of sludge that we produce in our
chemical waste treatment plants and the attendant environmental damage
and resource drain that accompanies the mining, production, transportation
and application of these chemicals, may well create a far more serious
ecological effect than the intelligent utilization of the assimilative
capacity. If one takes an old fashioned and true ecology course, one
of the first principals he learns is that the environment affects all
organisms and all organisms affect the environment. Man is an organism.
He has, does now, and always will affect the environment in which he
lives, and our goal must be to affect the environment in the least
adverse way. I am convinced that it would be easy for us in our efforts
189
-------
to avoid aquatic impact "at any price" to produce a far greater ecological
impact on the total system, that is more undesirable from man's point of
view—all because we made a mistake in the past by expecting the impossible
from our aquatic systems.
REFERENCES
American Public Health Association, American Water Works Association,
and Water Pollution Control Federation. 1975. Standard Methods for the
Examination of Water and Wastewater. 14th ed. Washington, B.C. 1193 p.
Chiou, C.T., V.H. Freed, D.W. Schmedding, and R.L. Hohnert. 1977.
Partition coefficient and bioaccumulation of selected organic chemicals.
Environ. Sci. & Tech., 11: 475-478.
Drummond, R.A., G.F. Olson, and A.R. Batterman. 1974. Cough response
and uptake of mercury by brook trout, Salvelinus fontinalis, exposed to
mercuric compounds at different hydrogen-ion concentrations. Trans.
Amer. Fish Soc. 101: 244-249.
Drummond, R.A., W.A. Spoor, and G.F. Olson. 1973. Some short-term
indicators of sub-lethal effects of copper on brook trout, Salvelinus
fontinalis. J. Fish. Res. Board Can. 30: 698-701.
Macek, K.J. and B.H. Sleight, III. 1977. The utilitiy of toxicity
tests with embryos and fry of fish in evaluating hazards associated with
the chronic toxicity of chemicals to fishes. In: Aquatic Toxicology
and Hazard Evaluation. (F.L. Mayer and J.M. Hamelink, editors). American
Society for Testing and Materials. STP 634. In press.
McKim, J.M. 1977. Evaluation of tests with early life stages of fish
for predicting long-term toxicity. J. Fish. Res. Board Can. In Press.
Neely. W.B., D. R. Branson, and G.E. Blau. 1974. Partition coefficient
to measure bioconcentration potential of organic chemicals in fish.
Environ. Sci. & Tech. 8: 113-1115.
Smith, W.E. A cyprinodontid fish. Jordanella floridae, as a laboratory
animal for rapid chronic bioassays. 1973. Jour. Fish. Res. Board Can.
30: 329-330.
Spoor, W.A., T.W. Neiheisel, and R.A. Drummond. 1971. An electrode
chamber for recording respiratory and other movements of free-swimming
animals. Trans. Amer. Fish. Soc. 100: 22-28.
The Committee on Methods for Toxicity Tests with Aquatic Organisms.
1975. Methods for acute toxicity tests with fish, macroinvertebrates,
and amphibians. U.S. Environmental Protection Agency, Duluth, Minn.
Ecological Research Series EPA-660/3-75-009.
U.S. Environmental Protection Agency. 1976. Bioassay procedures for
the ocean disposal permit program. Environmental Research Laboratory,
Gulf Breeze, Fla. Ecological Research Series EPA-600/9-76-010.
190
-------
U.S. Environmental Protection Agency. 1973. Biological field and
laboratory methods for measuring the quality of surface waters and
effluents. National Environmental Research Center, Cincinnati, Ohio.
Ecological Research Series EPA-670/4-73-001.
Veith, G.D. and N.M. Austin. 1976. Detection and isolation of bio-
accumuable chemicals in complex effluents. P. 297-302. In; Identi-
fication & Analysis of Organic Pollutants in Water. (Lawrence H. Keith,
editor). Ann Arbor Science, Ann Arbor, Mich.
Woelke, C. E. 1972. Development of a receiving water quality bioassay
criterion based on the 48-hour pacific oyster (Crassostrea gigas)
embryo. State of Washington, Dept. of Fisheries, Spokane, Wash. Technical
Rpt. No. 9.
DISCUSSION
L. Duffy, Standard Oil of Indiana: With the demands and problems of the
analysis of the problem organics, do you view bio-assay as a better
analytical tool?
Mount; Yes. The only valid way to measure toxicity is with an organism.
You can't do it with, an analytical instrument and yet we seem to rely on
the analytical approach with no regard for biological response even
though that really is the goal toward which much of this work is aimed.
I think that the toxicity test or the bio-assay should be considered as
an analytical tool and that it can do much to reduce the costs of doing
complex analytical work. Also, the high pressure liquid chromatographic
column, for example, might separate away 90% of the compounds in a waste
or a mixture of materials that we are not all that concerned about and
help us zero in on those which are going to be problems.
Paul Mikolaj, Lion Oil Company: What is the state of the art in the
future of continuous-flow bio-assays?
Mount: The chronic test that I mentioned must be done in a continuous
flow system and as I said, I think it must remain the foundation of our
aquatic toxicity work, but it doesn't have to be a routine workhorse. It
is through chronic tests, analagous to the two-year rat study which is
so common in other toxicology* work, that you find the mode of toxicity.
Such information is necessary for rapid test development and predictive
toxicology. So I see the chronic test as being the workhorse in the
research laboratory, so to speak, where one develops the fundamental
toxicology to evaluate the suitability of much faster methods. In the
prepared paper I have pleaded for a proper balance between the effort
that goes on fundamental research and developing numbers in a production
laboratory. The chronic test is an extremely essential tool as a fundamental
test and of course, depending on what problems you are trying to mimic
in the real world, it may be very useful for other purposes. If one is
concerned about a relatively confiftnuous discharge, then obviously the
flow-through system is the way to go. If one is concerned about a
Pesticide application, then the exposure period is likely to be short,
191
-------
and then perhaps it is not the right test. I think we should be careful
that we don't confuse continuous flow with continuous exposure, because
they are different tests.
BIOGRAPHY Donald I. Mount
Donald I. Mount holds a B.S. degree in
Wildlife Conservation from Ohio State University.
He also earned a M.S. and a Ph.D. from Ohio
State University, concentrating on fish toxicology
and physiology. He is currently Director of the
Environmental Protection Agency Research Labora-
tory-Duluth,Minnesota. Dr. Mount has served on
various national and federal committees in the
interest of water quality control. He has also
published over 40 articles.
192
-------
AN OVERVIEW OF ADVANCED TREATMENT SYSTEMS
Davis L, Ford, Ph.D., P.E.
Senior Vice President
Engineering-Science, Inc.
The term "advanced treatment systems" can have many connotations,
ranging from what is generally considered a 1977 level of technology
(Best Practicable Control Technology Currently Available) to Best Avail-
able Treatment Economically Available required by 1983, or some level in
between. Translating these levels of technology into unit process
requirements, the systems discussed in this overview include biological
treatment, bio logically-treated effluent polishing, and physical-chemical
treatment, primarily related to petroleum refining, petrochemical, and
organic chemical wastewaters. Recent developments relative to process
optimization and limitations will be included, as well as documentation
of process performance.
BIOLOGICAL TREATMENT
Although biological treatment systems per se are normally not considered
"advanced," they do serve as the most important component of most treat-
ment facilities which either now or in the future must produce effluents
with a quality consistent with advanced wastewater technology. The
trend during the past decade has been toward the use of high rate biological
processes for the treatment of organic juidustrial wastewaters. The
systems have generally included either ^fuspended-growth (activated
sludge) or fixed-growth (rotating biological surface) processes or some
modifications thereof. The completely mixed activated sludge process is the
most widely applied biological system in treating industrial wastewaters
with relatively high organic concentrations. The problem most common
with activated sludge systems treating industrial wastes is accomplishing
effective solids-liquids separation in the gravity clarifier which
follows the aeration basin. Many industrial wastewaters will tend to
generate a significant fraction of dispersed biomass which do not adequately
separate in the ,^^rifier. For example, the average effluent suspended
solids (TSS) wil^range from 25 to 75 mg/1 from an activated sludge
process. If the effluent limitations are more restrictive than the
indicated range, then effective effluent polishing systems must be
included in the process design, such as the ability to add organic
polymer flocculants and/or granular media filtration.
More recently, fix^-growth systems are becoming quite popular in the
petroleum refining and petrochemical industrial categories (Ref. 1). One
of the more popular fixed-growth systems is known as the rotating biological
surface (RBS). In this system the biological mass grows on the surface of
large-diameter discs whicjfe are placed side by side on a rotat^jig shaft.
The bottom portion of the rotating discs are emersed in a basin through
which the wastewater flows. It is a facultative system, with the oxygen
transferred by direct contact between the slime and atmosphere as well as
air entrainment in the turbulence associated with the rotation. The system
overcomes some of the disadvantages of the stationary trickling filter
193
-------
approach which includes the continuous shearing of the fixed biomass as the
discs pass through the water thus preventing an accumulation of surface
growth, and an effective penetration of oxygen through the outer layers of
the biomass. It has lower energy requirements than the activated sludge
system, particularly when mixing rather than oxygen controls design. Two
other advantages which are proposed but not yet demonstrated in the refining
and petrochemical industry is the fact that suspended solids in RBS effluents
are lower than those from activated sludge systems and nitrification can be
better accomplished by concentrating the nitrifying microorganisms in the
latter stages of discs on the shaft, and adjusting the pH in the nitrifying
stages to maximize the biochemical oxidation of ammonia. If the organic
concentration of the industrial wastewater is high (the BOD5 exceeds
1,000 mg/1 for example) the capital costs and energy requirements of the
RBS system over activated sludge may be reduced. Moreover, if many of the
organic compounds are refractory and require long contact periods for
adequate degradation, the effective biological growth-substrate contact may
be insufficient to reduce these refractory compounds to the required level.
Therefore, one should be certain that the RBS system can reduce the COD
level adequately before making the final process selection.
Process Flexibility for Biological Systems
One of the primary limitations in applying the biological method of
treating industrial wastewaters has been the failure to incorporate proper
pretreatment and process flexibility facets into the basic design. As the
biochemical oxidative mechanisms are complex, particularly for industrial
wastewaters discharged from the refining, petrochemical and organic chemical
industries, every effort must be made to accommodate the biological popula-
tion to the maximum extent. Some of the approaches which can be used in
insuring process flexibility are discussed in the following paragraph.
Equalization can be one of the most critical single processes in the
overall biological treatment facility. The deleterious effect of transient
loadings on biological systems, both hydraulic and organic, is well docu-
mented (Ref. 2 and 3). There are several rational methods which can be
utilized from raw waste load variations in size and equalization basins in
order to dampen influent variations (Ref. 4 and 5). It is also prudent to
include auxiliary "off-specification" basins in the biological process
design in order to temporarily receive and store waters of inordinately
high organic concentrations or those with potential toxicity. This water
can then be pumped from inventory back to the biological system at a con-
trolled hydraulic rate. Diversion of the wastewater stream can be accom-
plished automatically using an on-line analyzer. Such basins, along with
equalization, reduce the hydraulic and organic variations to the biological
systems and normally result in significantly higher overall removal
efficiency.
Specific pretreatment steps of industrial,waste are often effective
in enhancing the overall performance. For example, it may be prudent to
dilute the concentration of highly degradable organic constituents to a
concentration level which will allow more effective biochemical oxidation
of organic compounds. This is true when biochemical inhibition can occur
194
-------
which is attributed strictly to influent organic concentration rather than
to constituent complexity or resistance to biodegradation. Predilution in
this case is a legitimate and appropriate pretreatment step to improve the
overall performance of the biological system, particularly when recognizing
that kinetics of removal are more concentration-sensitive than mass-sensitive.
Another pretreatment approach which can be considered is the steam or sol-
vent stripping of selected waste streams, with proper air emission control
measures as applicable. This approach can reduce high organic loads,
sequester organic load variations, and remove potentially toxic or inhibitory
contaminants, improving the amenability of the stream to biological treat-
ment. Probably the most common example of this pretreatment is sour water
stripping in petroleum refineries, but there are numerous other instances
in the chemical processing industry where this is an effective pretreatment
step. Recent practice has indicated that hydrolyzing selected organic
wastewater streams by adding caustic and exercising pH control can enhance
the biodegradability of the hydrolyzed stream (Ref. 6). This practice has
been applied as a pretreatment step in the biological treatment of pesticide
and herbicide waste streams with positive results.
One operational technique which can provide additional resistance to
biological upset is increasing the inventory of biological solids in the
aeration basin of a fluidized activated sludge process. This can be
accomplished by increasing the sludge recycle ratio and/or reducing sludge
wastage. The increased inventory simply implies that the quantity of bio-
toxic or biostatic constituents per bacteria is reduced. The design MLSS
levels in activated sludge systems typically range from 2500 to 3000 mg/1
while in some cases for industrial applications, the MLSS level are main-
tained from 8000 to 10,000 mg/1 (Ref. 7).
Process Optimization of Biological Systems
When considering advanced treatment systems, one assumes the biological
portion of the facility is designed and operated for maximum performance.
This may not be the case and some discussion is therefore merited.
Sludge Ape Sludge age, or the average contact time between the microorganism
and the substrate, is becoming increasingly popular as a process
control parameter (Ref. 6, 7). Sludge age can be defined mathematically
using several approaches, the most common for activated sludge being:
e
c
where:
9 = sludge age, days
X -V = average aeration basin MLVSS, mass (V = basin volume)
Si
AX = sludge wastage, mass per day
In a controlled reactor, the sludge age is similarly defined as:
195
-------
where: -.,-,
Q = total system flow
Q = solids wastage flow
X = solids wastage concentration
X = effluent solids concentration
e
The sludge age and the more commonly used Food-to-Microorganism (F/M)
ratio can be related in the following manner:
I _ a(AF/AT) -b
9 X -V (3)
c a
where:
a = sludge yield coefficient
AF/AT = the mass (Ibs) of food (COD or BOD) per day removed
X *V = the mass (Ibs) of microorganisms in the aeration basin
a
b = the endogenous rate coefficient
The hydraulic retention time, t, and its interrelationship to sludge
age has not been adequately defined in terms of process kinetics, although
such a relationship would be particularly meaningful in developing design
equations. There are many indications that contact time between the
biomass and the waste constituents, as measured by hydraulic retention time,
can also be an important process parameter for treatment of complex organic
wastewaters.
Recent investigations have suggested that sludge age is the best con-
trol parameter, and, contrary to some theories, an extended sludge age of
forty days or more maximizes performance in terms of sludge settleability,
process control, and organic removal efficiency (Ref. 8). Other studies
have shown that the critical sludge age (defined as the minimum 0 necessary
to achieve maximum organic removal) is a function of substrate anS tempera-
ture, but does not exceed six to seven days even for a complex chemical
waste at temperatures of less than 10°C (Ref. 9). It can only be concluded
from these investigations that the optimum sludge age for an activated
sludge system treating industrial wastewaters is dependent on the nature
of the influent, namely, its concentration and complexity, and the operating
temperature of the aeration basin. In other words, sludge age alone does
not adequately define the ability of a suspended-growth biological
system to provide maximum removal of organics from a specific wastewatef.
The Food-to-Microorganism ratio (F/M) and hydraulic retention time are also
important control criteria in some cases. Treatability studies are thereby
justified to establish the design 0 and other control parameters specific
to the industrial waste and the mosi severe operating condition.
Temperature
It is important to pursue the temperature effects on biological systems,
as the Streeter-Phelp's empirical modification of Arrhenius' law has not
196
-------
always held in predicting reaction rate coefficients at defined temperatures.
This equation is stated as follows:
where:
KT = the reaction rate at temperature, T
K_, = the reaction rate at temperature Tn
Ll 1
9 = the temperature activity coefficient which is a
constant for a given wastewater
It has been proposed for several years that colder temperatures in the
aeration basins have a more pronounced effect in terms of reduced process
efficiency for wastewaters of higher molecular complexity and solubility
(Ref. 8). This has recently been confirmed by determining the critical
sludge age for wastewaters of varying complexity undergoing aeration at
several temperatures as shown in Figures 1, 2, and 3 (Ref. 9). Based on
these studies, the critical sludge age for each temperature and wastewater
can be approximated as follows:
Temperature. °C Critical Sludge Age.
Days
Domestic Wastewater 30° 2
10° 3.5
Chemical Wastewater 30° 2.5
10° 5.5
Petrochemical Wastewater 30° 3.5
10° 8
It is of paramount importance, therefore, that designers provide sludge
ages which are adequate for maximum performance predicated on wastewater
complexity and swings in operating temperatures. This is particularly
important for systems with long hydraulic retention times since aeration
basin temperatures will approach ambient air temperature even if the waste-
water is quite warm before aeration.
Bulking Sludge
The solids-liquid separation phase of biological treatment has always
been one of the more important elements in successfully treating wastewater
using this system. Bulking sludge is one of the main, precursors to high
effluent TSS levels and consequently has received much attention in the
attempt to optimize biological treatment facilities. Sludge bulking is
particularly prominent in the food processing industries, primarily based
on the fact that an easily available carbon source tends to promote fila-
mentous microorganisms. However, sludge bulking also occurs in many other
industrial categories, and consequently occupies a role of primary importance
in evaluating methods of process control. Historically, bulking is extremely
197
-------
difficult to correct once it has occurred and thus, a preventive approach
is generally most successful. As filamentous microorganisms tend to
thrive at lower pH, oxygen tension, and nutrient levels than do the floc-
culating microflora, it is important to insure adequate pH control,
sufficient aeration, and an adequate supply of nitrogen and phosphorus to
the system. The use of dissolved air flotation with polymer addition as
the final clarification step is receiving an increasing amount of attention.
If the wastewater contains an easily available carbon and filamentous
organisms tend to historically persist, dissolved air flotation with chemi-
cal addition facilities should be given careful consideration. It must
be recognized, however, that effluent TSS concentrations from biological
systems using DAF as the first cell separation step will range, as for
conventional activated sludge systems, from 25 to 75 mg/1 which still may
be inadequate to meet criteria, necessitating effluent polishing.
Preliminary results in evaluating contact stabilization against com-
pletely mixed activated sludge for treating one industrial waste indicates
that the activated sludge system is less prone to produce a filamentous
population and thus less susceptible to sludge bulking. This is possibly
attributed to the fact that in the contact stabilization approach, a higher
food-to-microorganism ratio in the contact tank promotes more filaments
which persist through the reaeration phase of the process. For this reason,
conversion of contact stabilization to completely mixed activated sludge
may enhance overall sludge settleability and process performance in systems
with bulking problems.
The addition of chemicals such as hydrogen peroxide to the aeration
basin or recycle sludge to minimize sludge bulking has had mixed success,
although operating costs are a significant factor if the procedure must be
implemented continuously to prevent recurrence of the problem. The concept
of biodynamic control using a controlled seeding of sludge microflora to
a biological waste system treating food and dairy wastewaters has recently
been reported as an effective method of minimizing sludge bulking (Ref. 10).
Although this is a theoretically sound concept, the practicality of con-
trolled seeding in large biological systems should be verified in terms
of process and cost effectiveness before such an approach is given serious
consideration. r
Optimization of Biological Nitrification
Stringent effluent ammonia concentration levels required by many per-
mits has necessitated the use of nitrification in many industrial biological
treatment systems. As nitrifying microorganisms are extremely sensitive
to pH and temperature, and since many process variables and trace chemical
constituents affect their performance, it is capricious to predict nitri-
fication strictly on sludge age or hydraulic retention time. Consequently,
designers of biological systems are having to use more sophisticated con-
cepts in order to insure the biological removal of ammonia nitrogen from
industrial wastes. The concept of two-stage activated sludge has been
proposed, utilizing the advantages of isolating the nitrifying microorganisms
in a second state while minimizing nitrifying inhibition factors in the first
stage of aeration (Ref. 11). Such a concept, although considerably more
198
-------
expensive than a single-stage system, does provide more process control for
enhancing nitrification. For example, the nitrifying reduction attributed
to winter temperatures can be partially offset by increasing the recycle
ratio of the nitrify ing-rich sludge in the second stage, enhancing the
overall nitrification during the critical seasons of the year.
A recent study indicated that an aerobic submerged filter may be a
feasible approach for the economical nitrification of low-strength wastes
(Ref. 12). This system provides an upward flow of liquid through plastic
or natural media and through efficient solids capture and control of
hydraulic detention time, stable nitrification has been reported. Surface
area effects on nitrification are not well documented, but undoubtedly are
significant when comparing suspended and fixed-growth systems relative to
nitrification.
The rotating biological surface (RBS) has also been quite successful
in biologically treating ammonia nitrogen. A properly-designed RBS system
offers inherently the same advantages as a two-stage activated sludge
system, namely, allowing an enriched nitrifying population to develop in
the latter stages of the RBS process. As the pH tends to drop through
an RBS system via the production of carbon dioxide, it may be necessary
to adjust pH in the nitrification stages by the addition of caustic to
raise the pH level to the nitrification optimum of 7.5 to 8.3. It is
important to recognize that 7.1 of alkalinity (as calcium carbonate) can be
destroyed per unit of ammonium ion (as nitrogen) nitrified, underscoring
the need for good pH control, particularly for wastewaters with low
alkalinity.
It should be noted that there have been several process problems in
terms of nitrification when high amine concentrations are present in the
wastewater, probably through biochemical cleavage reactions of the amine
functional group, actually creating ammonia biochemically in the biological
system.
A particular problem in applying nitrification to industrial waste-
waters is the sensitivity of the nitrifying bacteria to a wide variety
of identified and unidentified organic and inorganic chemicals. This is
especially a problem with complex wastewaters such as those of the
chemical processing industries. All other conditions being proper,
nitrification may still not be obtained for a given wastewater unless
the inhibitory components are found and removed. This problem makes it
mandatory that bench and/or pilot-scale treatability studies on the
actual wastewater to be treated be conducted prior to design of nitrification
into a biological treatment system. Often, the treatability studies can
identify certain waste streams and/or components which can be removed or
pretreated to promote nitrification of the total waste stream.
EFFLUENT POLISHING
Polishing the effluent from biological treatment systems clearly con-
stitutes "advanced waste treatment" and is the model for BPCTCA and/or
BATEA levels of technology for several industrial categories. As previously
199
-------
mentioned, there are several approaches for improving biologically treated
effluent quality. The methods discussed here include chemical addition,
powdered activated carbon addition to activated sludge systems, post
filtration, and fixed-bed carbon polishing units.
Chemical Addition
The addition of coagulants or coagulant aids between the aeration basin
and the final clarifier often can enhance quality by reducing TSS and
colloidal materials in the final effluent. The organic oxygen demanding
substances associated with these suspended and colloidal materials are
correspondingly removed.
There have been mixed results as to the efficacy of this approach. Nor-
mally, polymer additions in the range of 1 to 5 mg/1 result in a 10 to 30
percent reduction in effluent TSS from an activated sludge system, although
this may vary. One of the more favorable experiences of adding polyelectro-
lytes to an activated sludge system treating petroleum refinery waste-
waters is shown in Figure 4. As shown in this case, better than fifty per-
cent of the TSS were removed upon the additon of approximately 5 mg/1 of the
polyelectrolyte. There are many variables which influence the applicability
of adding coagulants or coagulant aids to enhance the clarification of
biologically treated effluents, however, so test conformation studies should
be performed. Moreover, polyelectrolytes are very expensive and high con-
centration demands would result in excessive operating costs. The instal-
lation of a chemical feed system and mixing basin also would require a
capital expenditure.
Powdered Activated Carbon Treatment
The addition of powdered activated carbon to activated sludge systems
to enhance settleability and remove residual organic materials has been
proposed for several years and has been implemented on several occasions.
The addition of 450 mg/1 of powdered activated carbon to an activated
sludge aeration basin treating refinery wastewaters resulted in a relatively
substantial increase in process performance as shown in Figures 5 and 6
(Ref. 13). This approach was not considered as a long-term corrective
measure, however, based on the difficulties in handling the powdered carbon
around the aeration basin.
The most significant project currently in operation which utilizes
powdered^ activated carbon and a biological mass and aeration system is
the 40 MGD DuPont-Chambers Works Facility. The initial results indicate
good performance in terms of effluent quality, although the efficacy of
dewatering, incineration, and regeneration of the mixed biological-
carbon sludge has not been proven and could represent the critical path
in the overall applicability of this approach (Ref. 14). Required
carbon dosages and the ability to reuse the adsorbent material obviously
will dictate the cost-effectiveness of this approach as compared to
other alternatives.
As with the chemical addition option described above, only actual
testing of the method and careful cost analyses can determine its
200
-------
applicability to a particular wastewater. Both bench and pilot treatability
studies as well as the experimental use of powdered carbon in the full-
scale system, if possible, should be used in the evaluation process.
Post Filtration
Post filtration of biologically treated effluents has been applied in
several industrial wastewater treatment facilities, and in fact, used as
the BPCTCA model for developing the petroleum refining guidelines. Post
filtration systems, based on field performance, do a reasonably good job
of reducing the effluent TSS concentration to 10 to 15 mg/1, although
chemical addition to the filter influent is often required to strengthen
the floe and make it more filterable. Additionally, there is a practical
limit of 80 to 110 mg/1 of TSS which can be charged to the filter, an
excess of which causes inordinately short run times and reduces the practi-
cality of the filter process.
The operating biological treatment-post filtration systems treating
refinery and petrochemical wastewaters are producing effluents with long-
term TSS averages of 10-15 mg/1. The efficiency of a post filtration
system depends to some degree on the influent TSS concentration as shown
in Figures 7 and 8. The data presented in Figure 7 was developed from
studies using pilot-scale filters receiving a biologically treated effluent.
It is noted that the addition of polymers to both the deep bed and shallow
bed downflow filters made the overall filter performance less dependent
on influent TSS, an inherent advantage of the polymer addition. The data
in Figure 8 were developed in pilot-scale downflow filtration studies,
indicating the same trend and TSS residual concentrations. The reduction
of oil and grease (O&G) compounds attributed to the addition of post-filters
is not dramatic, as indicated in Figures 9 and 10 (Ref. 15). The probability
distribution of O&G in biologically treated effluent from case histories
in the petroleum refining industry indicate a range of 3 to 15 mg/1 (median
values) as shown in Figure 9 while the two systems in the refining industry
which have filters shown in Figure 10 produce a median value O&G concentra-
tion of 7 to 8 mg/1.
Activated Carbon Polishing
There are presently no full-scale operating biological fixed-bed
carbon polishing treatment facilities treating refinery, petrochemical, or
organic chemical wastewaters for which data are available, although some
are reportedly close to beginning operations. For this reason,
pilot-scale studies must be used as the data base. The COD removal in
carbon columns polishing activated sludge effluent as determined in various
pilot-plant studies for petrochemical and refinery plants is tabulated in
Table 1 (Ref. 16). A 59 to 83 percent removal is noted, indicating that
residual COD can be further reduced in such an application. This is
true because there is an inherent process compatibility between biological
and carbon treatment as many compounds resistant to biochemical degradation
are amenable to carbon adsorption (Ref. 17). It should be recognized,
however, that the cost-effectiveness of carbon polishing expressed in
Ibs of BOD or COD removed per cost unit is poor, based on the high cost of
201
-------
removing a relatively low mass of residual compounds. For this reason,
every effort should be made to select less costly modes of effluent polish-
ing if regulatory constraints so allow.
Activated carbon, applied as a process in a physical-chemical system
treating these industrial wastewaters is less applicable, as discussed in
the following section.
PHYSICAL-CHEMICAL SYSTEMS
Most of the previous studies relative to the physical-chemical treat-
ment of industrial wastewaters has centered around activated carbon treat-
ment, although steam and solvent stripping, chemical oxidation, chemical
coagulation and precipitation, and other forms of non-biological treatment
are possibly applicable.
It should first be understood that activated carbon treatment of organic
industrial wastewaters should be carefully investigated prior to making
process commitments. Several studies, for example, have underscored the
limitations of activated carbon as total physical-chemical treatment process
as compared to carbon polishing of biologically treated effluent (Ref. 16,
17, 18, 19). The estimated effluent quality for the activated sludge,
carbon, and combined treatment of refinery wastewaters are tabulated in
Table 2 (Ref. 18). A more recent study comparing activated carbon as a
physical-chemical or polishing process was conducted by the Environmental
Protection Agency (Ref. 20). Both API Separator effluent and biologically-
treated effluent from a petroleum refinery were charged to pilot-scale
columns in order to obtain a comparative evaluation. These quality data
indicated that the carbon system was significantly more effective when
operated in conjunction with the biological process than when applied
singularly, both in terms of BOD and COD. This is consistent with the
results observed in pilot studies conducted by the author. The limitations
of physical-chemical systems designed around the activated carbon adsorp-
tion process therefore are a function of the organic compounds in the
wastewater which are not amenable to adsorption, reducing overall efficiency.
Even though physical-chemical systems have more of an "advanced waste
treatment" connotation, they could, in fact, produce a lower effluent
quality than biological treatment processes.
SUMMARY
In summary, it is the author's opinion that "advanced waste treatment"
for the petroleum refinery and organic chemical industries centers
around some form of biological treatment, at least for the next decade.
There are no direct alternatives which are presently as cost-effective in
terms of chemicals or energy, and there are few likely process candidates
which are likely to be more attractive, at least through the 1983 date
for implementation of Best Available Treatment Economically Achievable.
Physical-chemical treatment is and will continue to serve as an important
adjunct, primarily as pretreatment of specialty streams or as polishing
units in series with biological processes. Although sole physical-chemical
processes are possibly applicable in certain cases, careful conceptual
202
-------
planning and process confirmation through treatability studies should precede
final selection.
The applicability of the activated carbon process in treating indus-
trial wastewaters, for example, is contingent on many factors, including
the amenability of the dissolved constituents to sorption, the presence
of other substances which enhance or impede the sorption process, the sound-
ness of engineering, the degree of pretreatment and proper operation and
maintenance of the system. As activated carbon was one of the primary
processes factored into the development of the 1983 Best Available Tech-
nology (BAT) guidelines for many industrial categories, some of which have
been remanded by the Courts, it is important to fully understand the process
and its limitations.
Finally, it is important to recognize that "advanced waste treatment"
in effect will have an EPA definition when the BPT, BAT, and new source
performance standards for the Organic Chemicals and Plastics and Synthetics
categories, and the BAT standards for the Petroleum Refining category
are repromulgated later this year.
DISCUSSION
Garr M. Jones/ Brown and Caldwell: We have noticed that in many industries and their
municipal wastewater treatment systems the advantages of what we call a coupled system,
a fixed-growth system followed immediately by activated sludge. I noticed that you
passed over this particular combination and I would like to ask you to comment on the
advantages that we see, first of all lower operating costs, smaller clarifiers because of
improved solid settling of characteristics, and a far more stable process.
Davis Ford; Yes, I would say that would have the same basic process concept as the
trickling filters did in the past, and I think from the process point of view it makes a lot
of sense. You say there are lower operating costs; I would only caution there that if you
take the capital cost of that system and amortize that and include the operating costs and
it still is cheaper, then that certainly can be justified on an economical basis. From the
process point of view I would say the addition of a fixed-growth reactor before or even
after activated sludge has some merit certainly to be investigated. I certainly agree
with that concept.
203
-------
REFERENCES
(1) Ford, D.L., and Tischler, L.F., "Recent Developments in
Biological Treatment of Industrial Wastes," submitted
to Chemical Engineering. (May, 1977) .
(2) Ford, D.L., and Eckenfelder, W.W., "The Effect of Process
Variables on Sludge Floe Formation and Sludge Settling
Characteristics," Water Pollution Control, Federation
Meeting, Kansas City, Missouri (Sept. 1966).
(3) Ford, D.L., "Factors Affecting Variability from Wastewater
Treatment Plants," Prog. Water Technology, Vol. 8,
No. 1, pp. 91-111, Pergamon Press, London (1976).
(4) LaGrega, M.D., and Keenan, John D., "Effects of Equalizing
Wastewater Flows," Journal WPCF, Vol. 46, No. 1,
(January 1974).
(5) Speece, R.E., and LaGrega, M.D., "Flow Equalization by Use of
Aeration Tank Volume," Journal WPCF, Vol. 48, No. 11,
(November 1976).
(6) Shell Chemical Company, unpublished internal report (1975).
(7) Ford, D.L., "Water Pollution Control in the Petroleum Industry,"
Industrial Wastewater Management Handbook, pp. 8-1 through
8-75, edited by H. Azad, McGraw-Hill, New York (1977).
(8) Grutsch, J.F., "A New Perspective on the Role of the Activated
Sludge Process and Ancillary Facilities," Proceedings of
the Open Forum on Management of Petroleum Refinery
Wastewaters, sponsored by Environmental Proection Agency,
American Petroleum Institute, the National Petroleum Refiners
Association, and the University of Tulsa, Tulsa, Oklahoma,
(January 1976)..
(9) Sayigh, B.A., "Temperature Effects on the Activated Sludge
Process," Doctoral Dissertation, the University of Texas
at Austin, (May 1977).
(10) Chambers, J.V., "Bioengineering an Activated Sludge Microflora
to Improve Waste Removal Performance," Proceedings of the
Fifth Annual Industrial Pollution Conference, WWEMA,
Atlanta, Georgia, (April 1977).
(11) Adams, C.E., and Eckenfelder, W.W., "Nitrification Design
Approach For High Strength Ammonia Wastewater," Journal
WPCF, (March 1977).
204
-------
REFERENCES (CONTINUED')
(12) McCarty, P.L., and Haug, Roger T., "Nitrogen Removal From Waste-
waters by Biological Nitrification and Denitrification,"
presented at Society For Applied Bacteriology. Liverpool
England, (197lT '
(13) Rizzo, J.A., "Case History: Use of Powdered Activated Carbon
in an Activated Sludge System," Proceedings of the Open
Forum on Management of Petroleum Refinery Wastewaterst
sponsored by Environmental Protection Agency, American
Petroleum Institute, the National Petroleum Refiners
Association, and the University of Tulsa, Tulsa, Oklahoma,
(January 1976).
(14) Davis, J.C., "Activated Carbon: Prime Choice to Boost Secondary
Treatment," News Features, Chemical Engineering. (April 11,
1977) .
(15) Engineering-Science, Inc., Report to the National Commission on
. Water Quality, Petroleum Refinery Industry - Technology and
Cost of Wastewater Control, (June 1975).
(16) ' Ford, D.L., "Putting Activated Carbon In Perspective to 1983
Guidelines," presented at the 1977 National Conference on
Treatment and Disposal of Industrial Wastewaters and
Residues, Houston, Texas, (April 26-28, 1977).
(17) Ford, D.L., "Advanced Wastewater Treatment of Industrial Waste-
waters Using Carbon Adsorption," Proceedings of the Fifth
Annual Industrial Pollution Conference, Water and Wastewater
Equipment Manufacturer's Association, Inc., Atlanta, Georgia,
(April 19-21, 1977).
(18) Ford, D.L., "The Applicability of Carbon Adsorption in the
Treatment of Petrochemical 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,
Tennessee, (September 1972).
(19) Ford, D.L., "Current State of the Art of Activated Carbon
Treatment," Proceedings, Open Forum on Management of
Petroleum Refining Wastewaters. sponsored by the Environ-
mental Protection Agency, American Petroleum Institute,
the National Petroleum Refiners Association, and the
University of Tulsa, Tulsa, Oklahoma, (January 1976).
(20) 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).
205
-------
BIOGRAPHY Davis L. Ford
Davis L. Ford holds a B.S. degree in
Civil Engineering from Texas A & M University and
a M.S. and Ph.D. degrees in Environmental Health
Engineering from fhe 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 insutries, the United Nations
(WHO and PAHO), the EPA and various state and
municipal agencies.
TABLE 1
CARBON PILOT-PLANT RESULTS FOR POLISHING
PETROCHEMICAL AND REFINING WASTEWATERS (REF. 17)
Design
Q Influent COD Effluent COD Percent
Type of Wastewater (MGD) (mg/1) (mg/1) Removal
Petrochemical 3 150 49 67
Refinery 26 100 41 59
Refinery 28 300 50 83
Refinery 8 100 40 60
Petrochemical 29 150 48 68
206
-------
Table 2
Estimated Effluent Quality for the Activated Sludge, Carbon, and Combined
Treatment of Refinery Wastewaters*(Ref. 19)
Constituent
Mean Value Range
Primary Effluent
Activated Sludge
Effluent
Total
Carbon
Effluent
Combined
Activated Sludge-
Carbon Effluent
Remarks
COD
500-700 mg/1
100-200 mg/1 100-200 mg/1 30-100 mg/1
Exact COD residuals vary with
complexity of refinery & design
contact times in the Act.S. and
Carbon Treatment Plants.
BODC
250-350 mg/1
20-50 mg/1 40-100 mg/1 5-30 mg/1
BOD residual depends on BOD/COD
ratio which characterizes rela-
tive biodegradability of waste-
water.
Phenols 10-100 mg/1
NJ
mg/1
<1 mg/1 <1 mg/1 Phenols(ics) are generally amen-
able to biological and sorption
removal.
PH
8.5-9.5
7-8.5
7-8.5
7-8.5
pH drop in Act. S. systems attri-
buted to biological production of
C02 and intermediate acids. pH
change in carbon columns depends
on preferential adsorption of
acidic and basic oreanics.
SS
50-200 mg/1
20-50 mg/1
<20 mg/1 <20 mg/1
Primary effluent solids depend on
design and operation of oil removal
units. Act. S. effluent solids
depend on effectiveness of second-
ary clarified. Low effluent solids
characterize carbon column effluent.
TDS 1500-3000 mg/1
NH3-N 15-150 mg/1
1
1500-3000 mg/1
1-30 mg/1
1500-3000 mg/1
10-140 mg/1
1500-
3000 mg/1
1-30 mg/1
TDS is essentially unchanged
through all three treatment systems
Exact concentration depends on pre-
stripping facilities, nitrogen
content of crudge charge, cor-
rosion additive practice and
biological nitrification.
1-10 mg/1
-------
FIGURE 1 EFFECTS OF SLUDGE AGE AND TEMPERATURE ON BIODEGRADABILITY
IUO
Jsso
•_•
< 60
o
S40
ŁE
g20
O
0
<
3O°C
D
^*~~
/
/
-
^^—
i^^^HBBMM
... •
IO°C
••••I*
l^*~~
i
1OMESTI
cmsn
"WATER
234 56789 10
SLUDGE AGE ,0C, days
FIGURE 2 EFFECTS OF SLUDGE AGE AND TEMPERATURE ON BIODEGRADABILITY
(REF. 9)
100
CHEMICAL WASTEWATER
34567
SLUDGE AGE ,0C, days
8
10
FIGURE 3 EFFECTS OF SLUDGE AGE AND TEMPERATURE ON BIODEGRADABILITY
(REF. 9)
100 v
PETROCHEMICAL WASTEWATER
34567
SLUDGE AGE, 0C, days
8
9
10
208
-------
FIGURE 4
EFFECT OF POLYELECTROLYTE ADDITION
ON EFFLUENT TSS
FLOW PROBABILITY DISTRIBUTION
5 10 20 30 40 50 60 70 80 90 95
% OF THE VALUES LESS THAN STATED VALUE
98 99
KEY
O WITHOUT POLYELECTROLYTE
A WITH POLYELECTROLYTE
-------
FIGURE 5 EFFECT OF EFFLUENT TSS OF CARBON ADDITION TO AERATION
BASIN - PETROLEUM REFINERY (REF. 13)
240
POWDERED CARBON
ADDITION
5 10 30 50 70 90 95 99
% OF VALUES EQUAL TO OR LESS THAN
FIGURE 6 EFFECT OF EFFLUENT COD OF CARBON ADDITION TO AERATION
BASIN - PETROLEUM REFINERY (REF. 13)
280
240
^ 200
o>
J
Q 160
0
o
1- 120
LU
n! 80
U.
LU
40
0
•
«
•
+**
-—*
^-*^
x""
.-
^
-
BA,
X
-
9ŁC
4
~~
4SŁ
/
/
.«*
/
V ^^^
^^«
POWDER ED L
ADDITION
r
^
CARBON
5 10 30 50 70 90 95 99
% OF VALUES EQUAL TO OR LESS THAN
210
-------
Figure 7
POST FILTRATION PERFORMANCE-
PILOT SCALE DEEP BED AND SHALLOW BED
DOWNFLOW FILTERS WITH AND WITHOUT
POLYMER ADDITION
PRACTICAL POST
I— FILTER LIMITATION
40 60 80
INFLUENT TSS (mg/l)
211
-------
Figure 8
POST FILTRATION PERFORMANCE-
PILOT SCALE DOWNFLOW FILTERS
100
o>
CO
h-
UJ
ID
50
NOTE: DATA DEVELOPED
AT HYDRAULIC LOADINGS
OF 2-5
100
INFLUENT TSS(mg/l)
212
-------
Figure 9
OIL AND GREASE PROBABILITY
BIOLOGICAL TREATMENT
(REF. 15)
500
N,
O>
LJ
CO
LJ
a:
CD
o
<
O.I I 10 50 90 99 99.9
% OF THE VALUES LESS THAN STATED VALUE
213
-------
Figure 10
OIL AND GREASE PROBABILITY
BIOLOGICAL TREATMENT WITH
POST-FILTRATION
(REF. 15)
100
50
o>
JE
UJ
UJ
oc
a
<
10
I
0.1 I 10 50 90 99
% OF THE VALUES LESS THAN STATED VALUE
99.9
214
-------
SESSION VI
FUTURE CONSIDERATIONS IN BIOTREATMENT
Chairman
George W. Reid
Regents Professor of Civil Engineering
University of Oklahoma, Norman, Oklahoma
Speakers
J. F. Grutsch
R. C. Mailatt
"Design and Operation: Bases for an Activated Sludge
Route to BAT (1983) Water Quality Goals"
Paul Goldstein and Robert W. Griffin
"Considerations in Reuse of Refinery Wastewater"
Milton R. Beychok
"State-of-the-Art in Sour Water Stripping"
Irv Kornfeld and Jay G. Kremer
"Refinery Discharges to a Large Municipal Sewerage
System"
215
-------
BIOGRAPHY George W.Reid
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 recent Urban Water Systems Conferences and
seminars in Colorado, New Mexico, Bogota, Lima, Europe,
etc.
216
-------
DESIGN AND OPERATION: BASES FOR AN ACTIVATED
SLUDGE ROUTE TO BAT (1983) WATER QUALITY GOALS
J. F. Grutsch
Coordinator, Environmental Projects, Standard Oil (Indiana)
R. C. Mallatt
Manager, Environmental & Energy Conservation, Standard Oil (Indiana)
ABSTRACT
Since its debut 65 years ago, the activated sludge process (ASP)
has been developed into a widely applicable process with a tremendous
capacity for water purification. Unfortunately, the role of the ASP has
become essentially totally identified with secondary treatment, that is,
best practicable control technology currently available (BPT) goals for
1977. When effluent water quality superior to that normally associated
with BPT levels is a goal, for example, best available technology eco-
nomically achievable (BAT), an add-on stage of granular carbon treatment
is the typical response for organics reductions.
Current work is demonstrating that viewing the ASP solely as a
secondary treatment process is shortsighted and, in fact, the ASP pro-
vides the preferred means for achieving BAT water quality goals when
compared to alternatives. A cost effective route for using the ASP to
achieve BAT effluent quality requires that the ASP be operated at
unusually high sludge age and enhanced with high surface-area activated
carbon.
Other reports at this Second Open Forum will describe the success-
ful enhancement of the ASP using powdered activated carbon; this paper,
therefore, will focus on describing: 1) How ASP operation at very high
sludge age is achieved; 2) The support biological and kinetic data give
to defining, as being optimal, the ASP operating conditions required to
operate at very high sludge age, and 3) ASP process-design optimization
at very high sludge age.
INTRODUCTION
At the May, 1977, API meeting, two papers (1,2) presented an alter-
native to granular carbon adsorption treatment of activated sludge unit
(ASU) effluent to achieve "best available technology economically achiev-
able" (BAT) proposed for 1983. The alternate approach to BAT incor-
porates the use of newly developed, ultra-high surface-area powdered
active carbon (PAC) in the ASU of the "best practicable control tech-
nology currently available" (BPT) model sequence for 1977. The alternate
eliminates essentially all the capital costs of the granular carbon_
system. Further, operating costs for PAC can be minimized by optimizing
operation of the end-of-pipe sequence.
217
-------
Optimization of the end-of-pipe sequence for the PAC alternate to
granular carbon treatment requires that the role of the ASU and other
end-of-pipe treatment elements be defined as we described at the First
Open Forum (3). A summary of the operating guidelines follows:
1. A systems optimization of a refinery end-of-pipe treatment
sequence points to the ASU as the key element.
2. To achieve maximum water quality, system optimization points
to reversing the historic work-horse role of the ASU; i.e., it
should be used only for the removal of essentially soluble
contaminants.
3. Reversing the role of the ASU yields a dramatic series of
beneficial effects. At very high sludge age:
a. The SVI characteristics of the activated sludge mass are
excellent.
b. Process control is greatly simplified.
c. The need for many process control tests is eliminated.
d. An exemplary effluent low in TOG and other contaminants
is produced.
e. The net cell yield is remarkably low.
f. The population dynamics of the sludge mass improve.
g. Maximum ASU capacity for purification is achieved.
4. Systems optimization, wherein the key element is using the ASU
for removal of only soluble contaminants, permits clear defini-
tion of the roles the other elements play; i.e., colloidal and
suspended matter must be essentially all removed in pretreat-
ment 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 ASU.
7. Cell genetics, wherein inducible enzymes or mutant species are
required, support operating the ASU at very high sludge age
for maximum purification capacity.
8. Bacteria are essentially enzyme factories. Enzymes are sensi-
tive to temperature, pH, excessive concentrations of heavy
metals, oxidizing agents, salinity, UV, and other radiations.
218
-------
Using the ASU for removal of only soluble contaminants is
consistent with obtaining a stable system less influenced by
transient changes in environmental conditions since the endo-
enzyme 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.
The objectives of this paper are:
1. Outline how an ASU can be operated at high sludge age (50+
days) by outlining and describing in detail the chemistry of
colloid destabilization which must be addressed if the pre-
••- treatment (filter and dissolved air flotation) are to remove
colloidal solids essentially completely.
2. Propose a model for chemical destabilization of negative
colloids by weakly anionic polyelectrolytes.
3. Describe how the electrical properties of the activated sludge
floe impact on the mechanical design of the unit.
KINETIC AND BIOLOGICAL DATA SUPPORTING ASU OPERATIONS ON SOLUBLE CON-
TAMINANTS AT HIGH SLUDGE AGE
Wastes Properties
An intrinsic property of solids in the presence of water is an
electrical surface charge. When colloids are being considered, the
electrical charge is called zeta potential (ZP). Almost all matter
dispersed in spent process water such as oil particles, silt, biocol-
loids, inorganic colloids, etc., has a negative ZP. This repulsive
coulombic charge causes many particles to resist aggregation and settl-
ing in primary clarification facilities. In the case of refineries, the
chemical oxygen demand (COD) of the contaminants in the effluent from
API separators average about 50 per cent soluble and 50 per cent sus-
pended matter (Figure 1).
By comparison, the soluble COD in the effluent from a municipal
primary clarifier is substantially less, being about 15-30 per cent
(Figure 2). Thus, the colloids and suspended matter represent the major
COD component entering secondary treatment.
219
-------
ASU Kinetic Effects Achieved by the Removal of Colloids and Suspended
Solids from the Feed Stream
Provides a Response to Biological Oxidation Kinetics. A recent
kinetic equation (equation 1) by Adams, Eckenfelder, and Hovious (4, 5)
applicable to a completely mixed ASU predicts that best effluent quality
(lowest Se) will be achieved with the lowest feed strength
Se = Si (Si - Se)/KMt ' ' Equation 1
Where, Se = Soluble organics in effluent (mg/1)
S^ = organics in influent (mg/1)
K = kinetic constant
M = biomass (mg/1)
t = time
The authors points out that Si/Mt = F/M, and letting F = Si/Mt,
equation 1 becomes:
Se = Si/(KF~1 +1) Equation 2
Equation 2 clearly points out that a low F/M ratio (high sludge
age) and low feed strength are associated with optimized ASU operations.
Removing colloids and solids in pretreatment facilities has been demon-
strated to be an effective means to achieve dramatically better ASU
effluent quality (1, 2, 3).
Reasons for improved effluent quality with reduced feed strength
can be visualized better with reference to Figure 3. Transport of sub-
strates through the cell membrane counter to concentration gradients is
achieved by enzyme transport systems (permeases) . Once inside the cell,
the substrate molecules are acted upon by a coordinated and sequential
series of enzymes; there are several thousand endoenzymes in a single
bacterium (6) . A small part of the enzyme sequence may be as pictoria-
lized in Figure 4.
Enzymic reactions are reversible. Accumulation of products affects
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 concentration of sub-
strate. 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 the 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, the 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 affect
the rate of reaction. In many instances, enzyme-catalyzed reactions
220
-------
appear to proceed in only one direction because the equilibrium point is
very far in that direction. In other cases, one or more of the end
products may be removed constantly by some mechanism so that the equi-
librium is never reached. Under normal conditions in the living cell
enzyme reactions are constantly pushed in this manner toward one or the
other side of the reactions. Theoretically, reversible reactions cannot
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 (7, 8).
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 con-
trol called feedback inhibition. Thus, if a component of the waste
substrate periodically is the same as biological intermediate 814, the
internal concentration of 8^4 increases which increases the concen-
tration of the preceding intermediates causing more 8^3, and Sg to be
transported externally, and less So to be transported internally; i.e.,
high residual organics are observed in the solvent phase.
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 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. The result is, of course, higher
residual organics in the solvent phase. This is called feedback repres-
sion. 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 accumulation of end products
(feedback repression) in the enzyme sequence (Figure 4). In a sequen-
tially operating system, the end product of each enzyme can be the
inducer of the next enzyme in the series and the inhibitor or repressor
of the preceding enzyme, thus carrying forward the work of the enzyme
factory (9).
If the principle component of the substrate being treated is S0,
various biological intermediates (identified as S6, S13, S^, and S22J
may also have a propensity to be transported externally by the enzyme
transport system. Further, if the biological intermediate is three
carbon atoms or less, it can diffuse across the cell membrane, inus,
221
-------
with increasing concentration of So for fixed biomass, the concentration
of internal biological intermediates will increase. Since enzyme equi-
libria are similar to chemical equilibria, increasing the concentration
of internal biological intermediates (i.e., 85, 813, 8^4, and 822)
causes their transport outside the membrane as a response and the concen-
tration of external organics increases. Starved systems (high sludge
age, low F/M) are predicted from biochemical principles to be lowest in
residual soluble substrates.
On the other hand, a complex substrate, So, could,result in unex-
pectedly high residual soluble substrates by intermittently supplying a
biological intermediate of a more complex substrate which shuts down the
degradation of the complex substrate by feedback inhibition or repression.
Permits Operation at High Sludge Age. Minimizing the feed strength
by removing colloids and suspended matter not only yields a purer ef-
fluent but provides the means to respond to sludge age. Outlined in the
schematic of Figure 5 are properties of the BOD/COD materials. The
nonorganic and slowly bioxed materials increasingly accumulate in the
sludge mass with increasing sludge age causing deterioration of the
sludge settling properties (3). >This deterioration of sludge settling
properties is apparently caused by changing the electrical properties
(zeta potential) of the sludge which is discussed in a later section.
ASUs can be operated easily at a very high sludge age (50+ days) if the
process is protected from colloids which are inert or only extremely
slowly biodegraded. Successful operation at high sludge age responds to
those soluble components requiring acclimation of the biomass or having
a slow biox rate.
The improvement in ASU effluent quality with increasing sludge age
typically is uniformly demonstrable up to about 20 days sludge age; for
operation higher than 20 days sludge age, improved effluent quality may
or may not be observed. This appears to be related to operating tem-
perature and the nature of the substrates. For example, with reference
to Figure 6, Curve A is typical of the ease with which organics in
municipal and many industrial effluents are removed biologically. ASU
operation at low SA is very effective, increasing operation to 20 days
SA yields modest effluent-quality improvement and beyond 20 days, no
improvement is observable.
At the opposite end of the spectrum are organics that yield a
removal curve like E; i.e..similar to nitrification in a refinery
effluent. This curve indicates that microorganisms needed for such slow
reproducers that high sludge age operation is needed to accumulate
sufficient appropriate organisms for organics removal.
Curve B typifies an organic originally removed at a high rate but
which generates a biological intermediate that is removed at a much
slower rate. The biological intermediate is rate limiting and increas-
ing sludge age is more effective for decreasing organics in the effluent
for this example than for an organic typified by curve A.
222
-------
Curve D is illustrative of organics whose rate limiting reaction is
probably the initial, or an early, enzyme reation in the degration
sequence and increasing sludge age increases the removal by increasing
the enzyme concentration.
Curve C is probably representative of a sequential enzymic reaction
series wherein a series of rate limiting reactions are encountered.
Thus, increasing sludge age has the effect of increasing the supply of
enzymes needed for the reactions in the later stages which measures as
reduced organics in the effluent.
As operating temperatures decrease, biological oxidation rates
decrease and more microorganisms are needed to achieve the same sub-
strate removal. Thus, the lower the operating temperature, the more
pronounced the impact of increasing sludge age. In an Amoco study (1)
with parallel units operating at a median temperature of 14°C (57°F)
compared to a control unit at 20 days sludge age, the percentage improve-
ment in effluent quality by prefiltration and increasing sludge age from
20 to 60 days is:
Per Cent Improvement in Effluent Quality By
Prefiltration and Increased Sludge Age
Prefiltration and
Parameter Prefiltration Increased Sludge Age
SOC 9.4 20.3
SCOD 20 36.3
NH3-N - 58.7
Phenolics - 29.6
On the other hand, when operating at 85°F, Crame (2) reported
similar residual organics at sludge ages beyond 20 days.
Minimizes Waste Sludge. Minimizing the feed strength by removing
colloids and suspended matter not only yields a purer effluent but
provides the means to achieve operation at very high sludge age which
provides for minimum waste biosludge production. Bacteria in the acti-
vated sludge mass use the energy available in the substrate for two
general purposes. As shown in Figure 3, the substrates can be used for
1) cell maintenance energy and 2) the repair and generation of new cell
material. The cell maintenance energy requirement gets first call on
the substrate resources. If there is substrate left over, the bio-
logical system proceeds to increase in response to the available food
supply. Thus, to minimize production of waste sludge, it is obvious
that the ASU should have a large mass of activated sludge relative to
the food supply; i.e., a high sludge age (SA) which is, of course, the
same as a low F/M.
223
-------
The relationship between SA, F/M, maintenance energy, and cell
generation has been known for some time (10):
1^ = a (dF/dt) - b Equation 3
SA M
where, SA = sludge age (days)
a = cell yield coefficient
b = cell maintenance energy coefficient
M = biomass, Ibs.
F = COD, Ibs.
t = Time
Using this relationship and data from Crame (2) and Grieves et al
(1) which are optimized, completely mixed ASUs operating on the soluble
substrate in refinery effluents, the following values were calculated
for the coefficients -a and b at two different temperatures:
Temp. (F) a b
Refinery A 57 .317 .015
Refinery B 85 .3 .03
Comparison with literature data for various wastes suggest these
are appropriate coefficient values.
These coefficients and equation 3 can be used to estimate the
equilibrium biomass inventory and the resultant waste sludge generation
as a function of sludge age as shown in Figure 7. Considering the
curves for refinery B to be most typical operating conditions, it is
difficult to rationalize the prevailing practice of operating ASUs at 5-
15 days SA, a condition generating maximum waste biomass with its atten-
dant high disposal cost. The waste sludge can be reduced by 2/3 by
operation at very high SA. The time span to 150 days SA is included
because Amoco is operating a test unit at this condition.
Simplifies Process Control. The sludge age method of process
control has many advantages which are discussed in detail by Walker (11).
Operating at a very high sludge age is even simpler. For example, at 50
days sludge age, 2 per cent of the inventory-activated sludge is wasted.
This is such a small amount that sludge wasting can be done on about a
twice/week basis.
ASU Biological Effects Achieved by the Removal of Colloids and Suspended
Solids from the Feed Stream
Improves Flocculating Capacity of the Activated Sludge. As shown
in Figure 8, activated sludge has a bimodal floe size distribution. A
good flocculating and settling sludge has a preponderance of large floe
mass compared to the fine particle fraction. The large floe mass is, in
224
-------
essence, a biopolymer analogous to a weakly anionic polyelectrolyte-type
polymer. Thus, the large floe mass is a biopolymer flocculant and the
fine particles represent the phase to be flocculated. Qualitatively, -
when the ratio of the two phases overwhelmingly favors the biopolymer as
illustrated by the solid line, a good flocculating sludge results. When
the fine particles predominate as illustrated by the dotted line, their
surface area overwhelms the capacity of the biopolymer for flocculation,
and "arms and legs" (turbidity) is observed in the supernatant.
Inert and slowly biologically-oxidized colloids and suspended
matter contribute to increasing the amount of fine particles in the
sludge mass. Increasing SA increases their accumulation, thereby con-
tributing to the deterioration in sludge flocculating properties.
Removing colloids and suspended matter before ASU treatment results in
an excellent activated sludge at very high sludge age.
Improves Zeta Potential Probability Distribution of Activated Sludge
Particles. Another way of looking at the settling properties of acti-
vated sludge and the impact of the fine particle fraction on the settling
properties that is much more enlightening in terms of understanding why
solids are lost over an ASU final clarifier weir is to examine the zeta
potential probability distribution of the activated sludge particles.
Zeta potentials are typically reported as averages which is mis-
leading, always, and particularly in the case of activated sludge. For
example, the probability distribution curves in Figure 9 show the median
ZP values of a good settling sludge to be -11 mV and a poor settling
sludge to be -12 mV. This difference is not only difficult to measure
but is really insignificant. What is significant is the slope of the
distribution curve and the ZP of the highest-charged particles. If, for
example, a ZP of -14 mV or more negative provides enough repulsive force
that the fine particles will not flocculate, fully 8 per cent of the
solids in the poor settling sludge mass resist flocculation and settling.
On the other hand, essentially all the particles in the good settling
sludge are well below -14 mV ZP and flocculate well. Limiting the
accumulation of colloids in the activated sludge by effective phase
removal in pretreatment facilities is a principal means to control the
ZP probability distribution of activated sludge. '
ACTIVATED SLUDGE UNIT DESIGN
Mechanical
Mechanical elements that impact most strongly on the ASU are aera-
tion units and pumps—both have been observed to be a principal cause of
process failure through shear of sludge, increasing the detritus, and
poor SVI, ZP properties.
Aeration. High-speed aeration and brush aerators used in mixed
liquor tanks have been observed to disrupt sludge severely causing tne
generation of colloidal fines. This upsets the bimodal floe size distri-
bution as previously discussed, and the biopolymer component cannot
225
-------
supply the flocculant capacity to completely reflocculate the detritus
into the sludge mass; i.e., the ZP probability curve slope increases
because of the colloidal detritus. Low-speed aerators and compressed
air systems are alternates that do not have these undesirable properties.
Pumps. High-speed centrifugal pumps and high-pressure drops across
valves, etc., wreak havoc on recycled, activated sludge floe. Low-speed
recessed impeller, low head pumps have been successfully applied in
systems with fairly long aeration tank retention times of 12 hours or
more. These long retention times may have contributed to the success by
supplying reflocculation time. An attractive sludge recycle system is
one used in the municipal sector, air-lift pumps. Since these are low
head pumps, proper hydraulics are required and retrofit is not attrac-
tive in many existing plants.
Process
Reactor Design. Much has been written about completely mixed and
plug-flow reactor designs; completely mixed gaining in popularity over
the earlier plug-flow design. Plug flow theoretically yields a slightly
better effluent on well-equalized feed streams, and completely mixed
systems are less sensitive to some toxic transient loadings.
One area ignored is the role reactor design plays in excess sludge
generation; the plug-flow design inherently yields more excess sludge.
As an example, Lau (12) recently reported a comparison of results from a
completely mixed unit and volumetrically equally sized three-staged
reactor completely mixed unit. Treating high-strength, readily biode-
gradable wastes at various F/M (COD/MLSS) ratios, Lau observed the
following waste sludge yields:
Sludge Yield, Per Cent of Yield for Single
Stage Unit at F/M of 0.5
F/M Single Stage Three Stage
0.5 100 150
0.83 400 650
1.0 400 750
These data support the previous discussion that there is a minimum
of substrate necessary to maintain cell integrity, active transport, and
other mechanical events referred to as maintenance energy. When sub-
strate beyond that needed for maintenance energy is supplied, it is used
for cell growth.
Additionally, in the front end of a plug flow or multiple stage
system, the microorganisms are exposed to a high concentration of sub-
strate compared to a single stage completely mixed unit. They respond
to the increased substrate supply by generating more cell mass. Once
the substrate energy has been committed to new cell mass by the micro-
organisms, most of the commitment is irretrievable and the result is
greater sludge production. In aerobic stabilization of waste sludge
226
-------
where some small reduction in cell mass is observed, probably the mono-
mers and polymers in the cell which are the normal building blocks for
RNA, DNA, enzymes, peptidoglycan, polysaccharide, and lipid formation
that can be salvaged for maintenance energy functions accounts for the
loss in mass.
Preferred ASU reactor design for comparatively weak refinery ef-
fluent treatment is probably two stage with the lead reactor preferably
80+ per cent of the total reactor volume.
Clarifier Design. A clarifier is usually considered to provide for
separation of activated sludge with process design considerations being
chiefly overflow and recycle rates. Actually, this is short sighted,
and a clarifier design should consider in addition to solid flux rates
the elements of: flocculation, capture of solids trapped by surface or
interfacial tension, and a stage of reaction. All of these latter
elements are provided for in a wide-well clarifier design where the
diameter wide well is about one-half the diameter of the clarifier
(Figure 10).
It has been well established that gently flocculating activated
sludge before clarification improves greatly the sludge settling pro-
perties. Flocculation provides for aggregating the fines into the main
floe mass and taking advantage of the biopolymers naturally in the
system. The wide-well zone provides a region to achieve f locculation,
and the improvement in sludge settling more than compensates for the
clarifier surface area reduction due to the wide well.
CHEMICAL DESTABILIZATION REQUIREMENTS FOR OPTIMIZING THE PERFORMANCE OF
DISSOLVED AIR FLOTATION AND GRANULAR MEDIA FILTRATION UNITS
Addressing water chemistry principles determine the phase separa-
tion efficiency of the air flotation and granular media filtration
processes. Fundamentally, maximizing phase separation efficiency by the
air flotation process requires recognizing that an intrinsic property of
solids in the presence of water is a negative, electrical-surface charge
(zeta potential). Flotation air bubbles also have a negative zeta
potential as does the surface of granular media in a filter. Maximizing
phase separation efficiency requires that these coulombic repulsive
forces be controlled by controlling the properties of the dispersed
phase (13).
Air Flotation and Granular Media Filtration Definitions
A proper place to start a discussion of air flotation is a defini-
tion that recognizes the principal limitation of the process; it will
not efficiently achieve the phase separation of colloidal solids. The
definition, therefore, should spell out the application more carefully
than done heretofore:
227
-------
Dissolved air flotation may be defined as clarification of a
suspension of flocculated material by contact with minute'
bubbles that attach to the solids constituting the suspension
causing the suspensions to be separated from the clarified
water by flotation.
The definition puts proper emphasis on the fact that the material
being separated should not be colloidal with its inherent high, repul-
sive, negative zeta potential required to maintain the colloidal state,
but a flocculated suspension. A flocculated suspension implies a proper
chemical pretreatment consistent with the needs of a colloidal system.
Similarly:
Granular media filtration may be defined as clarification
of a suspension of dispersed material by passage through a
bed of porous media that separates and retains within the
media the solids constituting the suspension.
Properties of Suspended Solids
Refinery effluents from aerated lagoons are similar to surface
waters, API separator effluents, fire, and cooling water ponds, etc., in
that the suspended materials usually are predominately colloidal or a
combination of colloidal and very slightly flocculated suspensions. The
stability of these colloidal systems relates to the fact that the indi-
vidual particles carry like electrical charges causing their mutual
repulsion. Except for some isolated examples, the charge on organic,
inorganic, and biocolloids is negative when suspended in water. Col-
loidal destabilization by chemical treatment has the objective of neutra-
lizing or reducing the electrical charge so that mutual repulsion is
reduced to the extent that individual particles can approach each other
close enough for van der Waals and/or chemical forces to become effec-
tive. The attractive van der Waals1 forces cause the particles to
aggregate into agglomerates which facilitate their removal by sedimen-
tation, air flotation, or filtration processes. The surface charge on
colloidal particles may be estimated by electrophoretic, electroosmotic,
streaming, and sedimentation potential techniques.
We have found that the electrophoretic procedures and equipment of
Riddick (14) permits the rapid determination of colloidal charge to be
made and all our investigations involved use of Zeta Meter. Accordingly,
electrokinetic values reported herein are zeta potentials (ZP).
Electrokinetie Charge on Colloidal Particles
Microorganisms, dispersed oil colloids, and inert suspended matter
such as inorganic sulfides, silt, coke fines, etc., that are present in
refinery effluents are negatively charged.
228
-------
The efficiency of air flotation for phase removal depends on the
state of subdivision of the suspended matter; higher capture and removal
efficiencies are achieved with discrete particles of substantial size
and increasingly poorer efficiencies are observed with increasing col-
loidal solids fraction. Investigation revealed that the negatively-
charged flotation microbubble was repulsed by the negatively-charged
colloidal solids such that poor bubble adherence for flotation occurred.
Consequently, the key to maximizing the effectiveness of the air flota-'
tion process was found to be essentially neutralizing or reducing the
electrical charge of the colloidal particles to eliminate the coulombic
repulsion due to like charges.
Source of Charge. The net electrokinetic charge, i.e., zeta poten-
tial, on colloidal particles is a result of (1) ionization, (2) ion
adsorption, and (3) ion dissolution mechanisms.
The amino acids, proteins, and polysaccharides constituting the
surface of biocolloids, for example, acquire their charge mainly through
ionization of functional carboxyl and amino groups to give -COO" and -Nflt
ions. The degree of ionization of these functional groups and, thereby,
the net charge on the particle depends on the pH of the solution. The
pH at which the ZP is zero is called the isoelectric point; at a lower
pH, the ZP is positive, and at a higher pH, it is negative.
Solids dispersed in water typically have negatively-charged surfaces
because cations have a greater tendency to become hydrated and reside in
the aqueous solution than do anions which are smaller, less hydrated,
and more polarized, thereby having the greater tendency to be adsorbed.
The net surface charge, i.e., ZP, may be acquired by the unequal adsorp-
tion of oppositely-charged ions; however, ion adsorption may be positive
or negative.
Ionic substances may acquire a surface charge because the ions of
which they are composed dissolve unequally in solution. In the case of
the aluminum and iron primary coagulants, for example, hydrogen and
hydroxyl ions are in equilibrium with the solid phase hydrous oxide.
With excess hydrogen ions, the surface of the solid phase is positively
charged with excess hydroxyl ion; the surface is negatively charged.
Since the concentrations of the hydrogen and hydroxide ions determine
the chart at the particle surface, they are called potential determining
ions. This concept may be illustrated as follows:
IT
OH"
OH
Fe
OH
229
-------
As early as 1879, Helmholtz (13) described the charge on colloidal
surfaces in terms of an "electric double layer" and, subsequently, Guoy
and Chapman (14) described a diffuse electric double layer model that
permitted more quantitative treatment of electrokinetic data. Verwey
and Overbeek (15) set forth the classic work containing the mathematical
foundation for showing that the tendency of fine particles^ to remain
dispersed was due to the mutual repulsion of their electrical double
layers being sufficient to overcome the van der Waals1 attractive forces
pulling them together. The Russian investigators, Landau and Derjaguin,
had much the same idea and published the electrical double-layer repul-
sion part in 1941. The total treatment is now known as the DLVO Theory
(16).
Electric Double Layer. The electric double layer may be regarded
as consisting of two regions: (a) an inner region which may include
adsorbed ions and (b) a diffuse region in which ions are distributed
according to electrical forces and thermal motion.
Stern (17) proposed a model in which the boundary of the inner
region (Stern layer) was located by a plane (the Stern plane) about a
hydrated ion radius from the surface. Adsorbed ions attached to the
surface by electrostatic and/or van der Waals' forces may be dehydrated
in the direction of the surface. A certain amount of solvent will also
be bound to the charged surface in addition to the adsorbed ions. The
shear plane, therefore, is probably located farther from the surface
than the Stern plane. Ions with centers beyond the Stern plane are
considered to be in the diffuse part of the double layer. These con-
cepts are illustrated in Figure 11.
Electrokinetic potentials relate to the mobile part of the particle,
therefore, the electrokinetic unit consists of the volume enclosed by
the shear plane which is rather inexactly known. The potential differ-
ence between the surface of shear and the solution is called the zeta
potential (ZP).
Colloid Destabilization Mechanisms
Destabilization of the waterborne suspended solids may involve four
mechanisms: (1) colloid entrapment or removal via the sweep floe mecha-
nism, (2) reduction in surface charge by double-layer repression, (3)
charge neutralization by adsorption, and (4) bridging by polymers.
Colloid Entrapment. Colloid entrapment involves chemical treatment
with comparatively massive amounts of primary coagulants; the amount of
coagulant used is typically so great in relation to the amount of col-
loidal matter that the nature of the colloidal material is not relevant.
The amount of primary coagulant used may be 5 to 40 times as much as is
used for charge neutralization by adsorption. The rate at which the
primary coagulants form hydrous metal oxide polymers (Figure 12) is
relatively slow and depends chiefly upon water temperature and pH.
Coupled with the high concentration used, all negatively-charged colloidal
230
-------
material is initially exposed to charge neutralization by the transient
cationic species. The polymer matrix is 3-dimensional and voluminous as
illustrated by Figure 13, providing for entrapment of solids As the
polymer contracts, freeing solvent water molecules and settles the
suspended solids remain enmeshed in the settling floe and appear to be
swept from the water; hence, the description of the process as a "sweep
floe" mechanism. This destabilization mechanism can result in the
generation of large amounts of wet alum (or iron) sludges which are
difficult and costly to dewater. Even though it is by far the most
widely used mechanism for water clarification, it is not recommended
because of the sludge problem and because the use of other mechanisms
result in significantly lower operating and capital costs.
Double-Layer Repression. Reduction in surface charge by double-
layer repression is caused by the presence of an indifferent electrolyte
which in refineries is chiefly sodium chloride from brackish water usage
or salt water ballast. For water and monovalent electrolytes, the
thickness of the double layer is as follows:
Thickness of Double NaCl Concentration Specific Cond.,
Layer, Angstroms M mg/1 Micromhos
1,000 .00001 0.6
600 - 1.0
320 .0001 6
230 - 10 25
100 .001 59 115
75 - 100 200
32 .01 585 1,000
23 - 1,000 1,900
10 .1 - 8,800
7.3 - 10,000 15,000
For double-layer repression of colloid surface charge in brackish
waters, the sodium ions of the indifferent electrolyte which surrounds
the colloid particles in order to electrically balance their negatively-
charged surfaces have less tendency to diffuse away from the colloid
surface as the salinity increases. Some salt concentration may even-
tually be reached such that the thickness of the double layer may be
small enough that two colloids approach each other closely enough that
van der Waals' forces cause aggregation. An important aspect of double-
layer repression is that the quantity of colloidal charge is not signi-
ficantly reduced but just the extent to which it extends out from the
colloid surface. This relates to the nature of the destabilizing chemi-
cal (salt) and its mode of action; i.e., the sodium ions remain free in
the solvent and cause rapid dissipation of the charge as the distance
from the colloid surface increases.
231
-------
The above-mentioned principles can be visualized by the represen-
tations of the double layer in demineralized, fresh, brackish (or salt)
waters illustrated in Figure 14, 15, and 16 respectively.
For demineralized water there are not many ions available; there-
fore, the charge on the particle surface is not reduced much by adsorbed
ions in the Stern layer or counter-ions in the diffuse layers. As a
result, the zeta potential is high and extends for a considerable dis-
tance into the solvent; i.e., the double layer is thick.
Fresh water, on the other hand, comparatively may contain many
salts. As illustrated, the presence of counter-ions may lead to some
adsorption and potential drop across the Stern layer. The concentration
of counter-ions in the diffuse layer is much greater than the demineral-
ized water example and causes the charge to dissipate more rapidly;
i.e., the double layer is much thinner than the demineralized water
example.
In the example for brackish or salt water, the comparatively high
concentrations of sodium ions discourages their diffusion away from the
particle surface. The counter-ions occupying the Stern layer cause an
apparent reduction in potential but are not strongly adsorbed and,
therefore, do not permanently alter the surface potential charge to the
much lower charge of the Stern potential and the charge actually measured,
the zeta potential. The high electrolyte concentration causes any
residual charge to dissipate rapidly; i.e., the double layer is very
thin.
Charge Neutralization. Charge neutralization by adsorption of the
destabilizing chemical to the colloid is a key mechanism for optimizing
removal of waterborne solids from brackish waters by direct filtration.
The colloidal charge may not only be reduced to zero, but beyond zero,
i.e., reversed. Charge neutralization by adsorption infers that the
colloid-water interface is changed and, thereby, its physicochemical
properties. It doesn't require much extension of one's imagination to
see how this destabilization mechanism can explain those cases where
optimal chemical dosages were found and overdosing resulted in a deter-
ioration in, or failure of, direct filtration. This phenomenon is more
typically experienced using very low molecular weight polyelectrolytes
or surfactant-type molecules with little bridging properties. Some
examples of charge neutralization mechanisms are shown schematically in
Figure 17.
Bridging. Bridging by organic and inorganic polymers describes the
destabilization mechanism where the molecules of the added chemical
attach onto two or more colloids causing aggregation. There are two
kinds of bridging; polyelectrolyte bridging between dissimilarly and
similarly charged materials. An example of the first kind is the
bridging of negatively-charged colloids by cationic polyelectrolyte.
Because of the coulombic atraction involved, this destabilization is not
difficult to perceive. On the other hand, weakly anionic organic poly-
mers are negatively charged; however, they are especially useful for
232
-------
aggregating and binding together some negatively-charged aggregates into
agglomerates that resist redispersion. Thus, in this instance, attrac-
tive forces of a chemical nature seem to overcome electrostatic repul-
sion forces due to like charges. An electrical model proposing a
mechanism for this seeming anomaly has been offered by Grutsch and
Mallatt (18). Bridging by polymers proved to be an important destabili-
zation mechanism for optimizing phase removal processes.
DLVO Theory
The DLVO Theory quantifies particle stability in terms of energy
changes when particles approach one another. The total energy is deter-
mined by summation of the attraction (London-van der Waals' forces) and
repulsion (overlapping of electric double layers) energies in terms of
interparticle distance. The general character of the resulting inter-
action energy-distance curve illustrates the very significant conclusions:
1) attraction will predominate at small and large distances and 2)
repulsion may predominate at intermediate distances depending on the
actual values of the two forces. An important purpose of the chemicals
used for destabilization is to reduce or eliminate the repulsion force
at intermediate distances so that attractive forces will predominate and
the particles will aggregate.
These principles are readily illustrated by the interaction energy
curves in Figure 18. The energy of attraction curve (Va) and the energy
repulsion curve (Vr) are summed and yield the interaction energy curve
(Vt). The interaction energy curve shows a repulsive energy maximum
(Vm) which is an energy barrier to coagulation and the formation of
stable aggregates of particles by attainment of interparticle distances
which permit attractive forces at the primary minimum to react. This
example illustrates another characteristic feature of these energy
curves and that is the existence of a secondary minimum at relatively
large interparticle distances. If this secondary minimum is moderately
strong, it can give rise to a loose, easily-reversible form of floccu-
lation. For small particles with a diameter less than about 200 A°, the
secondary minimum cannot achieve this loose, reversible flocculation in
those cases where the energy barrier (Vm) is large enough to prevent
normal coagulation into the primary minimum. Thus, for complete desta-
bilization of systems composed of fine particles, a chemical approach
that responds to zeta potential charge neutralization is required.
A graphic illustration of how zeta potential charge neutralization
leads to destabilization is illustrated in Figure 19. Letting VRJ. be
the energy of repulsion curve for the particle system, summation of VR1
with the energy of attraction (Va) gives the curve for the energy of
interaction (VT1). This curve demonstrates that a repulsion always
exists which accounts for the stability of the original colloidal
system.
233
-------
The effect of adding an increment of chemical which reduces the
zeta potential by charge neutralization is illustrated by summing the
new energy of repulsion curve (Vj^) with Va which yields the energy of
interaction curve VT2« This curve has the secondary minimum which
provides for loosely-flocculated aggregates as previously described for
Figure 18. However, an energy barrier still exists which precludes
flocculation of particles smaller than 200 A° diameter.
Further addition of chemical which lowers the energy of repulsion
curve to that described by VRg yields the interaction curve V^ when
summed with V^. This interaction curve has a large secondary minimum,
no repulsion barrier, thus, all particles are coagulated into aggregates
by the attraction energy of the primary minimum.
Double-layer repression, therefore, can improve solids removal by
direct filtration, but this mechanism does not achieve the best results
and can conceal definition of optimal chemical pretreatment to achieve
best filtration results if the interference of this destabilization
mechanism is not recognized. Our refinery experience indicates that the
colloidal aggregates destabilized by double-layer repression appears
analogous to loose flocculation by the secondary minimum and the aggre-
gates are readily redispersed by hydraulic forces as if the net binding
forces are very weak. A simple procedure to identify the existence of
the double-layer repression mechanism in brackish waters so that it can
be avoided and optimized chemical treatment can be achieved has been
described (19).
Destabilizing Chemicals
Primary Coagulants. Efficient destabilization of colloidal suspen-
sions using salts of iron and aluminum as primary coagulants must recog-
nize the properties of these primary coagulants. The chief properties
of concern are the ZP-pH relationships and hydrolytic reactions.
Stumm and O'Melia (20) describe the equilibrium composition of
solutions in contact with precipitated primary coagulants in the interest-
ing manner shown in Figures 20 and 21. These diagrams are calculated
using constants for solubility and hydrolysis equilibria. The shaded
areas, A and B, we have added in each figure are approximate operating
regions for air flotation and clarifiers by colloid entrapment (region
A) and direct filtration by charge neutralization (region B). Both
regions are assumed to cover a pH range of 6.0 to 8.5. The coagulant
dosage ranges from 33 to 200 mg/1 in region A and 3.3 to 20 mg/1 in
region B. These figures are useful in the interpretation of some of our
filtration and air flotation unit results.
With reference to Figure 21, the isoele^tric point for ferric
hydroxide coincides with the region of minimum solubility, and the
operating regions for water treatment (destabilization) yield a hydro-
lyzed, primary coagulant with a desirable positive zeta potential.
234
-------
In many refinery situations, however, it is difficult to use this
attractive condition because the presence of sulfides and strongly
reducing conditions cause the reduction of ferric to ferrous iron and
the formation of mixed iron sulfides with no coagulation powers. In
fact, in some refinery waters, the use of iron coagulants at modest
dosages may contribute to stabilizing solids rather than destabilizing
them.
While alum has no redox or sulfide chemistry comparable to iron,
its amphoterism and solubility pose definite limitations on alum usage.
With reference to Figure 20, a substantial portion of operating region's
lies in the area where alum is soluble and the predominant equilibrium
Species is negative, Al(OH)^. In the more acidic part of region B,
however, the concentration of equilibrium ionic species is very much
lower and much less negative. Considering these data, it is not unex-
pected that investigators consistently report optimal coagulation/
flocculation results with alum at a pH of 5-6.
With inspection of Figure 20, one may question why alum is effec-
tive at all for neutralizing negatively-charged colloids in the indicated
operating regions. One approach to explaining observed performance
requires understanding that the data are equilibrium data; but before
equilibrium is reached, substantially different conditions exist.
Alum very readily hydrolyzes to form polymers in a complex manner
not well defined. The hydrolytic pathway and reaction rates are affected
by pH, temperature, other ions, etc. One hypothesized route which
includes different aluminum hydrolysis products which are known to exist
is outlined in Figure 12. When alum is added to water in amounts which
exceed the solubility limits, sequential kinetic reactions occur until
the ultimate precipitate is formed and the ionic species appropriate to
the pH equilibrate with the precipitate. The hydrolytic reactions are
formed which are available for colloid adsorption. The hydrolyzed
species have enhanced adsorption capabilities, possibly due to larger
size and less hydration and the presence of coordinated hydroxide groups
(20). In solutions more alkaline than the isoelectric point, the posi-
tively-charged polymers are transient and at equilibrium, anionic poly-
mers prevail.
In modestly alkaline solution, the transient positively-charged
polymers appear to contribute to destabilization of colloids. On the
other hand, in solutions more acidic than the isoelectric point, the
positively-charged polymers prevail at equilibrium and destabilization
of colloids may be achieved at significantly lower coagulant treatment
levels.
In Figure 22, the zeta potentials of colloidal iron hydroxide
solutions are plotted as a function of pH. The zeta potential decreases
in positive charge as the pH increases until the isoelectric point^is
reached at the pH of 8.3 at which the charge reverses. In the vicinity
of the isoelectric point, the charge may vary as indicated. Alum has
235
a
-------
similar zeta potential-pH relationship as shown in Figure 23. The zeta
potential may be negative or positive over the pH range of 7.0 to 7.8.
Salts appear to interfere with the coagulative powers of alum by
anion penetration of the alum polymer by chloride ion. Anion penetra-
tion, the replacement of a coordinated group such as aquo, hydroxo, or
another anion, can be visualized with reference to Figure 12. Chloride
ion is a highly mobile, nonhydrated, electronegative ion that at high
concentrations penetrates the hydrous aluminum oxide polymer, impairs
the olation of alum to polymers, and reduces the formation or charge of
the transient cationic alum polymers which are very important to colloid
charge reduction.
Surfactants. Certain substances, even when present in very low
concentrations, possess the unique property of altering the surface
energy of their solvents to an extreme degree. Almost always, a lower-
ing rather than an increase of the surface energy is affected. Sub-
stances or solutes possessing such properties are known as surface-
active agents or surfactants and their unique effect is known as surface
activity.
By broad definition then, surface-active chemicals are soluble
substances whose presence in solution markedly changes the properties of
the solvent and the surfaces they contact. They are categorized accord-
ing to the manner in which they dissociate or ionize in water and are
characterized structurally by possessing a molecular balance of a long
lipophilic, hydrocarbon "tail" and a polar, hydrophilic "head."
Surfactants owe their physicochemical behavior to their property of
being adsorbed at the interface between liquids and gases (where they
contribute to the electrical charge on the DAF bubble) or liquid and
solid phases (where they may contribute to the zeta potential). Surfac-
tants tend to concentrate in an oriented manner, at the interface, in
such a way that almost entirely, they turn a majority of their hydro-
philic groups toward the more polar phase and a majority of their
lipophilic groups away from the more polar phase and, perhaps, even into
a nonpolar medium. The surface-active molecule or ion, in a sense, acts
as sort of a bridge between two phases and makes any transition between
them less abrupt.
There are three types of chemical surface-active agents which are
classified according to their dissociation characteristics in water.
These are:
1. Anionic Surfactants—Where the electrovalent and polar hydro-
carbon group is part of the negatively-charged ion when the
compound ionizes:
ANIONIC CH3(CH2)16COO~Na+
236
-------
2. Nonionic Surfactante-Where the hydrophilic group is covalent
and polar and which dissolves without ionization:
NONIONIC CH3(CH2)16COO(CH2CH20)XH
3. Cationic Surfactants—Where the electrovalent and polar hydro-
carbon group is part of the positively-charged ion when the
compound ionizes:
CATIOKIC CH3(CH2)17NH3+CL~
Surfactants are powerful charge neutralizers (and charge reversers).
In the petroleum industry, anionic surfactants are used as emulsifiers
for asphalt by imparting a zeta potential on asphalt particles ranging
from -30 to -80 mV. Cationic types impart a zeta potential ranging from
+18 to 128 mV. Each surfactant possesses a distinct characteristic
capability of imparting quantitatively to asphalt during emulsification,
a specific zeta potential.
Surfactants have not found wide use for destabilizing colloidal
systems. In fact, they are an important cause for the existence of
colloidal systems, particularly in primary municipal effluents. The
principal organic colloidal destabilizing chemicals are polyelectrolytes.
Polyelectrolytes. Polyelectrolytes used as water-treating chemi-
cals are macromolecules having many charged groups and may be classified
as cationic, anionic, and nonionic depending upon the residual charge on
the polymer in solution. Examples of the structural types are shown in
Table 1.
In solution, the polyelectrolytes are dissociated into polyvalent
macroions and a large number of small ions of opposite charge (counter
ions). The macroion is highly charged, which is the cause for the
characteristic properties of the polyelectrolytes. Most of the macro-
ions are long, flexible chains, their size and shape depending on the
macroion charge and interaction with counter ions. With increasing
charge, the macroion extends; with decreasing charge, the macroion
assumes a contracted random coil. The source of the charge is illu-
strated by the polyacrylates, a widely-used polymer. In distilled
water, polyacrylic acid's carboxylic functional group is only slightly
dissociated. The addition of NaOH reacts with the carboxylic acid
groups causing them to dissociate leaving a charge on the macroion and
producing sodium counter-ions as shown in Figure 24.
The Dimensions Involved. The dimensions of the various components
involved in colloid destabilization vary a million fold, from a few A°
to more than 10 A° as shown in Table 2.
237
-------
Where color is not a significant factor, the problem is usually one
of causing colloidal particles down to about 1000 A° in diameter to
aggregate. When a clarifier or DAF is used for phase separation, it is
desirable to build aggregates to fairly large size, say greater than 10^
A°. On the other hand, when filters are used, simply destabilizing the
colloidal particles is sufficient because the destabilized particles
will build aggregates in the filter bed as the destabilized suspension
passes through the media and the particles impinge and adhere to the
media or trapped suspended matter.
Systematic Approach to Determining Chemical Treatment Requirement
Broad experience in refinery effluent treatment led to outlining
the condition response schematic for chemical treatment of waterborne
colloids shown in Figure 25. In phase removal by filtration, or even
DAF, we are not concerned with, and indeed it is desirable to avoid, the
use of (1) the "sweep floe" of colloid entrapment and (2) the double-
layer repression mechanisms for colloid destabilization. Destabilization
efforts must focus on the charge neutralization and bridging mechanisms.
Charge neutralization correlated with plant performance as the optimum
destabilization mechanism. For plant control of direct filtration,
charge neutralization has been the key test parameter correlating with
performance of refinery filters. Brackish water required that charge
neutralization be measured after dilution with distilled water to sepa-
rate the effects of double-layer repression and charge neutralization;
i.e., under plant conditions of high salinity, the addition of desta-
bilization chemicals could reduce the ZP to approximately zero by a
range of chemical treatments; however, when double-layer repression was
the cause of reduced ZP, reduced filter run lengths and performance were
observed. Reducing the ZP to approximately zero, as measured by means
responsive to charge neutralization, point out more definitively the
required destabilization chemical treatment and resulted in optimum
filter performance.
Waterborne colloids subject to chemical destabilization and phase
separation fall into two general categories: relatively inert substances
such as clays, sand, and organic materials; and microorganisms or bio-
colloids. Both categories of colloidal matter may be stabilized because
they are charged and/or are highly hydrophilic. Both categories of
colloidal matter also may vary in response to treatment by destabili-
zation chemicals and within each category, the state of subdivision
seems to require additional consideration; i.e., extremely small col-
loidal particles are sometimes more difficult to aggregate for removal
by filtration. Typically, destabilization of biocolloids, such as are
in aerated lagoon effluents, is a more demanding problem.
In the case of polyelectrolytes, some counter ions at high con-
centrations screen the charged functional groups with an ionic cloud as
previously described. Salinity, hydroxide, phenolics, sulfides, etc.,
are examples of the kinds of counter ions found to affect various cat-
ionics. Each waste water application of cationics must address the
contaminants present if the most cost-effective polyelectrolyte is to be
238
-------
used (21).
PROPOSED CHEMICAL AND BIOLOGICAL DESTABILIZATION MODELS
Anionic polyelectrolytes are frequently used to flocculate nega-
tively-charged colloidal systems. Even better acknowledged is the
capability of the activated sludge process to flocculate negatively-
charged sewage and industrial colloidal systems; an extreme case being
the Contact Stabilization variation of the activated sludge process.
Because the activated sludge mass and the anionic polyelectrolytes are
both negatively charged, both of these destabilization examples seem to
be counter to theory. A possible explanation for these extremely valu-
able properties lies in their electrical characteristics and the environ-
ment in which they operate.
Electrical Characteristics of Polyelectrolytes
As discussed previously ionizable groups on the polyelectrolyte are
the source of an electric field. Neglecting the effect of counter ions,
the field about an extended polyelectrolyte is shown qualitatively in
Figure 26. In this example, there are potential maxima in the region of
the charged functional groups. There is a lesser potential field,
outside the region of potential maxima, that might be described as a
"potential tunnel." This is illustrated isometrically in Figure 27.
When the polyelectrolyte is in the random coil conformation as shown in
Figure 28, there is an additional weak potential region (B) in the
polyelectrolyte's sphere of influence. A fourth potential region is the
solution where there is no potential effect due to the polyelectrolyte.
Each potential region has a different effect on counter ions. In
the three potential regions within the polyelectrolyte's sphere of
influence, counter ions can be considered as being bound to the poly-
electrolyte. In the region of potential maxima, the bound counter ion
may be localized at charged functional groups forming ion pairs. In the
potential tunnel region, the bound counter ions are mobile, as they are
in the weak potential region. Mobile counter ions establish an equili-
bria, therefore, between the potential tunnel, weak potential field in
the sphere of influence, and the solution. An especially unique
property of polyelectrolytes ±s that the bound counter ions in the
potentialTunnel area can move parallel _tŁ the polyelectrolyte molecule
in the apparent volume occupied by the potential tunnel; thus, counter
ions can "flow" in the potential tunnel areas of Figures 27 and 28
analogous to water in a garden hose. Polyelectrolyte solutions, there-
fore, show an extremely large dielectric constant.
The dielectric constant or polarizability of polyelectrolytes is
determined by the volume of polyelectrolyte in which counter ions are
retained (not by the charge density of the polyelectrolyte) as long as a
region of bound but mobile counter ions is formed. The dielectric
increment depends on the geometry of the polyelectrolyte; extended
polyelectrolytes give much larger dielectric increments than polyelec-
trolytes in the random coil conformation. Further, the dielectric
239
-------
increment of polyelectrolytes in the extended (or helix) conformation
increases rapidly with increasing length. Uniform and continuous dis-
tribution of charged groups and the mobility of bound counter ions are
essential for large polarizability (22).
Polyelectrolyte Destabilization Model
The means for negatively-charged polyelectrolytes to destabilize
negatively-charged colloids is hypothesized to lie in the electrical
properties of the polyelectrolyte. The proposed model in Figure 29
illustrates, approximately to scale, the colloid, its electrical double
layer, the anionic polyelectrolyte, and its potential tunnel. Counter
ions are cations in both cases.
Cations in the double layer are subject to at least two opposing
electrical forces, coulombic attraction to the colloid particle and,
also, to the solvent to maintain an electrically-balanced system.
When the sphere of influence of the anionic polyelectrolyte approaches
the electrical double layer of the colloid, repulsion due to encountering
like charges might be expected. However, in this instance, apparently:
1. The polyelectrolyte's electrical sphere of influence shields
the counter ions in a localized area of the colloid double
layer from the attractive, electrical solvent forces.
2. The resulting electrical imbalance results in an increase in
the negative coulombic potential in the localized area on the
colloid.
3. The increase in negative coulombic potential attracts the
mobile counter ions from the potential tunnel region of the
polyelectrolyte.
4. The potential tunnel of the polyelectrolyte serves as a conduit
for counter ions which neutralize the surface charge in the
localized area of the colloid surface.
5. The charge neutralization achieved reduces the energy of
repulsion at the localized site sufficiently so that the sum
of the energy of repulsion and energy of attraction curves
yields an interaction energy of attraction at the localized
site (Figure 18).
6. The positive interaction energy at the localized site permit
attractive London-van der Waals' forces at the primary minimum
between the polyelectrolyte and colloid to react.
The key to this hypothetical model is the mechanism by which the
anionic polyelectrolytes implement the initial shielding action of the
colloid surface from the electrical solvent forces. This capability
lies in the special properties of the polyelectrolyte: polarizability,
240
-------
bound but mobile counter ions ,in the potential tunnel, and equilibria
between counter ions in the potential tunnel and sphere of influence
established by coulombic forces.
Repulsion due to like coulombic charge is illustrated in Figure 30.
In Figure 30A, the double layer overlap for energy of repulsion curves A
(polyelectrolyte) and B (colloid) result in a net repulsion indicated by
curve C. If this were a system of lyophobic colloids, it would be a
stable colloidal system. Polyelectrolytes are extremely polarizable,
however, particularly the high molecular weight anionics. The polarizing
polyelectrolyte can supply an essentially infinite quantity of counter
ions via its potential tunnel to a localized site near a colloid particle.
This influx of counter ions reduces the repulsion energy curve A by
double-layer repression because of the equilibria established by the
bound mobile counter ions, and the surfaces can approach more closely as
shown in Figure SOB. The influx of counter ions to the localized charged
region between the polyelectrolyte and colloid surface also would be
expected to reduce the energy of repulsion (curve B) for the colloid in
the localized region. The distance between the charged surfaces can be
diminished further for this additional reason as illustrated in Figure
30C.
Interaction between the two surfaces is achieved at localized sites
if the counter ions reduce the energy of repulsion curves such that the
energy of interaction curve provides for a secondary minimum (see Figure
18). Particles larger than about 200 A° are flocculated. Further
reduction in the energy barrier of the repulsive energy maximum permits
flocculation of the particles less than 200 A°.
Sterically, when the two surfaces approach closely as illustrated
in Figure 30C, the polyelectrolyte provides a shielding or insulating
effect to the localized colloid site from the coulombic solution forces.
Negating the effect of solution forces results in an increase in attrac-
tion of counter ions to be localized colloid site and contributes further
to polyelectrolyte polarization and double-layer charge repression at
the site.
Electrical Characteristics of Microorganisms
The microorganisms constituting activated sludge also have a nega-
tive zeta potential. Glucuronic acid has been proposed as the source of
the electrical charge.
As shown in Figure 31, the outermost surface of most bacteria is a
slimy capsule varying in thickness up to 100,000+ A°. The capsule com-
position varies with species and may consist of polymers of glucose or
other sugars, amino and acid sugars, or polypeptides. Capsules generally
consist of about 98 per cent water. The function of the capsular layer
is proposed as serving the microorganisms as an osmotic barrier, i.e., a
mechanism for guarding against too rapid an influx or efflux of water
(23). We propose that the capsular layer performs an equally if not
more important function of providing the potential tunnel region.
241
-------
The zeta potential is the charge difference between the plane of
shear and the solution. In the case of microorganisms the shear plane
is some small, but indeterminate distance beyond the outer boundary of
the capsule. It is hypothesized that the region beyond the shear plane
is analogous to the diffuse layer of counter ions in the colloid double-
layer model (Figure 11), whereas the capsule on the capsule side of the
shear plane serves a function analogous to the potential tunnel region
of polyelectrolytes; i.e., it provides for a region of bound, but mobile
counter ions. The acidic polysaccharides in the outer layer of the soft
layer are at least partially responsible for the negative, electric
charge on the bacteria surface. As in the polyelectrolyte example,
mobile counter ions establish an equilibria between the capsule layer
(potential tunnel) and the solution and the bound counter ions in the
capsule layer can move parallel to the cell wall of the microorganism.
The chief difference between the polyelectrolyte potential tunnel and
the biological potential tunnel (capsule) is the comparatively large
volume of the capsule-a desirable property.
Biological Destabilization Model
As for polyelectrolytes, the means for negatively-charged micro-
organisms to destabilize negatively-charged colloids also is hypothe-
sized to lie in the electrical properties of the microorganisms. The
model proposed for polyelectrolytes (Figure 29) is analogous to the
model for the biological system. Once again, cations are counter ions
in both cases. Dimensionally, the activated sludge floe and the capsule
volume are larger than the comparative volumes occupied by the poly-
electrolyte and its potential tunnel at the very low concentrations of
polyelectrolyte used (frequently less than 1 mg/1). As the diffuse
counter ion layer of the double layers overlap (Figure 30A), compared to
polyelectrolytes, the massive size of the microorganisms floe more
readily shields the counter ions in the colloid double layer from
coulombic solvent forces. The microorganisms supply counter ions to
satisfy the electrical imbalance from their potential tunnel (capsule).
The localized increase in counter ions reduces the energy of repulsion
curve analogous to anionic polyelectrolytes and the particles flocculate.
The new technology, which is the subject of this article,
forms the basis for a number of U.S. and foreign patent
applications. Williams Brothers Waste Controls, Incorporated,
of Tulsa, Oklahoma, has been licensed to employ this tech-
nology and to sublicense others throughout the world.
242
-------
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, blend-
ing 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 & 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
Chemists, Chief Process Engineering
and Technical Services Superintendent
of the Amoco Oil refinery at Whiting,
Indiana.
243
-------
REFERENCES
1) Grieves, C. G., Stenstrom, M. K., Walk, J. D., and Grutsch, J. F.,
"Effluent Quality Improvement by Powdered Activated Carbon in
Refinery Activated Sludge Process," (Preprint 28-77), API 42nd Mid-
Year Refinery Meeting, May 11, 1977.
2) Crame, L. W., "API Bioenhancement Study," ibid.
3) Grutsch, J. F. and Mallatt, R. C., "A New Perspective on the Role
of the Activated Sludge Process and Ancillary Facilities," First
Open Forum on Management of Petroleum Refinery Wastewaters, USEPA,
API, NPRA, University of Tulsa, Tulsa, Oklahoma.
4) Eckenfelder, W. W., "Activated Sludge Treatment of Petroleum Refinery
Wastewaters-An Overview," ibid.
5) Adams, C. E., Eckenfelder, W. W., and Hovious, J. C., "A Kinetic
Model for Design of Completely Mixed Activated Sludge Treating
Variable Strength Industrial Wastewaters," Water Research, Vol. 9,
pp. 37 (1975).
6) Dickerson, R. E., Geis, I., The Structure and Action of Proteins,
Harper and Row, New York (1969), p. vii.
7) Frobisher, et al, Fundamentals of Microbiology, W. B. Saunders
Company, Philadelphia, Pennsylvania (1974), p. 92.
8) Aiba, S.; Humphrey, A. E., and Millis, N. F., Biochemical Engineer-
ing, Academic Press, New York (1965), pp. 38, 39.
9) Lehninger, A. L., "Energy Transformation in the Cell," Scientific
American, (May, 1960).
10) Heukelekian, H. Oxford, Manganelli, "Factors Affecting the Quantity
of Sludge Production in the Activated Sludge Process," Sewage
and Ind. Wastes. Vol. 23, p. 945 (1951).
11) Walker, L. F., "Hydraulically Controlling Solids Retention Time in
the Activated Sludge Process," Journal WPCF. Vol. 43, No. 1, p. 30
(1971).
12) Lau, C. M., "Staying Aeration for High, Efficient Treatment of
Aromatic Acids Plant Wastewater," 32nd Industrial Waste Conference,
Purdue (1977).
13) Grutsch, J. F^and Mallatt, R. C., "Optimizing Granular Media
Filtration of Refinery Waters Requires that the Electrical Charge
Phenomena Operative be Addressed," Chemical Engineering Progress,
Vol. 73, No. 4, p. 57, (1977).
244
-------
14) Riddick, T. M., Control of Colloid Stability Through Zeta Poteni-ial
Livingston Publishing Company, (1968). " '
15) Verwey, E. J. W. and Overbeek, J. Th. G., Theory of the Stability
of Lyophobic Colloids, Elsevier, Amsterdam, (1948).
16) Overbeek, J. Th. G., Colloid Science, Vol. II, Elsevier, Amsterdam,
1952.
17) Stern, 0., Z. Elektrochem, Vol. 30, p. 508, (1924).
18) Grutsch, J. F. and Mallatt, R. C., "Electrochemistry of Destabili-
zation," Hydrocarbon Processing, Vol. 55, No. 5, p. 221, (1976).
19) Grutsch, J. F., Mallatt, R. C., and Peters, A. W., "Chemical Coagu-
lation/Mixed-Media Filtration of Aerated Lagoon Effluent," EPA-
660/2-75-025, June, 1975.
20) Stumm, W. and O'Melia, C. R., Journal, AWWA, Vol. 60, p. 514,
(1968).
21) Grutsch, J. F. and Mallatt, R. C., "Approach to Chemical Treatment,"
Hydrocarbon Processing. Vol. 55, No. 6, p. 115, (1976).
22) Oosawa, F., Polyelectrolytes, Marcel Dekker, Incorporated, New
York, (1971).
245
-------
DISCUSSION
Milton Beychok, Consulting Engineer; Isn't the optimum sludge age for the
PAC addition related to equilibrium capacity of carbon? Why recycle spent
carbon?
Grutsch; It doesn't seem to work that way, Milton. It seems to work as a
sequential operation with some bioregeneration of the PAC. First, the
biological organisms appear to reduce the TOC/COD to a low level, and second,
the carbon then takes over and reduces the residual TOC/COD to even lower
levels as a direct function of carbon surface area. Bioregeneration of carbon
surface seems to be indicated. For example, at soluble TOC levels of 25 mg/1
in an activated sludge control unit, a parallel unit with 2,500 mg/1 PAC sees
25/2,500 or 0.01 Ib. TOC/lb. PAC at start-up, but at 100 days sludge age the
equilibrium exposure is 1 Ib. TOC/lb. PAC at 24 hours retention time and
increases with dilution rate. To achieve TOC reductions at these loadings in
this concentration range is not expected, therefore, some of the PAC surface
area must be bioregenerated to supply more surface area which, of course,
supplies more driving force for TOC reductions.
Experiments with active carbons with different surface areas indicates
equilibrium carbon surface area in the activated sludge unit determines
effluent quality (at fixed activated sludge unit operating conditions); the
larger the surface area, the lower the concentration of residual substrates.
George Reid, O.U. Professor: When you go to a long-term retention time,
don't you increase the capital and operating cost?
Grutsch: If by "long-term retention time" you mean high sludge age—the
answer for low and moderate strength wastewater typical of refineries is no,
you actually can decrease costs. An extended discussion of the advantages
of high sludge age activated sludge units with their required pretreatment
for reducing capital and operating costs is included in our paper in the
Proceedings of the First Open Forum on Management of Petroleum Refinery
Wastewaters.
Reid: Unless I misunderstood, is there a possibility of using a low sludge
age system?
Grutsch: I think that is the wrong way to go. It is the wrong way to go
because low sludge age systems maximize the conversion of the organic sub-
strate to biological solids (see Fig. 7). You don't want to do that because of
the dewatering and disposal costs. You want to remove them chemically, since
the art is available to remove them chemically. And second, a front end
biological system such as a trickling filter does not remove the troublesome
246
-------
colloidal fraction that ruins the activated sludge electrical properties so that
you can cycle up to high sludge age. I think chemical engineering principles
properly applied here give maximum returns in better effluent quality at lower
operating and capital costs.
Bob Carloni, Lion Oil Company; Do I understand from your discussion that
to flocculate a negatively charged colloid, one should use an anionic polyelec-
trolyte in preference to a cationic?
Grutsch: I don't want you to close your mind to that possibility. For example,
you can use a two-chemical or a one-chemical system. With the proper
facilities, a two-chemical system may be cheaper. Using a two-chemical
system in a municipal plant, the addition of an anionic polyelectrolyte at
0.3 to 0.5 mg/1 to the influent of the primary clarifier dramatically improves
the removal of most of the suspended matter. The residual colloidal material
can then be destabilized more economically by cationic polyelectrolyte charge
neutralization for essentially complete removal in a following DAF or granular
media filter. By contrast, a one-chemical system involves use of only a
cationic polyelectrolyte for charge neutralization and bridging. One-chemical
systems may be more costly in chemical requirement than a two-chemical
system.
John Penniman, Pen Kern, Incorporated: The paper industry, using the
dissolved air flotation process under careful zeta potential control, has
reduced particulate concentration from a save-all to well under 10 ppm.
Bearing in mind that zeta potential can fluctuate in refinery effluent, even
after equalization, from-6mV to-18mV, and that automatic zeta potential
instrumentation is now available, would you speculate on its applicability to
improving the reliability and economics of water clarification.
Grutsch: Well, that is hard for me to put into perspective. We like to follow
the KISS principle; that is, "keep it simple, stupid!" So we don't like to put
too big a demand on the plant personnel. For example, in chemical addition
we would like the plant operator only to look at the chemical addition pump to
determine if the pump is running and pumping, and that is all he has to
concern himself about.
So we use the zeta potential determinations to screen various poly electrolytes
and look for a polyelectrolyte that is insensitive to the system that we are
trying to destabilize. Then we try to pick a treatment concentration where the
suspended solids or pH variation doesn't significantly change the chemical
dosage. There will be sensitive systems that might use feed-back control.
Of course, if the chemical prices increase inordinately, we will have to get a
little bit more conservative in our use and that might pay off a feed-back
control system.
247
-------
Is)
TABLE 1. EXAMPLES OF CATIONIC, NONIONIC AND ANIONIC POLYELECTROLYTES
Polyelectrolyte
Description Structural Type Functional Group fficnole
H
1
Catloolc Anlnes — N — R
R
R
1
Quaternary — H— R
R
0
Nonlonlc Polyamide — C — NHg
Polyalcohol —OH
•(— CHg— CH— )*•
CH,
*•(— CHg— • CH— •)*•
0*0
IH,,
•*• ( — CHg — CH— )*•
OH
R
^^-K
CacO
0
— (— CHg — CH— ) •*
Rydrochlorlde
Poly(n-»ethyl-'*-
ylnyl pyrldlaiuB
chloride)
Polyacrylaalde
Polyrlqylalcohol
Poly(Mth)acryllc
Acid
Polyrliqrlaulfaoat*
I _
-------
TABLE 2. DIMENSIONS INVOLVED IN COLLOID DESTABILIZATION
A. SOME COLLOIDAL SYSTEMS DIAMETER. ANGSTROMS
COLOR BODIES 50 - 1000
INERT COLLOIDS (CLAY, SILT,
INORGANIC SALTS, ETC.) 1,000 - 30,000
EMULSIONS 2,000 - 100,000
BACTERIA 5,000 - 100,000
ALGAE 50,000 - 8,000,000
B. CATIONS
Na+ 1.9
Ca++ 2
C. POLYELECTROLYTES
POTENTIAL TUNNEL 7-11
CHAIN LENGTH, 100,000 - 15,000,000
M.W. 250,000 - 40,000,000
D. ELECTRICAL DOUBLE LAYER
RANGE OF EXPECTED VALUES 5-100
EXPECTED TYPICAL IN REFINERY 30
E. SOLVENT
H20
249
-------
2000
ORaw Waste Load API Separator
• Waste Load after Intermediate Treatment
O
Ui
O
-o
I
o>
O
OJ
X
O
1000
900
800
700
600
500
400
300
CD
O
200
100
I I
I I
0.5 1
10 20 30 40 50 60 70 80
% Probability less than Indicated Value
90 95
98 99 99.5
Fig. 1 Protoablli-ty Plots of COD 'Before and Af-tep In-fcermeaia-te Treatment
-------
300
315"
JO
(J1
200
T3
C
CO
E
c
CD
O)
X
O
15
o
E
0)
CJ
100
50
Before
After
195
53
110—
I I I I I I I I I I I I I
10 50 90 95 98
% Probability Less Than Indicated Value
Fig. 2 Primary Municipal Effluent - Before
And After Filtration (WPCF, V.48 #7, p.1801)
-------
Soluble Substrates
Soluble Substrates
from Actions of
Exoenzymes ,
Cell
mbrane
Cell Wall
I
CO2 and Other
Metabolic By-Products
Fig. 3 Schematic of a Bacterial Cell, Its Biochemical Activities,
and Exoenzyme Solubilization of Insoluble Substrates
252
-------
NO
I
1
1
1
I
1
1
EB ^ 0 Ł13 ^ 0 Ł12 ^
1 i3
i V
1 4
Ec^ 1
T*^" Sl4 -^^™^^^~ Sl5 -^*— •" etc-
\
\
\
1
^^"^x Cell Membrane
1E' JE7
-E* -** ^
L
E« *0
^ ^ 810
! /
" Ł20 Ł22 /
/
/
*l
Cell Membrane ^
Eo
So, Si, 82, etc. — Organic molecules
EA, EB, EC, etc. — Enzyme transport systems
Ł1, Ł2, Ea, etc. — Endoenzymes
Fig. 4 Sequential Endoenzyme Biological Oxidation Showing
Permease Active Transport
-------
Total
BOD/COD
Soluble
BOD/COD
Difficult to
Biodegrade
Readily
Biodegradable
Requires
Acclimation
Colloidal &
Susp. BOD/COD
Fig. 5 Impacts of BOD/COD Components On
Activated Sludge Unit
254
-------
c
0)
i
i
E
(0
0)
c
'w
(0
Ł
u
10
15
Days
20
25
Fig. 6 Biological Removal Of Contaminants With
Time-Types Of Removal Curves
255
-------
to
Ui
13,000 Ibs COD/D
For Refinery A: Alpha
For Refinery B; Alpha
015;°F= 57
03 ; °F = 85
.317; Beta
.3 ; Beta
Ł 2000
0
50 100
Sludge Age (Days)
Fig. 7 Activated Sludge Unit Waste Sludge
Generation and Equilibrium Biomass
Inventory Dependence on Sludge Age
-------
Fraction of Particles
|_ Inert Colloids-]^
Emulsions
Flocculated
.-Sheared
10* 105
Particle Size, Am stroms
10
Fig. 8 Activated Sludge Bimodal Floe Size
Distribution
257
-------
I II III
Zeta Potential, mV
2 5 10 50 90 95
% Probability Equal to, or less than, Indicated Value
Fig. 9 Zeta Potential - Probability Curves for Poor and Good
Settling Activated Sludges
258
-------
Fig. 10 Wide Well Clarifier
Scum
259
-------
0
©
0
©
©
Stern Potential -\
Nernst Potential
Zeta Potential
Bulk of Solution
Particle Surface
Stern (Rigid) Layer—'
Shear Plane
Diffuse Layer of Counter-ions-*
Fig. 11 Representation of Electric Double Layer
According to Stern's Theory
260
-------
[AL(H20)6
3
OH
Ni
[AL(H20)5(OH>]
[(H20){OH) Al
.OH^ OH,
Al
,3,«..,,->, ^ „, Al AI(OH)(H2O}3]
OH^ OH^ OH^
OH,/ ^OH
HH20)2(OH)2AI ' lAlx. lAI(OH)3(H20)]
OH"\ OH
[(H20)2(OH)2AI
.
OH
AI(OH)3(H20)]
•°H< ^°"\
l(H20}(OH)3M -^ ( Al JAI(OH)3(H20)]
"7.
[(H20)4AI AI(H20)4]
+4
OH"
[For solid phase see isometric in following figure]
OH"
[(H20)3«OH)AI ^AI(OH)2(H20)2
OH
Fig.12 Sequential Formation of Hydrous Aluminum Oxide Polymers
-------
o
NJ
O>
N>
H20
OH2
Fig. 13 Example of Complex which may exist in Precipitated
Hydrous Aluminum Oxide Polymers
-------
Particle Surface
Stern Plane
Surface of Shear
Diffuse Layer
Stern Layer
Surface Potentia
Stern Potential
Zeta Potential
Distance
Fig. 14 Representation of Electric Double
Layer in Demineralized Water
263
-------
Particle Surface
Stern Plane
^Surface of Shear
+
$
10
© dPn
ooTo;;
00^00
p
L ,
©0Gp°
1
>i
i
" •
Diffuse Layer
t '
Stern Layer
ourrace roiennai
• .
1
l
I
1
1
1
l
1
Q+Arrt Do+^n+iol
-Zeta Potential
l
Distance
Fig. 16 Representation of Electric Double Layer
in Brackish Water
264
-------
'article Surface
pStern Plane
\ ^Surface of Shear
© & ©
•"^0eo
©
o
Diffuse
Stern
Surface Potential
Distance
O
©
Fig. 15 Representation of Electric
Double Layer in Fresh Water
265
-------
No Specific Adsorption
Specific Adsorption of
Cations
Specific Adsorption of
Anions
Strong Specific Adsorption
of Cations
A Rigid Layer Migrating with Particle as Single Kinetic Unit
B Slipping or Frictional Boundary Layer of Ions
X Distance from an Arbitrary Point Inside Solid Phase
i// Electrical Potential
f Zeta or Electrophoretic Potential
Fig. 17 Distribution of Ions and Potentials in the Double Layer
Surrounding Colloids (After Overbeek 8)
266
-------
Primary
.Minimum
Secondary
Minimum
VA - Energy of Attraction
V - Energy of Repulsion
R
VT - Energy of Interaction
VM - Energy Barrier
D - Distance Between
Charged Surfaces
Fig. 18 Potential Energy of Interaction
for Two Charged Surfaces
267
-------
VA - Energy of Attraction
VR1" VR2' VR3' Decreasing
Energies of Repulsion
VT1- VT2-VT3- Interaction
Energies for VR1, VR2 & VR3
D - Distance Between
Charged Surfaces
Fig. 19 Influence of Electrolytes
on Interaction Energy
268
-------
NJ
O>
VO
O
-2
-4
-6
o
D)
O
-8
-10
-12
AI(OH)3 (s)
///>'//*
\\VtK\\
i\
\^\
M\ J-.V-
A - Operating Region for Air
Flotation and Clarif iers
B - Operating Region for
Direct Filtration
C - AI(OH)~
D - AI13(OH)|+
A-L.
E - Al
/
3+
pH
8
10
12
Fig.
20 Equilibrium Compositions of Solutions in Contact
withAI(OH)3
-------
N)
-~J
O
u
-2
1 -6
-8
10
-12
A - Operating Region for Air Flotation and Clarif iers
B - Operating Region for Direct Filtration
C-Fe(OH)4 F-Fe3+
D - Fe(OH)2 Q . pe
E - Fe(OH)2+
\\\A\\\
V//.6///A
Fe(OH)o (s)
\G
\ \ D
6 PH 8
10
i i
12
Fig. 21 Equilibrium Compositions of Solutions in Contact
with Fe(OH)3
-------
bo
+40
+30
+20
>
J +10
.2
'+5
Q)
4-»
o
Q_
(0
+j
CD
N
0
-10
-20
-30
-40
Data Scatter
I I I I
5 6
9 10
Iso Electric Point pH = 8.3
Fig. 22 Zeta Potential of Colloidal Iron Hydroxide
Solutions Plotted As A Function of pH
-------
+10
0
-10
I I I I
I
3456789 10
pH
Fig. 23 Zeta Potential — pH Plot for Aluminum
Hydroxide
272
-------
— C — C— + NaOH-*—C —
N>
•xl
OJ
H
1
— C—
1
C
0
H
1
C —
1
C
/\
r \
o
1
H
H
1
C
1
H
H
1
— C —
1
C
//\
H \
O O
1
H
LJX
H
1
— C
1
H
H
1
— C —
1
C
/A
0 O
e
X*— N.
(Na+)
H
1
C
1
H
H
O O
e
Fig. 24 Dissociation Of Polyacrylic Acid
By NaOH
-------
Waterborne
Suspended Solids
1) Charge
Neutralization
2) Bridging
1) "Sweep Floe"
Mechanism
2) Double Layer
Repression
Inert
Solids
Biological
Cell Material
Cold Water
Warm Water
Fresh Water
Brackish Water
Fresh Water
Brackish Water
Fig. 25 Condition—Response Flow Schematic for
Chemical Treatment of Waterborne Colloids
274
-------
Potential Along Dashed Profile
Extended Macroion
A B
Potential
A = Solution Potential; B = Weak Potential Field;
C =• Potential Tunnel; and D = Potential Maxima
Fig. 26 An Extended Macroion and Its Associated Potential Field
275
-------
ho
~J
CT>
Charged Functional
Groups on Macroion
Potential
Maxima
Potential
Tunnel
Fig. 27 Isometric Showing Potential Maxima and Potential Tunnel Fields
-------
A = No Potential Effect; i.e. the Solution
B = Weak Potential in the Macroion's Sphere
of Influence
C « Potential Tunnel
D * Potential Maxima Near Charged Functional
Groups
28 Macroion in Random Coil Conformation Showing Four Potential Regions
277
-------
Electrical Double Layer
Counter Ions
Shear Plane
Stern Layer
30A°
A = Anionic Polyelectrolyte Showing Potential Tunnel
and Mobile Counter Ions
B = Polyelectrolyte Flooding Double Layer with
Counter Ions
C = Chemically Bound Polyelectrolyte
Fie 29 Electrical Model for Destabilization of Negative
Colloid with Anionic Polyelectrolyte
278
-------
Potential of
Polyelectrolyte
or
Microorganism
Potential
Tunnel Surface
A. Stable Colloid System
B
Potential of
Colloid
Surface
• Distance Between Surfaces
B. Polarizing Polyelectrolyte
or Bacteria Systems
C. Double Layer
Repressions By
Influx Of Counter-ions
A = Energy of Repulsion Curve
for Polyelectrolyte or
Microorganisms
B = Energy of Repulsion Curve
for Colloid
C = Summed Energy of
Repulsion
30 Polarization Affects Distance Between Surfaces
279
-------
Membrane
Periplasmic Space
Rigid Layer
:Soft Layer
Capsule-
X "Increasing Negative Potential
Zeta
Potential —
Capsule
"Potential Tunnel'
Distance-
Fig. 31 Activated Sludge Floe Potential Model
280
-------
"CONSIDERATIONS IN REUSE OF REFINERY WASTEWATER"
Robert W. Griffin
Senior Staff Consultant
Cyrus Win. Rice Division, NUS Corporation
Paul Goldstein
General Manager, Cyrus Wm. Rice Division
Vice President, NUS Corporation
ABSTRACT
Refineries typically recover all usable heat at process units; the waste
heat dissipated by evaporative cooling towers and air coolers is of such
low quality that recovery historically has been neither economic nor prac-
ticable. Refinery effluent water reuse, however, presents a unique oppor-
tunity for utilization of this waste heat for the reduction of wastewater
volumes because the process technology used for evaporation can use this
low grade waste heat as an energy source. Increasing present day and future
energy costs increases the attractiveness of the approach for those situa-
tions where it is needed. Results from two successful experimental programs
conducted as part of studies involving refinery wastewater reuse are dis-
cussed.
DISCUSSION
In 1973, NUS Corporation initiated a project for the API to develop methods
of water reuse that could be employed to reduce wastewater flows from grass
roots oil refineries. The studies were undertaken with the premise that all
characteristic refinery water use practices and patterns would be considered
so that comprehensive water management programs could be evolved. In
essence, the multiplicity of refinery water and wastewater streams were
viewed as parts of a single system. This approach permitted the development
of bases for defining optimum approaches to water use and reuse.
The advantage of studying a hypothetical grass roots refinery was the freedom
to employ any practical water use patterns and add treating equipment to
alter water characteristics without the necessity of considering backfit
penalties.
The model refinery was established with all of the normal unit operations
of class D refineries including the following:
281
-------
1. Atmospheric and vacuum distillation
2. Hydrocracking
3. FCC
4. Sulfuric acid alkylation
5. Catalytic reforming
6. Coking
7. Lube processing
8. Gas recovery
9- Gasoline sweetening
10. Desulfurization
11. Sulfur recovery
12. BTX
13. Gasoline blending
During the course of these studies, operating refineries for the most part
already had BPT end-of-pipe treatment systems in place or under construction
to meet the 1977 compliance date. This specific project was initiated look-
ing ahead toward the probability of more stringent controls for 1983. It
was recognized that additional processes to further reduce pollutants might
be necessary and that treatment of large wastewater volumes would probably
be exceedingly costly. On this basis effluent reduction could effect a dual
benefit, i.e., the improvement in treatment efficiency of current systems
and the reduction in future capital outlays and operating costs for 1983
BAT compliance.
The model refinery used for the studies had wastewater flows and qualities
typical of a conventional 150,000 bbl/day class D refinery as a base case
situation. The studies considered three different water supplies among many
variables. Included was a low solids water typical of the Gulf Coast area,
an intermediate solids water similar to Mississippi River water and a high
bicarbonate alkalinity water similar to Lake Michigan.
A complete water balance was developed for each supply with resulting dis-
charge volumes from 31 to 36 gallons per barrel of crude for the base con-
ditions. In addition the following utilities requirements were assumed for
the studies:
Steam Generation 20,000,000 Ib/day
Cooling Load 50 x 109 BTU/day
Electrical Load 681,500 KW/hr/day
Pretreatment methodology and recycle schemes developed in the study showed
that reduced effluent volumes of 7 to 11 gallons per barrel of crude were
possible.
With this smaller volume of effluent as a basis for further investigations,
two additional processes for further effluent reduction were applied. Both
systems had the advantage of utilizing waste heat from the refinery.
The concepts employed to reduce flow from 7 to 11 gallons per barrel to
lower values included a concentrating cooling tower and a brine concentrator
as shown schematically in Figure 1. Figure 1 also shows a non-specific
282
-------
waste treatment plant upstream of the above referenced equipment. Deter-
mining the extent of waste treatment necessary in order to use wastewater as
makeup to either or both systems was an important objective of the studies.
Concentrating Cooling Tower Tests
A cooling tower for concentrating wastewater could theoretically be used to
reduce effluent volumes by evaporating water up to the solubility limit of
the least soluble inorganic salt present. The concept was considered to be
attractive since the capital cost of such a tower was considerably less per
unit of water evaporated than other alternatives (such as a brine concen-
trator) and only slightly more expensive than normal cooling tower capacity.
The cooling capacity of a concentrating tower would be substituted for other
refinery cooling and the operating cost (electrical power required) would be
essentially a trade-off.
In order to examine the feasibility of such a system a pilot plant was tested
using refinery wastewater as makeup. During the operation of the pilot plant,
scaling and corrosion data was of primary importance as was the fate of or-
ganics and other volatile compounds introduced in the makeup water.
The cooling tower selected for the tests was a coil shed type as depicted in
Figure 2. It was in essence a normal induced draft tower with distributor
nozzles in the bottom of an elevated collecting basin which distributed water
to underlying cooling coils. The heat load to the tower was steam at 35 psig
which admitted to the coils condensed and subcooled.
The pilot plant tests were conducted at a refinery. Extensive analyses of
the wastewater to be used as makeup were conducted. Inorganic analyses
focused on the phosphate, carbonate, sulfate, silica content which with cal-
cium could form scale. Suspended solids was also considered as a possible
depositing material. It was determined from the analyses that phosphate and
silica would not be significant scaling materials and that calcium carbonate
scale could be avoided by careful pH control using acid. It was simulta-
neously determined that calcium sulfate and suspended solids would be the
likely potential sources of deposits.
During test runs suspended solids were concentrated to values exceeding
500 ppm in the tower without deposition. This left the deposition of calcium
sulfate of primary concern.
Figure 3 shows the solubility limit of calcium sulfate as a function of tem-
perature. During tests, the bulk water temperature was 105°F as indicated
by Point No. 1. This according to literature should have been the maximum
concentration of calcium sulfate attainable without scaling. Point No. 2
corresponds to a skin temperature of 200°F. Point No. 3 indicates the solu-
bility limit at the skin temperature of the coil inlet equal to a 285°F
steam temperature. Scaling actually occurred at Point No. 4 conditions as
a result of the film boiling. Scaling started at the steam inlet to the
coil and progressed across the first tubes.
283
-------
Although the coil shed tower was selected primarily for accessibility of the
heat exchangers, the direct cooling of process fluids in the tower limits
the usefulness of the concept and an intermediate cooling loop to eliminate
the occurrence of film boiling as well as to reduce skin temperature is re-
quired to achieve higher concentrations of dissolved salts and thereby ex-
tend the usefulness of the concept.
Figure 4 shows by arrows the conditions in the test program where crystalline
deposits developed. It was determined that a short chain polyacrylate with
acid pH control utilizing hydrochloric for Test 7 and sulfuric for Tests 9
and 10 was effective in suppressing crystalline deposit formation on the
tubes. It was therefore concluded that these materials would be very effec-
tive in assuring reliable operation of a system where calcium sulfate was
controlling. With proper design, system pH control and with the short chain
polyacrylate as an additive, it is realistic to conclude that calcium sul-
fate concentrations of 2500 ppm could have been tolerated without crystalline
deposit formation. During Tests 7, 9 and 10, only a light powdery film
developed in insufficient quantities to sample, pie material could be re-
moved easily by water sprays.
The full test program encompassed evaluations of the following chemical con-
trol conditions:
1. No chemical addition
2. Sulfuric acid for pH control, pH 7.1-7.2
3. Repeat of No. 2, pH 6.8-7.1
4. Organic phosphate (100 ppm), sulfuric acid, pH below 8.0
5. Organic phosphate - polyacrylate (100 ppm)
6. Combination short & long chain polyacrylate (100 ppm) with
sulfuric acid pH control
7. Short chain polyacrylate (100 ppm), HC1 pH control
8. HC1 alone, pH 6.9-7.1
9. Short chain polyacrylate, repeat of Test No. 7, with
sulfuric acid
10. Short chain polyacrylate, repeat of Test No. 9
During pilot plant operation, data was collected to determine the fate of
the organics present in makeup water. Test data is included in Table 1.
The concentration of both the phenols and ammonia in the recycle water were
lower than those in the makeup water.
Data was collected under both sterile and unsterilized conditions with the
same results. The reduction in total carbon, ammonia and phenols was
284
-------
therefore attributed to air stripping in the tower and not due to biological
oxidation. Consequently, it was further concluded that thorough and effec-
tive treatment of refinery wastewater makeup is necessary to prevent air
emissions due to stripping. Conventional wastewater treatment with oil
separation, equalization, air flotation or filtration, biological oxidation
and final solids removal would be desirable.
Brine Concentrator Tests
A survey of refinery wastewater conditions indicated that a brine concentra-
tor could be economically employed for effluent reduction if the stream to
be treated was of low volume and if the device would operate successfully
on refinery wastewater without scaling and produce water of condensate
quality which could be used as a substitute for boiler makeup water. A small
pilot evaporator using the calcium sulfate seed slurry scale control process
was operated on a Texas refinery wastewater. A series of four tests were
conducted during which the volume of evaporator blowdown was successfully
reduced to 1% of the evaporator makeup water. Soluble salt concentrations
in recirculating brine reached 300,000 ppm and suspended solids reached
80,000 ppm during the tests without deposition on the heat transfer surfaces.
Oil in the makeup water to the evaporator ranged from 1 to 3 ppm but caused
no problem.
No corrosion of the 316 stainless steel evaporator occurred except during
an early run when the feedwater was poorly deaerated which in conjunction
with iron concentrations of approximately 30 ppm led to slight pitting of
the stainless steel heat transfer surfaces.
The distillate water produced by the brine concentrator contained 60% to
70% of the phenolic materials, 25% of the freon soluble oils, 25% to 50%
of the ammonia and 50% to 60% of the total organic carbon introduced in
the feedwater. It was therefore concluded that if the distillate was to
be employed as boiler feedwater, conventional wastewater treatment processes
would be required upstream of the evaporator.
In summary, it was determined that with the refinery wastewaters tested,
both the concentrating cooling tower and brine concentrator are viable
methods for reducing wastewater flow.
The concentrating cooling tower enjoys the universal advantage of utilizing
waste heat normally rejected in conventional cooling towers. The brine
concentrator finds its most economical application in plants where low
pressure steam that would otherwise be vented is available as an energy
source. However, in some situations where the value of the recovered water
for boiler use is great enough based upon alternative raw water treatment
costs, operation on the vapor compression cycle using electric energy input
"lay be justified.
285
-------
BIOGRAPHY Paul Goldstein
Paul Goldstein is Vice President and
General Manager of the Cyrus Wm. Rice Division
of NUS Corporation. He holds a B.S. degree in
Marine Engineering from the United States Mer-
chant Marine Academy. Prior experience in-
cludes techanical management with Combustion
Engineering, Inc. and Senior Research Engineer
with Foster Wheeler Corporation.
Robert W. Griffin received a B.S.
degree in Chemical Engineering from Grove
City College. He is currently Manager of
Special Projects fro NUS Corporation, Pitts-
burgh, Pennsylvania. He has been with NUS
Corporation since graduation in 1950. Mr.
Griffin is also a Consulting Engineer.
286
-------
PAPER DISCUSSION
Question - Bill Ruggles, Phillips Petroleum Company
What corrosion rates were observed during use of highly concentrated waste-
water during cooling tower tests?
Answer
Each test conducted was of short duration making it impractical to collect
meaningful corrosion data. We were concentrating on the scaling data and
once the tubes were scaled it was necessary to lower pH dramatically to
remove the deposit. The coil was made with four passes and the tubes were
two rows of stainless, one row each of Admiralty and 90-10 cupro-nickel.
After the test runs there was no evidence of corrosion on the tubes either
generalized or pitting. We would expect corrosion rates on carbon steel
to be similar to those experienced on brackish waters where somewhat higher
inhibitor concentrations are required.
Question - Jeffrey Chen, Dravo Corporation
What do you do with the concentrated brine?
Answer
There is no specific answer to the problem of brine disposal, however, it
is somewhat site specific. The brine concentrator decreased effluent volume
to one or two gallons of water per barrel of crude processed. It could be
decreased further by a drier. Of the total dry salt produced over 50% came
from ballast water (150,000 gal/day) which was included in the study and
represented a salt on a dry basis of 15 to 20 tons per day. Answers to
the question depending on the location could be deep well disposal, ocean
disposal, on-land in arid regions or possibly reprocessing.
287
-------
TABLE 1
ANALYSES DURING TEST RUN
Date
CARBON
2/6
2/17
2/20
2/26
3/1
3/5
3/10
3/19
3/26
AMMONIA
2/6
2/17
2/20
2/26
3/1
3/5
3/10
3/19
3/26
PHENOLS
2/6
2/17
2/20
2/26
3/1
3/5
3/10
3/19
3/26
Makeup Water
92
79
81
69
60
74
69
60
68
3.7
5.4
6.1
8.0
1.9
2.6
3.2
3.3
0.2
0.26
0.20
0.254
0.299
0.330
0.495
Recycle Water
185
374
458
278
326
461
323
380
300
4.8
13.0
0.84
1.3
0.4
0.82
3.0
1.2
2.9
0.1
0.25
0.26
0.092
0.229
0.239
0.02
0.286
0.411
Calculated From
Cycles
506
671.
753,
607.
531,
488.
828
834
595
20.35
45.9
56.73
45.75
14.63
31.2
44.5
28.9
1.1
2.21
1.86
1.5
1.97
4.59
4.33
288
-------
FIGURE I
PROPOSED EQUIPMENT USE
REFINERY
to
oo
CONDENSATE
OIL
SEPARATOR
WASTE
TREATMENT!
I J
BLOWDOWNl^-cb
TO WASTEJ f
-------
K>
FIGURE 2
CONCENTRATING COOLING TOWER
-------
25OO-
FIGURE 3
THEORETICAL AND ACTUAL
CALCIUM SULFATE SOLUBILITY VS.TEMPERATURE
FOR CONCENTRATING COOLING TOWER DEMONSTRATION
ANTICIPATED AT BULK WATER TEMP 105 °F
PRODUCT. INLET TEMPERATURE 200°F
STEAM INLET TEMPERATURE 35ps!g 280 °F
0 ACTUAL SOLUBILITY 280°F DUE TO FILM BOILING
2000-
Q-
1500-
10
\D
u.
_J
±>'
en
O
_J-
<
O
1000-
500-
100
200
300
400
TEMPERATURE (°F)
-------
IO
FIGURE 4
PERORMANCE DATA CONCENTRATING COOLING TOWER TEST
-»• POINT OF CRYSTALLINE SCALE FORMATION
CALCIUM CARBONATE AND CALCIUM SULFATE
2500-
10
O
o
o
0 2000-
o.
Q.
co 1500
UJ
z
o
cc
<
X
-------
STATE-OF-THE-ART IN SOUR WATER STRIPPING
M. R. Beychok
Consulting Engineer
Sour water effluents from refining and petrochemical plants
originate primarily from the use and subsequent condensation of
process steam. The condensation usually occurs in the presence
of a hydrocarbon vapor phase containing various amounts of NH3
and H2S. Thus the condensed steam often contains NH3 and H2S in
amounts ranging from 1,000 to 10,000 ppm which imparts the un-
pleasant odor characteristic of sour waters. Some sour waters,
particularly from hydrocrackers, may contain as much as 30,000
to 50,000 ppm of NHs and H2S.
Sour waters may also contain significant amounts of CX>2>
phenols, cyanides, fatty acids and other contaminants. Fortun-
ately, the principal contaminants, NH3 and H9S, can be removed
by relatively simple steam distillation (stripping). Tradition-
ally, the refining and petrochemical industries have stripped
NH3 and H2S from their sour waters by steam distillation at
5-10 psig and 230-240 °F. The stripping steam is either injected
directly into the distillation tower or generated in reboilers.
A systematic study of sour waters and a tray-by-tray design
method for sour water strippers was first published in 1967(1).
At that time, the typical refinery sour water stripper involved:
• About 8-10 trays
• A stripping steam rate of about 0.8 pounds of steam per
gallon of raw feed (Ibs/gal RF)
• 69 % (or more) average NH3 removal
• 95 % (or more) average H2S removal
• Tray efficiencies of about 40-50 %
In 1972, the American Petroleum Institute (API) undertook a de-
tailed survey of sour water strippers(2). The results of that
survey are summarized in Table 1 herein. Briefly, the 1972
survey indicated that the average sour water stripper involved:
• 15 trays
• A stripping steam rate of about 0.8 Ibs/gal RF
« 78 % NH3 removal
• 96 % H2S removal
« Tray efficiencies of about 45 %
293
-------
Clearly, very little had changed in the design or perform-
ance of sour water strippers between 1967 and 1972 other than a
trend toward using more trays (15 rather than 10, as an average).
CURRENT DESIGN REQUIREMENT
In recent years, a number of 25 to 30 tray sour water strip-
pers have been built. This trend toward a dramatic increase in
trays has been necessitated by ever more stringent environmental
regulations on the NHa content of discharged effluent waters.
Whereas the traditional designs of sour water strippers had emph-
asized H2S removals, environmental regulations now make it nec-
essary that strippers be designed primarily for NH3 removal.
As an order of magnitude, Table 2 illustrates that a typical
125,000 BSD refinery (within the EPA's cracking category) may
require sour water stripping down to a level of 25-60 ppm of NHs
to meet the EPA's refinery effluent guidelines for Best Available
Technology Economically Achievable (BATEA). That level of 25-60
ppm of NH3 in the stripped water is based on these criteria:
— The BATEA 30-day average guidelines for BOD and NH3
translated to annual averages using the EPA's variability
factors for BOD and NHa
— Assuming the 125,000 BSD cracking re-finery has a sour
water rate of 300-350 gpm
— Assuming that 50-100 % of the BOD is removed in a bio-
treater (operating at 90 % efficiency) which consumes
4 pounds of Nitrogen per 100 pounds of BOD removed
— Assuming that 70 % of the nitrogen entering the bio-
treater comes from stripped sour water
Given a sour water containing 7,500 ppm of NH3, it will
require 99.2 to 99.7 % NH3 removal to achieve a stripped water
NH3 level of 25-60 ppm. Regardless of the precise accuracy of
the illustrative case in Table 2, it is fairly obvious that the
current design requirement for NH3 removal in sour water strip-
pers should be at least 99 % and perhaps in excess of 99.5 %.
Achieving such NH3 removals requires more trays and/or steam as
compared to the average stripper in the 1972 API survey. In the
current era of high fuel costs and emphasis on energy conserva-
tion, it is important that the design engineer carefully evaluate
the tradeoff between trays and stripping steam "in achieving a
desired NH3 removal. Figure 1 is from a recent publicationt3) of
a stripper design study, and it illustrates the tradeoff between
incremental trays and incremental stripping steam.. It shows that
99.4 % NH3 removal could be achieved, for the specific study, by
either of the combinations of equilibrium stages and stripping
steam listed below (from which we can obtain the tradeoff between
incremental trays and incremental steam):
294
-------
Stages Stripping Steam
Equilibrium Actual Ibs/gal RF Ibs/qal TF
9.3 21 1.1 i.o
6.0 13 1.8 1.5
Thus an increment of 8 actual trays could replace an increment
of 0.7 Ibs of steam/gal RF. For a stripper raw feed rate of 350
gpm, that amounts to 353,000 Ibs/day of steam savings as the
tradeoff against using 8 more trays. The savings are in fact
even larger since the higher steam rate would increase the size
of the stripper tower, reboiler and overhead condenser. In this
specific case, it is obvious that the correct design choice would
be to use more trays rather than more stripping steam.
RECENT AND ONGOING RESEARCH ON SOUR WATERS
After completing the stripper survey in 1972, the API re-
tained the Bechtel Corporation to:
• Determine if phenols or cyanides affected the stripping
of synthetic sour waters in a bench-scale stripper
• Determine if actual refinery sour waters stripped in
the same manner as did synthetic sour waters
• Evaluate the validity of Van Krevelen's vapor-liquid
equilibrium (VLB) data for the NH3-H2S-H20 system
The results of Bechtel's work, have been published by the API(4)
and summarized in an excellent paper by Gantz(5). Briefly,
Bechtel found that:
— Fresh synthetic solutions of NH3 and H2$, as well as
actual refinery sour waters, both exhibited rapid and
pronounced oxidation. This resulted in a "fixed" amount
of NH3 residual in stripped waters which could not be
removed by intense stripping or batch boiling. Similar
results had been reported earlier by Dobrzanski and
Thompson(6).
— Fresh synthetic solutions of NHa and H2S (protected from
oxidation) could readily be stripped to very low levels
of both components. Levels of 10-15 ppm NH3 and 0-5 ppm
H2S were achieved with 5-10 bench-scale trays and strip-
ping rates of 1.0-1.8 Ibs/gal RF. (The 5-10 bench-scale
trays were probably equivalent to 10-20 plant scale
trays).
— The addition of as much as 800 ppm phenols and 120 ppm
cyanide did not affect the ability to strip NH3 and H2S
from the fresh synthetic sour waters.
— Actual refinery sour waters all contained a varying
amount of "fixed" residual NH3 after stripping which
was attributed to either oxidation or some unknown acid-
ic compounds present in the sour waters.
— Caustic injection into the stripper could be used_quite
effectively to release the "fixed" NH3 and to achieve
295
-------
low levels of NH3 in stripped waters.
— The Van Krevelen VLB data were valid for synthetic and
actual refinery sour waters, if proper allowance was made
for any "fixed" NHa residual as determined by a batch
boiling test.
Following Bechtel's work, the API and the EPA jointly funded
Stanford Research Institute (SRI) during 1976-1977 to:
• Determine what acidic material were present in sour water
and if there were other causes for NH3 fixation.
• Study caustic injection strategies.
• Determine the behavior of cyanides during stripping.
The results of SRI's work have been described by Bomberger and
SmithC"7). A final report for publication is in preparation.
Briefly, SRI found that:
— There were numerous problems associated with the standard
analytical procedures when applied to sour waters. In
particular, the NH3 determination procedures need much
further developmental work.
— The analytical procedures for cyanide were inadequate and
the cyanide stripping studies were therefore not con-
cluded.
— Heavy metal contents were so low in the stripped waters
that ammonia-metal complexes were eliminated as a cause
of NH3 fixation.
— Most of the refinery sour water samples had significant
amounts of oxidized sulfur compounds, and the oxidation
had occurred in the refineries.
— 12-40 % of the organics in the refinery sour waters were
phenol and cresols.
— NH3 fixation was caused by weak sulfidic acids, weak
organic acids and strong sulfidic acidsi
Refinery
B
C
D
F
G
H
— The optimum caustic injection strategy for sour water
stripping was single-point injection in the tower feed.
This was more effective than single-point injection at
the tower middle or bottom, and more effective than
multiple point injection.
— The optimum caustic injection strategy (in the tower
feed) freed practically all of the fixed NH3 and did not
interfere with H2S removal.
296
Weak acids
(meq/1)
<2.0
<2.0
2.9
3.9
<2.0
<2.0
Strong acids
(meq/1 )
___
4.8
0.5
1.0
_ —
3.4
Total acid
(meq/1 )
2.0
6.8
3.4
4.9
2.0
5.4
Fixed NHs
(meq/1 )
1.0
11.0
2.1
6.2
2.1
4.2
-------
The API has also funded work by,Dr. Grant Wilson at the
Thermochemical Institute of Brigham Young University (BYU). The
objectives of that program were:
• To obtain new and additional VLB data for the NH3-H2S-H70
system. Since Van Krevelen's work was essentially at
atmospheric pressure and 70-140 °F, BYU was to obtain
data ranging from 150-250 psia and 175-250 °F. Whereas
Van Krevelen's VLB data could only be extrapolated and
used for NHa/J^S molar ratios above 1.5, the BYU results
would provide VLB data applicable at any NH3/H2S molar
ratio.
• To provide a correlation that would calculate system pH's
as well as VLB data.
• To extend the VLB data base to include other species such
as mercaptans, cyanides, phenols, C02 and others.
The BYU results have been published in a series of draft
reports, the latest of which is dated May 1977(8). ^ finai re_
port is in preparation. The BYU results are summarized below:
— New VLB data have been obtained as follows:
System Data points
H2S-H20 9
HCN-H20 8
C2H5SH-H20 6
NH3-H2S-HCN-H20 4
NH3-H2S-H20-Phenol 6
NH3-H2S-H20-Xylenol 6
NH3-H2S-H20 18*
(* 176-248 °F and 15-242 psia total pressure)
— In general, when the BYU data points for the NH3-H2S-H20
system are compared to temperature-extrapolated Van Krev-
elen data WHERE THE NH3/H2S MOLAR RATIO IS ABOVE 1.5 AND
VAN KREVELEN IS THEREFORE APPLICABLE, the BYU and the
Van Krevelen correlations agree within about 20 %. Any
other comparisons at NH3/H2S molar ratios of less than
1.5 are meaningless I1'.
— The BYU data should yield a VLB correlation that would
apply at any NH3/H2S molar ratio. Such a correlation
should be very useful in designing sour water fraction-
ators including both a rectifying and a stripping sect-
ion. It would also be useful in designing high pressure
(200 psia) sour water systems.
— BYU has developed a VLB correlation based on their meas-
ured data as well as data from Van Krevelen and many
other literature sources which is combined with a method
of calculating system pH as well.
297
-------
The BYU experimental data work has greatly extended our VLB data
base, but it requires a complex computer program for application.
In most cases, for typical refinery sour water strippers, the
Van Krevelen VLB correlation as recently modified(3; still pro-
vides a simple and reliable design basis.
Having been associated on a consulting basis with almost all
of the research programs discussed in this paper, it is my own
opinion that the most useful aspect of all the work has been the
determination of the optimum caustic injection strategy for free-
ing the fixed NH3 in sour water strippers. It is also my opinion
that the experimental VLB work at BYU will prove to be very
useful.
REFERENCES
(1) Beychok, M. R. , "AQUEOUS WASTES FROM PETROLEUM AND PETRO-
CHEMICAL PLANTS", John Wiley and Sons, 1967
(2) "1972 Sour Water Stripping Survey Evaluation", API Publica-
tion 927, June 1973
(3) Beychok, M. R., "Program Calculators For Design Study",
Hydrocarbon Processing, Sept. 1976
(4) "Sour Water Stripping Project", API Publication 946, June
1975
(5) Gantz, R. G., "Sour Water Stripper Operations", API Meeting,
Chicago, May 1975. (Also, Hydrocarbon Processing, May 1975)
(6) Dobrzanski, L. T. and Thompson, W. J., "Performance Eval-
uation Of Sour Water Strippers", 76th Annual AIChE Meeting,
Tulsa, March 1974
(7) Bomberger, D. C. and Smith, J. H., "An Experimental Study Of
Ammonia Fixation In Sour Water Strippers", API Meeting,
Chicago, May 1977
(8) Wilson, G. M., "A New Correlation Of NH3, C02 and H2S
Volatility Data From Aqueous Sour Water Systems...", draft
report to API, May 1977
298
-------
DISCUSSION
N.F. Seppi: How much caustic is typically added to raw sour water stripper feed?
Milton Beychok: SRI tested caustic injection at the top, the middle and the bottom of
tneir laboratory stripper, as well as combinations of those injection points. Their caustic
injection rates ranged from 60% to 120% of the stoichiometric amount based on the
fixed ammonia present. As I recall, 90% to 100% of the stoichiometric amount was
enough to free and to remove essentially all of the fixed ammonia.
Ed Bienhoff: Are you indicating that there is a shift away from two-stage strippers
wfiere caustic is injected between stages?
Milton Beychok: I did not realize that I had even mentioned that aspect. As you
know, some refiners have used two-stage strippers to remove the ammonia and the f-LS
as separate streams and to dispose of them separately. Basically, those operations use
two conventional strippers in series. Acid is injected into the first stripper to maximize
the removal of H«S and caustic is injected into the second stripper to maximize ammonia
removal. Personally, I think that it is better to use one of the proprietary licensed
processes for that purpose, such as Chevron's fractionating process of U.S. Steel's
PHOSAM process. My reason for preferring those processes over the two-stage system
with separate acid and caustic injection is that it must be quite difficult to achieve
good pH control in the two-stage system without running into a lot of corrosion problems.
N.F. Seppi: I must have misunderstood you. Were you referring to caustic injection
into a single tower or into the feed of the second tower of a two-stage system?
Milton Beychok: In terms of the SRI work, I was referring to a single conventional
stripper with caustic injection into the feed entering that single stripper (i.e. the top
of the column), or the middle of that stripper, or the bottom of that stripper as well
as combinations of those injection points. Feed injection proved to be the best.
299
-------
BIOGRAPHY Milton R. Beychok
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 registered 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 from 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.
300
-------
TABLE 1
1972 SOUR WATER STRIPPER SURVEY*
PERFORMANCE
Average NHs removal = 78.1 % (50 towers reported)
Average H2S removal = 95.8 % (51 towers reported)
TRAYED TOWERS
Average number of trays =15 (44 towers reported)
Average number of trays, excluding three highest and three
lowest values = 15
Average tray efficiency = 46 % (12 towers evaluated)
Average tray efficiency, excluding one highest and one
lowest value = 45 %
PACKED TOWERS
Average packed height = 15 ft (14 towers reported)
STRIPPING STEAM
Average of all towers = 0.8 Ibs/gal of raw feed
Average of all towers removing more than 90 % NH3
= 1.2 Ibs/gal of raw feed
TOTAL STEAM
Average of all towers = 1.1 Ibs/gal of raw feed
Average of all towers removing more than 90 % NH3
= 1.4 Ibs/gal of raw feed
API Publication No. 927, June 1973 (73 survey questionnaires)
301
-------
TABLE 2
ALLOWABLE NH3 CONTENT IN STRIPPED SOUR WATER
TO MEET 1983 BATEA LIMITS
(Typical 125,000 BSD Refinery)
EPA category cracking
Process configuration 6.32
Process factor 1.09
Size factor 1.47
Stripped sour water = 300 gpm = 3,600,000 Ibs/day
BATEA limits (30-day average):
BOD =1.0 lbs/1000 bbls crude oil
NH3-N =1.2 lbs/1000 bbls crude oil
Variability factors:
BOD =1.7
NHs-N =1.5
BATEA allowable discharges (equivalent annual averages)i
BOD = (1.09)(1.47)(1.0)(125)/1.7 = 118 Ibs/day
NH3-N = (1.09)(1.47)(1.2)(125)/1.5 = 160 Ibs/day
Assuming a biotreater removes 90 % of the BOD and consumes
4 Ibs of Nitrogen/100 Ibs of BOD removed* :
Nitrogen consumed = 118(0.9/0.1)(4/100) = 42 Ibs/day
Nitrogen entering = 160 +42 = 202 Ibs/day
Assuming 70 % of Nitrogen entering the biotreater comes from the
stripped sour waters
Allowable NH3 in stripped water
= 202(0.7/3.6)(17/14) = 48 ppm as NH3
Repeating the above calculations for a range of values:
stripped % of NH3 entering
sour water biotreater coming allowable ppm NH3
(gpm) from SWS in stripped water
61
48
34
53
41
29
300
300
300
350
350
350
90
70
50
90
70
50
range: 30 to 60 ppm
* If only half of the BOD is removed
in a biotreater, the range is 25 to 50 ppm
302
-------
8
CO
a
w
w
I
CO
11
10
9
8
7 (-
6 -
5 -
4 -
g
Fresh feed
Reflux drum
- FIGURE 1
3
2
T
T
1 1
10,OOO ppm H2S and 7,500 ppm NH3
200 OF after exchange with bottoms
2.3 psig with a 2.0 psi pressure drop
across condenser and 0.5 psi
pressure drop per equilibrium stage
— --o-
<=> -
*1> '
x
- o-
—O Stripping steam @ 1.0 Ibs/gal of TF (1.1 ibs/gal RF)
O—— —O Stripping steam @ 1.5 Ibs/gal of TF (1.8 ibs/gal RF)
(1) Includes reboiler equilibrium
stage. Excludes overhead cond-
enser equilibrium stage.
1
1
95
96 97
% REMOVAL OF
98
99
AND
-------
PETROLEUM REFINERY DISCHARGES TO A
LARGE SANITATION DISTRICT
Irv Kornfeld
Lead Project Engineer, Sanitation Districts (Los Angeles County)
Jay G. Kremer
Head, Industrial Waste Section, Sanitation Districts (Los Angeles County)
INTRODUCTION
The Los Angeles County area is one of the major oil producing and
petroleum refining areas in the United States. Half of the reported 26
U.S. petroleum refineries discharging to publicly owned treatment works
(POTW) are served by the County Sanitation Districts of Los Angeles County
(Districts or LACSD) sewerage system (Figure 1). Refinery industrial waste-
water discharges to the Districts total in the neighborhood of 17 to 20
million gallons per day (mgd), (64,345 to 75,700 m3/d). It has been
estimated that this volume constitutes over 50% of all refinery discharges
to POTW in the United States.
Eight of the larger refineries discharging to the Districts are
refineries whose processes can be classified in EPA Category B; that is,
those refineries with topping and cracking operations (see Table 1).
The processes of the five smaller refineries can be classied in Category A;
that is, those refineries with topping and crude distillation operations
only (see Table 2). The crude capacity of these 13 refineries totals
over 800,000 barrels per stream day (b/sd), (127,176 m3/sd).
SEWERAGE SERVICE IN LOS ANGELES COUNTY
Consisting of 27 individual districts, the Sanitation Districts provide
sewerage service to the major portion of Los Angeles County outside of the
City of Los Angeles. Fifteen of these Districts collectively own and operate
the Joint Outfall System which provides a common sewerage system for
over 3.8 million people, 750 square miles (1943 km2) of area, and
approximately 8,000 industrial companies. The Districts own, operate and
maintain over 1100 miles (1770 km) of trunk sewers, and treatment facilities
for wastewater flows of 420 mgd (1,589,700 m3/d). The local 72 cities
within the Districts provide and maintain the small collection sewers.
Methods of wastewater treatment include primary, secondary and tertiary
at eleven treatment plant locations.
304
-------
THE JOINT WATER POLLUTION CONTROL PLANT
All but one of the petroleum refineries discharge to the Districts'
Joint Water Pollution Control Plant (JWPCP) located six miles inland from the
Pacific Ocean, in the City of Carson. The tributary sewers to this plant
accept primary and activated sludge from several upstream treatment plants,
in addition to domestic and industrial wastewater. The JWPCP presently
utilizes primary treatment with polymer addition and advanced solids
recovery procedures to treat a daily average of 350 mgd (1,324,750 m3/d)
of wastewater, which includes approximately 70 mgd (264,950 m3/d) of
industrial flow. The petroleum refineries' wastewater discharge amounts to
approximately 25% of this industrial flow and about 5% of the total JWPCP
wastewater influent.
The JWPCP solid wastewater material, except for grit, is digested
anaerobically. The digested material is processed by screening, centri-
fuging, and air drying for conversion to an innocuous end product suitable
for use as fertilizer. The JWPCP wastewater effluent is discharged
directly to the Pacific Ocean about two miles off-shore from the Palos
Verdes Peninsula through a system of tunnels and submarine outfalls. A
commitment has been made by the Districts and work is currently underway
to convert the JWPCP to a full secondary biological treatment plant.
Many components of refinery wastewater such as phenols, and oil and grease
will be more adequately treated by such a plant.
THE DISTRICTS' INDUSTRIAL WASTE ORDINANCE
In 1972, requirements defined in the Federal Water Pollution Control
Act Amendments (PL 92-500) and the State of California "Ocean Plan" man-
dated that the Districts establish a program which would control pollutant
levels in treatment plant effluents. To meet the Districts' treated
wastewater quality goals, control industrial pollutants, and to recover the
true cost of wastewater treatment from industrial companies, the Districts
adopted a Wastewater Ordinance on April 1, 1972.
This ordinance, which was amended on July 1, 1975, included a permit
program for industrial dischargers to the sewerage system. Information
required in the permits for major dischargers included:
1. Industrial process descriptions.
2. Industrial process equipment information and plans.
3. Description of wastewater pretreatment equipment.
4. Sewer plans of the industrial plant.
5. Pertinent wastewater constituent concentrations.
6. Wastewater flow volumes and methods of wastewater flow
measurement.
305
-------
7. Description of liquid waste materials disposed of other than
to the sewerage system.
Wastewater characterization information required from the refineries
includes the parameters listed in Table 3. The permit requirements for
the petroleum refinery industry (SIC 2911) specify minimum wastewater
pretreatment facilities of an oil-water separator and a sampling point.
THE INDUSTRIAL WASTE SECTION
The Districts' unit implementing the industrial waste regulatory
program is the Industrial Waste Section. Permit processing, field
inspection, and industrial waste engineering are the function designations
of the three subsections of the Industrial Waste Section. Graduate civil,
chemical or mechanical engineers fill over two-thirds of the 19 professional
positions in this section. One function of the industrial waste engineering
subsection is to provide technical support through a project engineer
competent in a specific field of industrial wastewater engineering. One
project engineer covers the oil producing and petroleum refining
industries.
WASTE DISCHARGES PROHIBITED BY ORDINANCE
Listed in the Districts ordinance as prohibited discharges and applic-
able to wastewaters from petroleum refineries are the following:
1. Any gasoline, benzene, naptha, solvent, or fuel oil.
2. Any waste containing toxic or poisonous solids, liquids
or gases.
3. Any waste having a pH lower than 6.0 or having any corrosive
or detrimental characteristic that may cause injury to
wastewater treatment or maintenance personnel.
4. Any solids or viscous substances of any size or in such
quantity that they may cause obstruction to flow in the
sewer or be detrimental to proper wastewater treatment
plant operations.
5. Any rainwater, stormwater, groundwater, or street drainage.
J
6. Any water added for the purpose of diluting wastes which
would otherwise exceed applicable maximum concentration
limitations.
7. Any excessive amounts of petroleum or mineral based cutting
oils (commonly called soluble oils) which form persistent
water emulsions;
8. Any excessive concentration of non-biodegradable oil,
petroleum oil or refined petroleum products.
306
-------
9- Any waste with an excessively high concentration of cyanide.
10. Any unreasonably large amounts of undissolved or dissolved
solids.
11. Any waste with excessively high BOD, COD or decomposable
organic content.
12. Any strongly odorous waste or waste tending to create
odors.
13. Any waste containing dissolved suIfides above a concen-
tration of 0.1 milligram per liter.
14. Any waste with a pH high enough to cause alkaline
encrustations on sewer walls or other adverse effects
on the sewerage system.
15. Any waste having a temperature of 120°F or higher.
16. Any excessive amounts of deionized water, steam
condensate, or distilled water.
17. Any waste containing substances which may precipitate,
solidify or become viscous at temperatures between
50°F and 100°F.
18. Any blowdown or bleed water from cooling towers or other
evaporative coolers exceeding 1/3 of the make-up water.
19. Any single-pass cooling water.
These regulations have had a major effect on the ability of Districts'
treatment plants to handle refinery wastes and the type of pretreatment
equipment required at each refinery.
INDUSTRIAL USER CHARGE
Prior to the advent of the Districts' industrial waste ordinance,
the petroleum refineries and all other large industrial dischargers
were charged only an ad valorem (property) tax for sewerage service.
This tax, however, did not cover the true costs of industrial wastewater
treatment. The Districts' ordinance established an industrial user
charge or surcharge to obtain the revenue needed to meet federal law
PL 92-500 requirements. Use was made of the charge parameters of total
wastewater flow, chemical oxygen demand, suspended solids and peak
flow in a Districts' surcharge formula (Table 4).
The surcharge rates for the last three fiscal years are given on
Table 5. The flat rate charge of $230 to $250 per million gallons can
be utilized by only industrial waste dischargers of less than 6 million
307
-------
gallons per year. However, this figure of $250 per million gallons is
the approximate amount that a large refinery will pay the Districts
for sewerage service.
The fiscal year 1975-76 ad valorem (property) taxes and surcharge
from the refineries to the Districts totaled about $1,100,000 of which
the surcharge portion amounted to about $715,000. The refineries'
surcharges amounted to approximately 16% of all industrial waste
surcharge payments received in fiscal year 1975-76.
SULFIDES AND THIOSULFATES
In the 10 years prior to promulgation of the Districts ordinance,
a Districts' policy governing the use of trunk sewers provided that
dissolved sulfides in wastewater discharged to the sewer must not
exceed a concentration^of 0.1 mg/1. The petroleum refineries, since the
early 1960's, had treated wastewater containing high dissolved sulfides
by oxidizing these sulfide compounds to thiosulfate.
In the early 1970's, new State of California discharge requirements
for the Districts' JWPCP included more stringent bacterial standards.
These standards required substantial chlorination of JWPCP effluent
when large quantities of refinery thiosulfate was present. In order
to maintain chlorination effectiveness, the Districts required that each
refinery discharging to the Districts' system must reduce its thio-
sulfate level to not more than 50 mg/1 as sulfur. The refineries were
given a period of two years, until July 1, 1973, to complete the required
construction for the reduction in thiosulfate concentration.
Only the refineries in the EPA Category B, which were high dischargers
of thiosulfate, were affected. These refineries, totaling over 90% of
all refinery discharges, accomplished the task of reducing thiosulfate by
constructing sulfide strippers for refinery sour water streams. These
sulfide strippers had a most significant effect in reducing overall
pollutant concentrations discharged from the refineries.
OIL AND GREASE
Concentrations of mineral oil and grease greater than 75 mg/1 are
considered excesssive in industrial discharges to the Districts' system.
Refinery wastewater dischargers are required, as a condition of permit
approval, to pretreat so that the oil and grease content is below
75 mg/1. All major refineries have used a combination of "good house-
keeping", improved oil water separators and dissolved air (or gas)
flotation units to obtain adequate oil and grease removals.
RAINWATER
It is the policy of the Sanitation Districts that rainwater will
not be permitted access to the Districts' sewerage system due to
limited capacity, primarily designed for dry weather flows, (separate
sanitary sewerage system). In certain situations, where discharge to the
308
-------
storm sewer is not feasible, refineries are permitted to discharge rain-
water to the sewer. Rainwater permitted to be discharged with industrial
wastewater during a rain storm is limited to the first 0.1 inch (2.54 cm)
of rainwater over the relevant surface area. Normally, rainwater diversion
devices discharge storm flow to surface drains after 0.1 inch (2.54 cm) of
rain has fallen. If greater than this amount of rainwater is too polluted
for discharge to a storm drain, it may be stored on the refinery property
during the rain storm for later discharge to the sewer. Discharge to
the Districts sewerage system can only be made 24-hours after the
cessation of rainfall and then only during the off-peak hours of sewer
flow, (10:00 p.m. to 8:00 a.m.).
TOXIC WASTEWATER CONSTITUENTS
Limits for the discharge of toxic pollutants, mainly heavy metals,
were presented to industry on July 1, 1975, with an enforcement date,
after an eighteen-month implementation period, of January 1, 1977 (see
Table 6). The Districts believe that these limits with some minor
modifications will allow the Districts to comply with it's NPDES permit
requirements for toxic pollutants discharged from its treatment plants.
The refineries' existing wastewater discharges are generally in compliance
with these toxic wastewater constituent limits. Typical concentrations
of these materials in refinery effluent are shown in Table 7.
REFINERY DISCHARGE IMPROVEMENTS
A significant reduction was noted in refinery wastewater pollutants
in 1974 following the start up of refinery sour water strippers and the
resultant reduction in wastewater thiosulfate to levels of less than
50 milligrams per liter. Tables 8 and 9 indicate typical wastewater
quality at the same category refineries during 1972-73. Table 10 and 11
indicate typical wastewater quality at the same category refineries
during 1975-76. The mass emission totals and percentage reductions for
refinery wastewater pollutants between the years 1972-73 and 1975-76 are
listed in Table 12.
In addition to thiosulfate, significant reductions were obtained in
other refinery constituent levels, including COD, ammonia, phenols, and
oil and grease. During the three year period, the refineries improved
their primary treatment facilities and upgraded their environmental
wastewater management efforts. These activities served to significantly
reduce refinery wastewater pollutant levels.
Reduction in refinery pollutant levels coincided with significant
reductions in BOD levels, about 70 mg/1, in the influent to the JWPCP.
An investigation determined that these reduced BOD levels were mainly
attributable to the decrease in refinery thiosulfate discharges.
It has been estimated that, since 1972, the cost to the LACSD
refineries for capital wastewater pretreatment improvements has amounted
to over 20 million dollars. Thiosulfate reduction, which mainly included
new sour water sulfur stripping facilities, accounted for approximately
309
-------
80 percent of these costs.
REFINERY WASTEWATER PRETREATMENT SYSTEMS
Pretreatment of wastewater at the refineries discharging to the
Districts sewerage system involves mainly advanced primary treatment.
A primary treatment system for a Category B refinery includes a sour
water stripper, a sour water oxidizer, an oil water separator, and a
dissolved gas or air flotation unit. A typical pretreatment system
for a large refinery is shown on Figure 3. Small Category A refineries
have nearly the same pretreatment systems except the sour water stripping
capability is usually not included. The small refineries, for the
most part, refine low-sulfur crude which does not generate a high-
sulfide wastewater.
Three of the large refineries discharging to the Districts' system
are able to discharge a portion of their low pollutant wastewater streams
to a storm water channel which flows to the Los Angeles Harbor area.
These channel discharges are under the jurisdiction of the local
California State Water Quality Control Board. The Board limits the mass
emission of pollutants, including BOD and oil and grease, which can be
discharged into a storm water channel.
REFINERY ODORS
The improvements in refinery pretreatment systems helped reduce
odor problems in the Districts' trunk sewers and in treatment plants
receiving refinery wastewater. One problem prevalent in the past was
refinery odors being released in a Districts' treatment plant.
With one company's cooperation, an in-plant refinery wastewater
survey was initiated in order to determine if an odorous refinery
wastewater stream could be selectively separated and treated for odor
removal prior to sewer disposal. The refinery, at the time of investi-
gation, discharged approximately 5 percent of the influent received at
the Districts' Los Coyotes Water Reclamation Plant (LCWRP), a secondary
treatment plant. At the time of the survey, the refinery did not have
a sour water stripper and all sulfides were oxidized to thiosulfate.
It was found that the sour water oxidizer, operated for wastewater
sulfide removal, was the refinery's major existing odor removal facility.
Also, selectively removing particularly odorous sour water streams
from the total refinery wastewater was not practical as over 60 percent
of the refinery sour water streams included highly odorous sulfide
concentrations. The installation of a sour water stripper facility, which
could reduce ammonia, sulfide, and thiosulfate discharges, was found to
result in substantial wastewater odor reductions. After the refinery's
sour water stripper began operation, wastewater odors at the LCWRP
were significantly reduced. Optimum stripper operation required
removing a particularly odorous refinery spent caustic stream from
sewer discharge and truck transportation of it to a landfill for
disposal.
310
-------
PETROLEUM REFINING PRETREATMENT STANDARDS
The Environmental Protection Agency (EPA) established petroleum
refinery pretreatment standards (Part 419, Federal Register, March 23,
1977) in an interim final form for pollutants discharged to POTW from'
existing sources. Established were two sets of pretreatment standards.
The first set, known as the prohibited discharge standards, is for control
of gross problems such as the discharge of flammable materials or
wastes that could plug sewers. The Districts' Wastewater Ordinance
also prohibits the discharge of such materials to the sewer system. The
second set of pretreatment standards establishes numerical limits on
ammonia and oil and grease and suggests local restrictions on chromium
sulfide and phenol.
The regulations established a maximum one-day concentration of 100
milligrams per liter (mg/1) for ammonia and 100 milligrams per liter for
oil and grease allowed to be discharged by petroleum refineries to POTW.
In addition, guidance was provided to the operators of POTW for control
of chromium, sulfide, and phenolic compounds,which may prove harmful
to or not be adequately treated by the POTW. The EPA recommended,
however, that sulfides, phenols, and chromium be controlled only as
needed by the local agency.
The Districts commented on the development document which established
these regulations. It was recommended that the EPA either propose that
sewerage agencies receiving refinery waste establish suitable local
source control programs, or that any EPA limits be established uniquely
for each of the 13 sewerage systems receiving such wastes. It is
believed that sewerage agencies such as the Sanitation Districts have
the staff and technical competence to operate a cost effective industrial
source control program and should be given the total responsibility
for such a program.
Impact on the Districts and the local petroleum industry for the
items prohibited is minimal as the Districts' ordinance has already
established most of the standards. The mean concentrations of oil and
grease, and ammonia in the refinery studies made by EPA were stated to
be well below the maximum pretreatment standard of 100 mg/1. This is
not true for all refineries discharging to the Districts' sewerage
system as two or three may discharge daily ammonia levels in excess of
100 milligrams per liter. If the Districts' treatment plants can meet
the required NPDES permit conditions for ammonia discharge, it is not
cost effective to require more severe refinery ammonia restrictions
than needed for environmental protection. All of the Districts' refineries
are required to discharge oil and grease below a daily level of 75
(<100) milligrams per liter.
SUMMARY
The Sanitation Districts of Los Angeles County provide sewerage
service for 13 (50%) of the 26 U.S. refineries reported to be discharging
to publically owned treatment works (POTW). Eight refineries are
311
-------
classified in EPA Category B and five in Category A. Total refinery
capacity is 800,000 barrels per stream day (127,176 m3/sd) with a
wastewater discharge of about 17 to 20 million gallons per day (64,345
to 75,700 m3/d).
The San_tation Districts have been able to satisfactorily treat
municipal wastewaters containing significant quantities of refinery
wastes. Wastewater discharge requirements for refineries have been
established by a Districts ordinance. This has resulted in wastewater
pretreatment installations at local refineries costing over 20 million
dollars. Most major refineries have pretreatment equipment which
includes a sour water stripper, a sour water oxidizer, an oil-water
separator, and a dissolved air (or gas) flotation unit.
A significant improvement in the quality of wastewater discharged
by refineries has occurred in the last few years. Installation of
improved pretreatment equipment by the refineries has significantly
reduced some problems at Districts' treatment plants such as high odor
levels and high wastewater organic content.
In the Sanitation Districts area, some pretreatment regulations
for refineries are quite stringent, such as 0.1 mg/1 of dissolved
sulfide and 75 mg/1 of oil and grease. Local conditions require that
these limits be more severe than proposed by EPA in the recently
published pretreatment regulations. Conversely, it appears that the
EPA pretreatment limit of a one-day maximum concentration of 100 mg/1
of ammonia from refineries may not be required for meeting pollution
control goals.
The Sanitation Districts have requested that EPA permit responsible
local sewering agencies such as the Districts to establish and enforce
its own industrial pretreatment regulatory program. Such a program,
aimed at meeting required receiving water quality, would comply with
pollution control goals without placing an excessive cost burden on
industry.
DISCUSSION
Randy Buttram; Can a 1 trotted amount of storm water be discharged along with
normal process effluent during a storm?
Irv Kornfeld: Rainwater permitted to be discharged with industrial waste-
water is limited to the first 0.1 inch (0.254 cm) of rainwater over the
relevant surface area. It is the policy of the Districts that rainwater
will not be permitted access to the Districts' sewerage system. This is
because the system has limited capacity as it is primarily designed for dry
weather flows. However, in certain situations where discharge to a storm
sewer is not feasible, refineries are permitted to discharge rainwater to
the sewer 24-hours after cessation of rainfall during the off-peak hours
of sewer flow; that is, between 10:00 p.m. and 8:00 a.m.
312
-------
BIOGRAPHIES
Irv Kornfeld is a Lead Project Engineer
in the Industrial Waste Section of the
Sanitation Districts of Los Angeles County.
He holds a B.S. degree in Mechanical
Engineering from the Illinois Institute of
Technology and a M.S. degree in Civil
Engineering from the California State
University at Long Beach. He is a
Registered Mechanical and Civil Engineer
in the State of California and a member
of ASCE and WPCF. Mr. Kornfeld has been
with the Sanitation Districts since 1972
and has prior employment with engineering
consultants and contractors to the petroleum
refining industry.
Jay G. Kremer is Head of the Industrial Waste
Section of the Sanitation Districts of Los
Angeles County. He has a B.S. degree in Civil
Engineering from Northwestern University, a M.S.
degree in Sanitary Engineering from the Illinois
Institute of Technology and a Masters of Public
Administration degree from the University of
Southern California. He is a registered Civil
Engineer in California and a member of ASCE,
WPCF and AAEE. He has been with the Sanitation
Districts since 1963 and previously worked for
various consulting engineer and governmental
agencies.
313
-------
UJ
LOS ANGELES
CITY OF
J
REFINERIES
I. ATLANTIC RICHFIELD CO.
2. DOUGLAS OIL CO.
3. GULF OIL CO.
4. MOBIL OIL CO.
5. POWERINE OIL CO.
6. SHELL OIL CO..
7. TEXACO INC.
8. UNION OIL CO.
9. EDGINGTON OIL CO.
10. FLETCHER OIL CO.
II. GOLDEN EAGLE REFINERY CO.
12. LUNDY THAGARD OIL CO.
13. MACMILLAN RING FREE OIL CO.
— OCEAN OUTFALLS
FIG. 1. SANITATION DISTRICTS OF LOS ANGELES COUNTY
PETROLEUM REFINERY LOCATIONS
-------
cr
o
H
O
O
z
>
a.
H
D
5
RATIO P/A
NOTE: Mathematical formula for "M11 is: M = 2.5 log(P/A)
2. VALUES OF MULTIPLYING FACTOR "M1
315
-------
SOUR WATER
STRIPPER
SOUR WATER
SOURCES
I
SOUR WATER
OXIDIZER
DESALTER
WATER
25% VOLUME
10%I VOLUME
NEUTRALIZATION BASIN
I
OIL WATER SEPARATOR
DISSOLVED AIR OR
GAS FLOTATION
i
DISCHARGE BASIN
FLOW METERING
SAMPLING POINT
f
PUBLIC SEWER
COOLING WATER
SLOWDOWN
BOILER SLOWDOWN
MISC. OILY WATER
65% VOLUME
SANITARY
WASTE
FIG. 3. TYPICAL LARGE REFINERY WASTEWATER
PRETREATMENT SYSTEM
316
-------
TABLE 1 REFINERIES DISCHARGING TO THE LACSD,
CATEGORY B (TOPPING AND CRACKING PLANTS)
Crude Wastewater
Capacity Discharge
Refinery Location b/sd mgd
Atlantic Richfield Co. Carson 186,000 > 4.0 < 6.0
Douglas Oil Co. Paramount 48,000 > 0.2 < 0.5
Gulf Oil Co. Santa Fe Springs 53,800 > 0.4 < 0.7
Mobil Oil Co. Torrance 131,000 > 3.0 < 5.0
Powerine Oil Co. Santa Fe Springs 46,000 > 0.2 < 0.3
Shell Oil Co. Carson 93,000 > 2.0 < 4.0
Texaco, Inc. Wilmington 79,000 > 0.6 < 0.8
Union Oil Co. Wilmington 111,000 > 3.0 < 5.0
Total 747,800 >13.4 <22.3
Metric Conversion
m3/d = (mgd) (3,785)
TABLE 2 REFINERIES DISCHARGING TO THE LACSD,
CATEGORY A (TOPPING PLANTS)
Crude Wastewater
Capacity Discharge
Refinery Location b/sd mgd
Edgington Oil Co. Long Beach 31,000 >0.10 <0.30
Fletcher Oil & Refinery Co. Carson 20,000 >0.05 <0.10
Golden Eagle Refinery Co. Carson 13,000 >0.04 <0.07
Lunday-Thagard Oil Co. South Gate 8,100 >0.02 <0.06
Macmillan Ring Free Oil Co. Signal Hill 12.200 _^0.05_ _fOJ.08_
Total 84,300 >0.26 <0.61
Metric Conversion
m3/d = (mgd)(3,785)
317
-------
TABLE 3 PETROLEUM REFINERIES WASTE CHARACTERIZATION
Parameter Quantity Values
Flow (Total) gals/day
Flow (Peak) gals/day
COD mg/1
Suspended Solids mg/1
pH units
Ammonia (N) mg/1
Oil and Grease mg/1
Phenols mg/1
Thiosulfate (S) mg/1
Arsenic mg/1
Cadmium mg/1
Chromium mg/1
Copper mg/1
Lead mg/1
Nickel mg/1
Zinc mg/1
Cyanide '--•• : - mg/1
TABLE 4 LACSD SURCHARGE
Surcharge = a (V) + b (COD) + c (SS) + dM (P) - TAX
and where surcharge equals, net,-annual industrial wastewater treatment
surcharge, in dollars. No refund is made if a negative number results.
V = Total annual volume of flow, in millions of gallons.
COD = Total annual discharge of chemical oxygen demand in thousands
of pounds.
SS = Total annual discharge of suspended solids in thousands of
pounds.
P = Peak discharge rate over a 30 minute period, occurring between
the hours of 8:00 a.m. and 10:00 p.m.
A = Average discharge rate, determined by dividing (V) by the total
annual hours of operation and working time for the industrial
discharger converted to gallons per minute.
a, b, c & d = Unit charge rates adopted annually by the individual
District based upon the projected annual total cost of wastewater
collection, treatment and disposal, in dollars per unit.
M = A multiplying factor accounting for increased Districts cost to
high ratios of industrial discharge to obtain flow rates (P/A).
Factor M is obtained from Figure 2.
TAX = The annual ad valorem taxes paid to the Districts during the accrual
years on land or property utilized for the generation of industrial
wastewater, in dollars.
318
-------
TABLE 5 INDUSTRIAL WASTEWATER TREATMENT SURCHARGE RATES
Unit Rate
a
b
c
d
Flat Rate
Charge
Parameter
Volume
Millions of gallons
per year
COD
Thousands of
pounds per year
Suspended Solids
Thousands of
pounds per year
Peak Flow
Gallons per
minute
Volume
Millions of gallons
per year
Surcharge Rates
1974-75
$104.00
$ 6.25
$ 14.25
$ 18.75
$230.00
1975-76
$104.00
$ 6.25
$ 14.25
$ 12.00
$230.00
1976-77
$127.00
$ 6.60
$ 16.10
$ 14.60
$250.00
TABLE 6 LACSD INDUSTRIAL WASTEWATER EFFLUENT
LIMITATIONS FOR JOINT OUTFALL DISTRICTS
Constituent
Arsenic
Cadmium
Chromium (Total)
Copper
Lead
Mercury
Nickel
Silver
Zinc
Cyanide (Total)
Total Identifiable
Chlorinated
Hydrocarbons
Phase I Control Period
(mg/1)
3
15
10
15
40
2
12
5
25
10
Essentially None
319
-------
TABLE 7 TYPICAL REFINERY HEAVY METALS DISCHARGE LEVELS
1975-76 (typical day)
Constituents
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Zinc
Cyanide
Concentrations
in mg/1
0.01
0.01
1.1
0.06
0.11
0.08
0.41
0.85
TABLE 8 LACSD REFINERY DISCHARGE LEVELS
1972-73 (Typical Day)
Category B Refineries
Refinery
Flow mgd
COD
Suspended Solids (SS)
pH
Ammonia (N)
Oil & Grease
Phenols
Thiosulfate (S)
Chromium (Cr)
A
4.968
mg/1
1,426
27.5
7.4
895
25
460
783
0.445
Ib/day
59,083
1,139
37,082
1,036
19,059
32,442
18.45
B
0.479
mg/1
2,640
97
8.9
1,075
165
80
865
1.10
Ib/day
10,546
388
4,294
659
320
3,455
4.39
C
3.297
mg/1
3,205
108
11.5
1,162
122
3.93
2,113
1.22
Ib/day
88,128
2,970
31,951
3,355
108
58,101
33.54
Metric Conversion
kg = (Ib)(0.454)
320
-------
TABLE 9 LACSD REFINERY DISCHARGE LEVELS
1972-73 (Typical Day)
Category B Refineries
Refinery
Flow mgd
COD
Suspended Solids (SS)
pH
Ammonia (N)
Oil and Grease
Phenols
Thiosulfate (S)
Chromium (Cr)
D
4.034
mg/1
2,956
595
8.7
181
460
1,617
688
1.06
Ib/day
99,450
20,117
6,089
15,476
54,401
23,148
35.66
E
0.739
mg/1
12,702
30
8.3
3,280
50
155
2,380
0.258
Ib/day
78,285
185
20,215
308
955
14,669
1.59
F
3.958
mg/1
1,795
8
6.9
776
9
30
1,427
1.34
Ib/day
59,252
264
25,615
297
990
47,105
44.23
Metric Conversion
kg = (Ib)(0.454)
TABLE 10 LACSD REFINERY DISCHARGE LEVELS
1975-76 (.Typical Day)
Category B Refineries
Refinery
Flow, mgd
COD
Suspended Solids (SS)
PH
Ammonia (N)
Oil and Grease
Phenols
Thiosulfate (S)
Chromium (Cr)
A
5
mg/1
774
11
10.6
35
31
32
11
0.78
.513
Ib/day
35,587
506
1,609
1,425
1,471
506
36
B
0.530
mg/1
746
49
51
~47
14
29
0.9
Ib/day
3,297
217
225
208
62
128
4
C
4.028
mg/1
1,093
48
9.8
39
54
76
28
1.05
Ib/day
36,717
1,612
- —
1,310
1,814
2,553
941
35
Metric Conversion
kg = (Ib)(0.454)
321
-------
TABLE 11 LACSD REFINERY DISCHARGE LEVELS
1975-76 (Typical Bay)
Category B Refineries
Refinery
Flow, mgd
COD
Suspended Solids (SS)
PH
Ammonia (N)
Oil and Grease
Phenols
Thiosulfate (S)
Chromium (Cr)
D
2.918
mg/1
1,223
53
8.5
547
96
78
18
1.1
Ib/day
29,763
1,290
13,312
2,336
1,898
438
27
E
0.703
mg/1
4,150
47
6.8
162
120
178
31
0.09
Ib/day
24,332
275
950
704
1,044
182
0.5
F
3.299
mg/1
282
32
7.1
92
5
18
30
0.69
Ib/day
7,759
880
2,531
138
495
825
19
Metric Conversion
kg = (Ib)(0.454)
TABLE 12 LACSD REFINERY DISCHARGE LEVELS (TYPICAL DAY),
Totals of Refineries A, B, C, D, E, F
(Over 90% of LACSD Refinery Flow)
Flow
COD
Suspended Solids
Ammonia (N)
Oil and Grease
Phenols
Thiosulfate (S)
Chromium (Cr)
(SS)
1972-73 1975-76 % Reduction
mgd
Ib/day
Ib/day
Ib/day
Ib/day
Ib/day
Ib/day
Ib/day
17.48
394,744
25,063
125,246
21,131
75,833
178,920
137.86
17.00
137,445
4,780
19,937
6,625
7,523
3,020
121.50
65
81
84
69
90
98
12
Metric Conversion
kg = (Ib)(0.454)
322
-------
SESSION Vil
POWDERED ACTIVATED CARBON
Chairman
Thomas P. Meloy
Director, Industrial and Extractive Processes
Division, U.S. E.P.A., Washington, D.C.
Speakers
Francis L. Robertaccio
"Combined Powdered Activated Carbon Treatment-
Biological Treatment: Theory and Results"
Colin G. Grieves and Michael K. Strenstrom
Joe D. Walk and James F. Grutsch
"Powdered Activated Carbon Enhancement of
Activated Sludge for BAT Refinery Wastewater Treat-
ment"
Paschal B. DeJohn and James P. Black
"Case Histories: Application of PAC in Treating
Petroleum Refinery Wastes"
James F. Dehnert
"Case History: Use of PAC With a Biodisc-Filtration
Process"
323
-------
BIOGRAPHY Thomas P. Meloy
Dr. Thomas P. Meloy is Director of the
Industrial and Extractive Processes Division, Office
of Research and Development, U.S. Environmental
Protection Agency. He holds an A.B. and B.S.
respectively from Harvard and MIT as well as a Ph.D.
from MIT. Formerly, he was Director of Engineering
Division at the National Science Foundation and
Vice-President for Research and Development at
Meloy Laboratories. His specialized field is particulate
systems. This fall he will become the Benedum Professor
at West Virginia University.
324
-------
"COMBINED POWDERED ACTIVATED CARBON - BIOLOGICAL TREATMENT:
THEORY AND RESULTS"
Francis L. Robertaccio
Senior Engineer
E. I, du Pont de Nemours & Co., inc.
Central Research and Development Department
Wilmington, Delaware
The purposes of this paper'are: to acquaint you with an
overview of some of the theoretical aspects of the combined
powdered activated carbon-biological treatment process, and to
present recent start-up experiences and results from the 1.5 x
105 M3/day installation based on the process at the Du Pont
Chambers Works Plant in Deepwater, New Jersey.
PACT is Du Pont's name for a patented process for purifica-
tion of sewage and/or industrial wastewater which comprises
subjecting the wastewater to an aerobic biological treatment
process in the presence of powdered activated carbon^' (Figure
1). The aerobic biological treatment vessel(s) can have many
geometric configurations. Single or multiple reactors can be
used. The reactors can be plug flow, completely mixed, or
somewhere in between. Powdered activated carbon addition is
compatible with activated sludge, contact stabilization, or
aerated lagoon systems; that is, any process in which the
carbon can be suspended. The rate of powdered activated carbon
addition for a given wastewater is a function of the effluent
quality desired. When the rate of addition is expressed in
terms of weight of carbon added per unit volume of incoming
wastewater, the rate becomes a function of the type of carbon
used.
In addition, certain internal process controls such as the
solids retention time, or sludge age, can be changed to in-
fluence the rate of application of a given type of carbon to
produce a desired result.(2) Some of these relationships are
illustrated in Figure 2. Here sludge age and carbon dose are
shown as variables affecting effluent quality as measured by
the total organic carbon (TOC) test- All data points represent
treatment conditions by which the effluent BOD of the industrial
wastewater tested was reduced to negligible concentration. The
effluent TOC is shown to be reduced by an independent increase
in either the sludge age or the carbon dose. Note that the im-
provement in effluent quality, by increasing sludge age, is
less apparent when carbon is absent. We have long postulated
this phenomena results because the adsorbed microorganisms
have the sludge age rather than the relatively shorter hydraulic
detention time to biodegrade adsorbed and difficult-to-degrade
molecules. It is important to recognize the economic advantage
associated with the ability to biodegrade these materials in
the biological reactor as an inherent advantage of the process.
325
-------
The alternative for this type of effluent quality improvement is
a more expensive add-on granular carbon adsorption step.
Table 1 illustrates the effectiveness of using carbon as
an adsorbent in a biological reactor.(3) The table shows the
dissolved organic carbon (DOC) removal from a PACT unit and
comparable data from a biological unit. Also shown is the
combined removal obtained from a separate adsorption of the
biological unit effluent using the same type and quantity of
carbon used in the PACT unit. The same carbon combined with
bacteria in the PACT unit removed more DOC and exceeded the
quantity expected from separate isotherm determinations.
Of course, as more molecules are biode.graded, adsorption
sites are filled with molecules that are more biorefractory and
a more rigorous form of regeneration is needed if the spent
carbon is to be reused. Alternatively, more active carbons
(i. e., higher surface area) can be used in throw away doses.
Economic considerations grouped as various capital and operating
expenses dictate the choice. To some extent the economics are
strongly influenced by the carbon usage rate, however, site
specific factors such as the local costs of alternative sludge
disposal methods must be considered. At the PACT treatment
facility for Chambers Works we will thermally regenerate carbon
from PACT sludge but wet oxidation can also be used.
!
The heart of the PACT system is a matrix of microorganisms
and powdered activated carbon. Figure 3 shows the matrix.^ '
The photo on the left is powdered activated carbon in the water;
the photo on the right the PACT matrix. The PACT matrix has
some interesting properties.
First, the carbon acts as a weighting agent. Sludge
settling rates are vastly improved as illustrated in the series
of pictures in Figure 4. Activated sludge and PACT mixed liquor
were taken from treatability units operating on the same waste-
water at the same sludge age. The series of photographs are
taken at different elapsed settling times shown on the timer in
the background. Note that the PACT sludge settles better and
has a clearer supernatant. The PACT sludge also compacts very
well. The PACT sludge had a mixed liquor suspended solids
concentration of 7700 mg/1 (about 65% carbon) and ,.a. sludge
volume index of 20 cc/g. The activated sludge had a mixed
liquor concentration of 2400 mg/1 and a sludge volume index of
46. We feel that the improved sludge settling, achieved by
simple carbon addition can result in the processing of more
wastewater through existing hydraulically overloaded treatment
plants. This is often a viable alternative to rather expensive
capital equipment expansion programs to accomplish similar
results. Of course, carbon addition will improve effluent
quality at the same time. A new treatment plant can incorporate
326
-------
this advantage in its design by substantially reducing the size
of secondary clarification. Another illustration of the im-
proved settling of PACT sludge is shown in Figure 5. Here the
initial settling velocity of activated sludge is compared at
various concentrations to a family of PACT sludge curves at two
carbon levels and two temperatures. The numbers on the PACT
labeled curves are the application rates for carbon in mg/1.
While on the topic of sludge handling, it should be
mentioned that PACT sludge dewaters much more readily than
conventional activated sludge. The manifestation of this
property is reduced size of sludge dewatering equipment. The
need to dewater more sludge as a result of the presence of
carbon is offset by reduced cycle times. Figure 6 compares the
specific resistance of activated sludge to two PACT sludges
at different carbon feed doses. (3)
A second property of the PACT sludge matrix is that it
contains an effective adsorbent. We have already explored one
aspect of the role of the adsorbent - removal of biorefractory
organic compounds - in the discussion of carbon dose and its
relationship to sludge age. In that discussion the effluent
total organic carbon content was a gross measure of biore-
fractory material. More specific measures of biorefractory
materials which might require control in specific instances
include materials contributing to final effluent color, oil
and grease, surfactants, chlorinated hydrocarbons, phenols and
toxicity to fish or other trophic level measures of toxicity
in receiving waters. No matter how you care to measure, or are
told to measure these biorefractory materials,it is apparent
PACT can control these substances to levels beyond the
capability of conventional biological systems. In complex
waste situations control of these substances at the PACT treat-
ment plant is often a more viable alternative than biological
systems followed by granular activated carbon columns or source
treatment.
Sometimes biorefractory materials are as much, or more of a
problem within a biological system as they are in its final
effluent. Examples include materials toxic or inhibitory to
biological reactor microorganisms, materials that are periodic-
ally present in high concentrations (shock loads), or materials
that cause severe foam, odor, or bulking sludge. Unlike post-
biological separate granular activated carbon treatment, the
Presence of carbon in the aerator often controls these in-
process problems as well as those normally associated with the
biologically treated effluent. Over the years, we have had a
difficult time sustaining bench scale biological treatability
units on Chambers Works wastewaters due to the periodic
Presence of toxic or inhibitory materials.(4) However, PACT
treatability units operated in parallel did not experience
327
-------
similar upsets. As a result of this property of the PACT
process the full scale PACT facility at Chambers Works has no
upstream equalization. We have even spiked Chambers Works
wastewater with various toxic substances including pentachloro-
phenol (190 1/min. pilot plant test) and found the PACT system
can sustain efficient operation when the conventional biological
system completely fails. Ferguson, et al , reported similar
findings in shock loading tests involving trichlorophenol.(5)
There are two intertwining reasons to consider any waste
treatment process: technical merits and economics. We have
touched upon the technical merits of the PACT process and
summarized them in Table II. In most instances any one of the
technical merits may be sufficient reason to consider PACT. In
most instances a number of technical merits must be simulta-
neously applied to the consideration of the process at a
specific site. The resulting matrix of reasons results in a
difficult appraisal of the full value of the use of the PACT
process versus alternative processes. Some of the economic
considerations are shown in Table III. At Du Pont we are con-
vinced PACT is a versatile, economic and technically viable
process. We have about 100 man years experience in PACT process
research and development. At the Chambers Works facility which
will be described next, we feel PACT represents a $7 million
capital savings and a $5 million/year operating cost savings
(1972 dollars) over the next best alternative which was granular
carbon treatment followed by activated sludge.(6)
The full scale PACT facility at Chambers Works has been in
a start-up phase since mid-November 1976, and it proceeded
smoothly through the coldest winter in decades. The liquid
train is on line and the solids handling train is expected to
be fully operational fairly soon. During March 1977 a half full
flowrate test, and during early May a full flowrate test were
conducted. This portion of this paper will highlight the start-
up operation and describe results of the tests.
Figure 7 shows the major components of the PACT portion
of the Chambers Works treatment plant. Construction of the PACT
facility started in February 1974 and was completed in December
1976 at an estimated capital cost of $22.5MM. Primary effluent
is split equally to each of three 15MM liter aeration tanks as
is the recycled PACT sludge. Five 1,000 hp blowers supply air
to static mixers in the aerators. Effluent from the aerators
is conveyed to the clarifier flowsplitter and then to two
secondary clarifiers. Treated effluent (overflow) is dis-
charged to a basin and then to the Delaware River. Secondary
clarifier underflow is returned to the aeration tanks via two
2 meter screw pumps. This part of the system constitutes the
"liquid train" and includes feed and unloading facilities for
carbon, phosphoric acid and polymer,
328
-------
Waste sludge can be removed from either the mixed liquor
or recycle lines and is pumped to a sludge thickener where it
is settled to 7 + 2 weight percent solids. The thickener
underflow serves as a feed stream for the filter press which is
designed to produce a 35 weight percent dry solids PACT cake.
The cake is mechanically conveyed to a five hearth furnace
where it is dryed, the biomass incinerated, adsorbed organics
pyrolyzed and the powdered carbon in the sludge regenerated for
reuse. The furnace off-gases pass through two scrubbers and
an afterburner before being discharged to atmosphere. The dry,
regenerated carbon recovered off the bottom hearth is slurried,
acid washed, and returned to the carbon feed tanks for recycle'
to the process. No waste sludge is produced. This part of the
system is called the "solid train".
The liquid train startup became evident in November 1976
when 2.7 x 10-> kg of powdered carbon and 2.0 x 105 kg of
bacterial solids were added to one aerator. Water temperature
at the time was 11-15°C. During January a second aerator and
clarifier were brought on line. During February the carbon
regeneration startup began and during March the sludge press
was brought on line. The more important operating problems
encountered and solved during startup have been presented in
a recent paper by Flynn.*'' The problems were of the type
found with the startup of a conventional activated sludge pro-
cess, that is, they were not at all related to the uniqueness
of the PACT process. These problems went through the classic
problem solving stages - initial definition, questioning of
assumptions, hypothesis forming, reobservation of the problem
in some cases, implementation of a solution and feedback on the
success of the solution.
In March a half-full flowrate test was conducted. Table IV
compares operating conditions, feed and effluent quality for
the full scale PACT facility and various bench scale controls.
The effluent color and dissolved organic carbon (DOC) are
important control parameters. During this test, flowsheet
dosages of virgin carbon (regenerated carbon not available at
this date) reduced effluent DOC to 20 ppm (43 ppm goal) for the
last seven days of the test and an average of 36 ppm for the
entire test. Effluent color was 310 (540 goal) despite the
feed color being 42% over design. The full scale, half-flow
test results compare favorably with the PACT bench scale control.
This table also presents insight into the improvement in
effluent quality offered by PACT. Note the marked decrease in
effluent DOC and color in the PACT full scale or bench scale
units versus the biological bench scale unit.
In early May a full flowrate test was conducted. Table V
Presents operating conditions, feed and effluent quality for
the test. The effluent color and dissolved organic carbon
329
-------
again met goals despite the feed color being 108% over design.
Once again this test was conducted using virgin carbon.
In summary, this paper has presented an overview of some
of the theoretical aspects of the combined powdered activated
carbon-biological treatment process (PACT) and updated recent
startup experiences from the 1.5 x 10^ M3/day installation
based on this process at the Du Pont Chambers Works. This
process is a versatile, viable wastewater treatment technology;
we expect its use to become an accepted solution to a variety
of existing and future wastewater treatment problems.
REFERENCES CITED
1. U. S. Patent 3,904,518
2. Flynn, B. P., F. L. Robertaccio and L. T. Barry,
"Truth or Consequences: Biological Fouling and Other
Considerations in the Powdered Carbon - Activated Sludge
System". Presented at the 31st Annual Purdue Industrial
Waste Conference, West Lafayette, Indiana, May 5, 1976
3. Heath, H. W., "Combined Activated Carbon-Biological
("PACT") Treatment of 40 MGD Industrial Waste" presented
to Symposium on Industrial Waste Pollution Control at
ACS National Meeting, New Orleans, LA., March 24, 1977
4. Robertaccio, F. L., D. G. Hutton, G. Grulich and H. L.
Glotzer, "Treatment of Organic Chemical Wastewater with
the Du Pont PACT Process". Presented at A. I. Ch. E.
National Meeting, Dallas, Texas, February 1972
5. Ferguson, J. F., G. F. P. Keoy, M. S. Merrill and
A. H. Benedict "Powdered Activated Carbon - Biological
Treatment: Low Detention Time Process" presented at
the 31st Annual Purdue Industrial Waste Conference, West
Lafayette, Indiana, May 4, 1976
6. Flynn, B. P. and L. T. Barry "Finding a Home for the
Carbon: Aerator (Powdered) or Column (Granular)".
Presented at the 31st Annual Purdue Industrial Waste
Conference, West Lafayette, Indiana, May 1976
7. Flynn, B. P. "Operating Problem Solving During a
Secondary-Tertiary Treatment Plant Start-Up". Presented
at llth Mid-Atlantic Regional ACS Meeting, University
of Delaware, Newark, Delaware, April 22, 1977
330
-------
DISCUSSION
Ed Sebesta, Brown S Root: What is the design hydraulic loading
for secondary clarifiers at the Chambers Works wastewater
treatment facility?
Robertaccio and B. P. Flynn, Du Pont; The solid flux rate is
designed at 250 Ibs. per day per square foot which is kind
of high. At that solid flux rate,you should be able to get
an underflow concentration of 3-1/2 weight percent. In
full scale testing we have been able to generate 7-1/2
weight percent solids which means we could operate without
a waste sludge thickener and could feed our filter press
directly from our return sludge line. The hydraulic over-
flow rate is in excess of about 1000 gallons/day/ft.2. We
have two secondary clarifiers but could send full flow
through one secondary clarifier.
Leonard W. Crame, Texaco; What does the Du Pont PACT process
patent mean to the refining industry in terms of using this
process?
Robertaccio; Du Pont will license any user of the process.
The royalty rate will be reasonable in order to encourage
use of the process.
J. E. Rucker, API; Please comment on economics and feasibility
of regeneration of powdered carbon from PACT sludge.
Robertaccio: Economics first. We think that powdered carbon
can be regenerated for an operating cost of about 5C a
pound. Capital costs would depend on the size of the
facility and the method used to annualize capital costs. At
Chambers Works capital costs would add another 5C a pound.
Now feasibility. We put as much effort into the regenera-
tion part of the Chambers Works facility as we did to the
PACT process. The regeneration system is being brought on
line. We have had some mechanical problems but we don't
expect to have any more trouble solving these as we had
solving other problems. Of course, a number of thermal and
wet oxidation regeneration equipment manufacturers will tell
you they think regeneration of powdered activated carbon
from PACT sludge is no problem.
Dave Skamenca, Envirotech: Did you pilot test mixing the
powdered activated carbon - biomass mixture with static
mixers?
Robertaccio; Yes.
331
-------
BIOGRAPHY Francis L. Robertaccio
Fran is a senior engineer for the DuPont Company
in Wilmington, Delaware. He has 12 years experience
in various industrial pollution control positions. He
holds a B.S. and M.S. in Chemical Engineering from
Clarkson College and a Ph.D. in Environmental
Engineering from the University of Delaware. He has
authored a dozen papers on industrial pollution control/
holds several U.S. and foreign patents and is a member
of AlChEand WPCF.
332
-------
TABLE I - SYNERGISTIC EFFECT ON DOC REMOVAL WITH "PACT"
DOC, mg/1
FEED TO
UNITS
183
178
167
BIO UNIT
EFFLUENT
80
70
79
BIO UNIT
EFFLUENT + BATCH
CARBON ADSORPTION (D
59
42
55
"PACT" <2)
EFFLUENT
44
18
25
TRIAL
1
2
3
(1) Take 500 cc filtered Bio Unit effluent, add 150 ppm
virgin carbon, stir 3 hours at room temperature, filter,
and analyze filtrate for DOC
(2) "PACT" unit operating at 20°C with 8.0 day sludge age
and 160 ppm carbon addition
333
-------
TABLE II - REASONS TO CONSIDER PACT
Existing Biological Treatment Plants
A. Need to improve effluent quality
treatment plant hydraulically overloaded
biological treatment unstable
odor, foam
new, more restrictive effluent limits
biological treatment inherently incapable of
removing some control parameters
• water quality limitations not met in geographic
area
B. Need to r.educe cost
• rising sludge disposal costs
• expensive chemicals used to aid biological treatment
C. Need to expand treatment plant
• want to accept new customers
• currently overloaded
• want to accept new product's wastewater; afraid
biological process will become unstable or in-
capable of removing new waste constituents
• can no longer use off plant sludge disposal site
D. This treatment plant will eventually be abandoned
(i. e., to join regional plant) but I have to get the
most out of what I have
II New (potentially biological) Treatment Plants
A. Want cost effective process
B. Concern about efficiency of biological treatment
have potentially toxic waste
face strict effluent limits
want stable process
have wastewater from changing product line
future regulations might outdate biological
treatment capabilities
C. Have components in waste not currently regulated, but
want them removed now.
D. Limited amount of land available for treatment
E. Want to minimize sludge disposal problems
• have undesirable components in waste that will
be concentrated in sludge; don't want these
released to environment
• don't have land, or availability of ocean disposal
III New Advanced Waste Treatment Plants
A. Want cost effective process
B. Want flexibility to alter treatment plant
• as regulations change
• as product mix dictates different treatment
need -
- over short intervals
- over a long period
C. Concerned about stability of alternate advanced
treatment processes
334
-------
D. Want to minimize investment
IV Miscellaneous
A. Pretreatment plus municipal disposal route too
expensive
B. PACT is a low risk process compatible with many
existing waste processing schemes, changes in
product mix, or changes in regulations
TABLE III - ECONOMIC CONSIDERATIONS OF POWDERED ACTIVATED
CARBON ADDITION
PRO CON
Eliminates granular carbon • Powdered Carbon Cost
adsorption equipment needs, - virgin
including initial GAC about 0.5-0.8C/1000 liters/
inventory lOppm
using 55 to 80Ł/kg carbon
• Regenerated (full cost*
- about 0.1-0.2C/1000
liters/lOppm
Minimizes need for equaliza-
tion facilities to control
wastewater variability
Eliminates separate second-
ary sludge disposal if re-
generation is used
Reduces or eliminates need • May require use of flocculant
for antifoam, odor control
Protects biological system
from inhibition or toxic
upset
Reduces size requirements
for secondary sludge
settling, thickening,
dewatering
Carbon addition rate
readily changed for changes
in wastewater character-
istics or regulations
335
-------
TABLE IV - CHAMBERS WORKS HALF FULL FLOWRATE TEST
TEST PERIOD 3/13/77-3/26/77 INCLUSIVE
Operating Conditions
Carbon Dose(ppm)
Aeration Temp (°C)
Hydraulic Residence
Time (hrs)
Sludge Age (days)
Feed Quality
BOD-Soluble (mg/1)
DOC (mg/1)
Color (APHA)
Effluent Quality
BOD-Soluble (mg/1)
DOC (mg/1)
Color (APHA)
FULL SCALE
PACT PLANT
182
22
14.6
304
214
1416
15.2
35.7
311
BENCH SCALE CONTROLS
CONVENTIONAL
PACT BIOLOGICAL
150 0
22 22
7.5
8
304
214
1416
19.3
28.4
369
7.5
8
304
214
1416
13.8
67.3
1900
*no steady state material balance available
TABLE V - CHAMBERS WORKS FULL FLOWRATE TEST
TEST PERIOD 4/26/77-5/6/77 INCLUSIVE
Operating Conditions
Carbon Dose (ppm) 189
Aeration Temp (°C) 28.5
Hydraulic Residence Time (hrs) 7.5
Feed Quality
BOD - Soluble (mg/1) 300
DOC - (mg/1) 214
Color (APHA) 2080
Effluent Quality
BOD - Soluble (mg/1) 9.6
DOC - (mg/1) 28
Color (APHA) 490
336
-------
Powdered Activated
Carbon
Sewage
and/or
Industrial
Wastewater
Aerobic
Biological
Treatment
Product
Water
to
Gas Containing
Oxygen
Fig. 1 A SIMPLE ILLUSTRATION OF PACT
-------
.20
u>
OJ
00
* Numbers shown in the bottom face of the exploded cube are
applied carbon doses, mg/liter
Fig. 2 EFFECT OF SLUDGE AGE AND CARBON DOSE ON
EFFLUENT TOG AND APPARENT LOADING
-------
Left:
Virgin Carbon in Water
h— 1
Right:
PACT Sludge
Fig. 3 PHOTOMICROGRAPHS
-------
PACT
SLUDGE
PACT
ACTIVATED SLUDGE
ACTIVATED
SLUDGE
PACT
SLUDGE
TKBp
INTIAL
SOLIDS. M6/L
T708 1 896
2412 Ł 150
ftCTIVATED PACT
SLUDGE SLUDGE
SLUDGE VOLUME
INDEX. CC/GRAM
20 ±2
461 3
ACTIVATED PACT
SLUDGE SLUDGE
ACTIVATED
SLUDGE
uo
t-
o
PACT
SLUDGE
ACTIVATED
. SL,U5?E SLUDGE
ACTIVATED
PACT
SLUDGE
ACTIVATED PACT
SLUDGE SLUDGE
ACTIVATED
SLUDGE
NOTE: TIMER HANDS MOVE COUNTERCLOCKWISE. READ ELAPSED TIME USING SMALL NUMBERS.
LONG HAND INDICATES MINUTES, SHORT HAND INDICATES SECONDS.
FOR EXAMPLE, LOWER LEFT TIMER READS I MINUTE, 21 SECONDS.
BOTH SLUDGES FROM TREATABILITY UNITS AT SAME SLUDGE AGE, HYDRAULIC DETENTION TIME.
. Fig.4 COMPARATIVE SETTLING CHARACTERISTICS
OF PACT AND ACTIVATED SLUDGE
-------
1.O
Initial Settling Velocity, n 1
Ft/Min
CO
0.01
PACT-ISO (winter, 10°C)
I I I
0-1 1.0
Suspended Solids Concentration, %
Fig. 5 SETTLING CHARACTERISTICS OF PACT
& ACTIVATED SLUDGES
10
-------
7.0
6.0
5.0
Bio Unit
C = .0047
M = .137
"r" = 128x1012 —
gm
M = Slope t/v vs v
C = gm solids Ice filtrate
P = Filtration Pressure, 1.5 x 107
A = Filter Area, 38.3 cm2
V = Filtrate Viscosity, .010 poise
2PA2
™ = 4.40 x 1012
N>
4.0
3.0 -
2.0 -
1.0 -
55 ppm "PACT"
C = .0167
M = .022
"r" = 5.8x1012|Ł
50
100 150
Filtrate Volume in Cm3
110 ppm "PACT"
C = .028
M = .0093
"r" = 1.5x
, cm
gm
200
250
Fig. 6 SPECIFIC RESISTANCE, "r" OF PACT vs
BIOLOGICAL SLUDGE
-------
Carbon ;
Unloading, Storage
and Feed
Primary
Effluent
-M
Five 1,000hpl
Blowers I
I Air
Three
Parallel
Aerators
-/
Polymer Unloading,
Storage and Feed
o-
Two
Parallel
Clarifiers
a
Two Parallel
Screw Pumps
Return
'Sludge"
Waste Sludge
"Liquid Train"
To Basin
'and River
yste
idge
kener
To Primary
Treatment
To Carbon Slurry
Storage Tanks *
Fuel Oil
Unloading, Storage
and Feed
Filter Procc
1
^
' 1
Afterburner
Dilution
Water
Carbon
Slurry
L_L
Acid
Contactor
Multiple
Hearth
Furnace
I Quench Tank
"Solid Train"
HCI Unloading,
Storage and Feed
Fig. 7 DUPONT PACT PROCESS: CHAMBERS WORKS
-------
POWDERED ACTIVATED CARBON ENHANCEMENT OF ACTIVATED
SLUDGE FOR BATEA REFINERY WASTEWATER TREATMENT
Colin G. Grieves
Research Engineer, Amoco Oil Company
Michael K. Stenstrom
Research Engineer, Amoco Oil Company
Joe D. Walk
Project Director, Standard Oil Company (Indiana)
James F. Grutsch
Coordinator of Environmental Projects, Standard Oil Company (Indiana)
ABSTRACT
Pilot plant studies show that powdered activated carbon enhancement of
activated sludge is a viable alternate to and less costly substitute for
granular carbon tertiary treatment of refinery wastewaters. Effluent quality
depends upon both the equilibrium concentration and the surface area of the
powdered carbon in the activated sludge mixed-liquor.
Operation at very high sludge ages—60 days or more—allows the carbon to
accumulate to high concentrations in the mixed-liquor even though only small
make-up amounts are added to the system. Also, carbons with a high surface
area are especially efficient in adsorbing contaminants. Consequently,
costly regeneration may be unnecessary because the spent carbon can simply be
discarded with the waste sludge. Powdered carbons may thus eliminate the
need for the add-on granular carbon adsorption process that the Environmental
Protection Agency has recommended for meeting proposed 1983 standards for
Best Available Technology Economically Achievable (BATEA).
INTRODUCTION
According to the EPA guidelines for treating refinery wastewaters , the
sequence shown in Figure 1 is recommended for meeting 1977 standards for Best
Practical Technology Currently Available (BPTCA). For meeting 1983 goals for
Best Available Technology Economically Achievable (BATEA), the guidelines
recommend an add-on process using granular carbon adsorption. However, this
approach may be both inefficient and very costly. So far as is known, its
effectiveness has never been adequately demonstrated. Moreover, preliminary
estimates indicate that capital and operating costs for the granular carbon
adsorption and regeneration facilities may equal or exceed those of the
entire current activated sludge process.
9—7 c
By contrast, both patents and research studies indicate that powdered
activated carbon may be a practical and economical substitute for granular
carbon. For example, powdered carbon costs only about one-half as much as
# References inserted at end of text.
344
-------
granular—$0.65/kg versus $1.20/kg15. In addition, recent studies have shown
that powdered carbon can be added directly to the mixed-liquor in activated
sludge aeration tanks Z1» 22» 23» 24. Thus, appropriate alterations in
operating procedures may eliminate the need for regeneration by making it
economically feasible to discard the spent carbon with the waste sludge.
In general, the cost effectiveness of a powdered carbon process increases
with the concentration of carbon maintained in the mixed-liquor. A mass bal-
ance of such a process is represented by the following equation:
(1)
where
C = Equilibrium mixed-liquor carbon concentration (mg/1)
Ci = Influent carbon concentration (mg/1)
9C = Sludge age (days)
9jj = Hydraulic retention time in the aeration tank (days)
Equation 1 reveals that the equilibrium mixed-liquor carbon concentration is
proportional to the product of the influent carbon concentration (carbon
dose) and the sludge age. Thus, equilibrium carbon concentration can be
increased by increasing the carbon dose, or the sludge age, or both. There-
fore, to keep carbon costs to a minimum, it is desirable to operate at as high
a sludge age as possible and not at an excessively long hydraulic retention
time.
A possible drawback to operation at a high sludge age is the increased risk
that toxic, inhibitory, or inert materials will build up in the aeration
tank. For example, a build-up of oily solids could reduce the oxygen trans-
fer efficiency and inhibit both the nitrifying and organic carbon utilizing
organisms. The dissolved oxygen concentration in the mixed-liquor could
also become too low for effective nitrification, and the final clarifiers
could become overloaded. Therefore, it is desirable in the pretreatment step
to remove as much solid material as possible from the wastewater before it
enters the aeration tank.
To evaluate the effects of such variables in a process using powdered carbon,
an extensive 15 month four-phase pilot plant study was carried out at Amoco
Oil Company's Texas City refinery. Pilot plants operating in parallel with
the refinery activated sludge process facility were fed the same wastewater
for treatment. Specific variables investigated were:
Carbon type, including surface area and pore volume
Carbon addition rate
Sludge Age
Pretreatment of feed to remove oil and solids
345
-------
EXPERIMENTAL EQUIPMENT
Figure 2 shows the configuration of the pilot plants. Each had a volume of
42 liters, and as many as eight units were operated in parallel during por-
tions of the study. They were housed in a rain-tight enclosure but were
neither heated nor cooled. Thus, the temperature of the mixed-liquor varied
from 4°C to 31°C.
Operating conditions and analytical procedures are summarized in Table 1.
The pH was checked daily and controlled by addition of caustic at a constant
rate. Dibasic potassium phosphate, K2HP04, was added to satisy the phos-
phorus requirement of the microorganisms.
The wastewater feed, a slipstream from the pressure filters of the refinery
treatment plant, was passed through a pilot gravity sand filter before being
fed to the pilot plants.
Table 2 summarizes the characteristics of the five powdered carbons evaluated.
Amoco's experimental high-surface-area carbons are designated as Al and A2,
PX-21 and PX-23, respectively. Those designated as B, C, and D are commer-
cially available carbons having a much lower surface area. Carbon A2 (PX-23)
has the highest pore volume.
Effectiveness was judged on the basis of the following effluent standards
proposed for a BATEA facility1:
Concentration,
mg/liter
Total Organic Carbon (TOC) 15
Chemical Oxygen Demand (COD) 24
Ammonia (NH3-N) 6.3
Phenolics 0.02
These standards are for a Class "C" refinery and are based on the guideline
effluent flow rate of 0.46 m^/m^ of crude throughput per stream day (19 gal/
bbl). Because the BATEA treatment sequence will undoubtedly result in very
low concentrations of effluent suspended solids, only the soluble components
of the effluent were measured.
To obtain high sludge ages, effluent suspended solids were allowed to settle
in 30-gallon plastic containers and then were returned to the pilot plants
periodically. At any given sludge age, all plants were allowed to reach
steady-state operation over an extended period of time. Then performance
data were taken over a 30-day period.
RESULTS AND DISCUSSION
The four phases of the study were carried out in sequence, with the design of
succeeding phases based on the results of the preceding ones. In summary,
they examined:
346
-------
phase Obi active
I Effect of carbon type at an addition rate of
100 mg/liter and a sludge age of 20 days with
prefiltered feed.
II Effect of carbon type at an addition "rate of
200 mg/liter and a sludge age of 20 days with
prefiltered feed.
Ill Effect of increasing sludge age to 60 days and
reducing carbon addition rate to 25 mg/liter
with unfiltered and prefiltered feed.
IV Effect of further increasing sludge age to 150
days while reducing carbon addition rate to
10 mg/liter.
Phases I and II
The results of Phases I and II, summarized in Table 3, indicate that powdered
activated carbon significantly enhances the performance of a refinery
activated sludge process. Improvement in the quality of the effluents from
carbon-fed plants ranged from 65% for soluble organic carbon up to 95% for
phenolics. At the 200 mg/liter addition rate, the results usually satisfied
the BATEA effluent quality goals. The high surface area carbon AT was
significantly more effective than the other three. The commercially avail-
able carbon B produced slightly better effluent than carbon C, which would be
expected if efficiency is proportional to surface area. Because nitrifica-
tion was essentially complete in the control unit, carbon addition could not
improve ammonia conversion. Carbon D, which is derived from wood charcoal
and has a significantly lower pore volume than the others, performed so
poorly in Phase I that it was dropped from further consideration. The per-
formance of carbon Al at 100 mg/liter dose was about as effective as carbon
B at 200 mg/liter, or about twice as effective as the best commercially
available carbon tested.
Phase III
Table 4 shows the effects of sludge age and feed'filtration upon performance.
The plant with filtered feed performed better than one with unfiltered feed,
and a sludge age of 60 days was better than one of 20 days. No deterioration
in the settling characteristics of the mixed-liquor suspended solids was
observed at this higher sludge age.
At a sludge age of 20 days the plant with filtered feed performed marginally
better than the one with unfiltered feed. Undoubtedly, greater differences
in effluent quality would have been observed ;Ln a plant operated at a sludge
age of 60 days with unfiltered feed. (Not recorded in these data, however,
is the complete failure of the plant fed unfiltered feed shortly after
cessation of data gathering for this steady-state period.)
347
-------
Table 4 also shows how pore size and surface area affect the performance of
the carbons. Carbons Al and A2 have approximately the same surface area, but
carbon A2 has much larger pores. Yet, at an equivalent addition rate of 50
mg/liter, both carbons showed about the same performance. Thus, large pore
diameters are not required for effective treatment of this refinery waste-
water. Moreover, plants fed 50 mg/liter of either Al or A2 performed much
better than the plant fed 100 mg/liter of carbon B. In fact, these high-
surface-area carbons are between two and four times more effective than
carbon B in enhancing SOC and soluble COD removal.
A comparison of the data in Tables 3 and 4 shows that a low carbon dose and a
high sludge age enhance an activated sludge process almost as much as do a
high carbon dose and a low sludge age.
It is possible that the difference in performance is solely due to difference
in temperature between the phases—mean operating temperature during Phase
III was only 14°C, whereas during Phases I and II temperature averaged 31°C
and 25°C, respectively.
Also observed during the lower operating temperature of Phase III was an
increase in the ammonia removal efficiency of the carbon-fed pilot plants.
This phenomenon was unexpected because activated carbon does not normally
adsorb ammonia. Possibly, the increased removal rate is due to the adsorp-
tion of potentially toxic or inhibitory organic materials which would reduce
the rate of nitrification if left in solution. The control plant in Phases I
and II had little difficulty in achieving full nitrification, perhaps because
of the higher temperature.
Phase IV
As shown in Table 5, Phase IV was designed to push the activated sludge sys-
tem to the limit by increasing sludge age to 150 days and decreasing carbon
addition to 10 mg/liter. Further, in one of the plants, hydraulic retention
time was reduced to 7.5 hours, compared with 15 hours in the other plants.
Despite similarities in influent quality during all four phases, during
Phase IV the effluent SOC and COD of the control increased by about 30-35%
over that observed during the first three phases, despite a mean temperature
of 27°c (c-f• 14°C during Phase III). All pilot plants essentially nitrified
completely.
Remarkably, however, the plant with 10 mg/liter of high surface area carbon
Al at a sludge age of 150 days produced an effluent whose soluble organic
carbon concentration was 50% lower than that of the control reactor and
slightly lower than that of all of the other pilot plants. The plant dosed
with 25 mg/liter carbon Al, with one-half the hydraulic capacity of the other
plants, produced the second best effluent.
The outstanding performance at a sludge age of 150 days indicates that
refinery activated sludge processes can be operated with very little added
348
-------
carbon. The dose may be low enough so that the carbon need not be regen-
erated but be discarded with the waste activated sludge. At a very high
sludge age, there will be smaller quantities of waste sludge to be disposed
of.
The data in Table 5 also indicates that powdered carbon can be used to
increase the hydraulic capacity of an activated sludge plant, as proposed by
others13, or to increase the effluent quality of an overloaded plant. The
carbon-fed plant that operated at one-half the hydraulic retention time of
the control produced an effluent 50% better than that of the control. Exper-
ience with pilot activated sludge plants operated at several of Amoco's other
refineries has shown that conventional activated sludge processes cannot be
operated successfully with a hydraulic residence time of only 7% hours.
Status of Powdered Carbon Enhancement of Activated Sludge
The data from Phase IV indicate that the limits of the powdered carbon
enhanced activated sludge process have not been reached. In addition, more
data are needed before economic studies can be made to weigh the possible
options for achieving a given effluent quality: high fresh carbon dose at
moderate sludge age (20-60 days) with regeneration of spent carbon; low
fresh carbon dose at high sludge age (60-150 days) with no regeneration of
spent carbon. Cost analyses should be made for each of these extreme options,
and several intermediate ones, and compared with those for tertiary treatment
with granular carbon technology.
Figure 3 shows the qualitative curves this pilot study has generated. Of
course, the one for the 150-day sludge age is purely speculative because
only one data point exists. However, the trend of the data does show that
effluent quality is a function of mixed-liquor carbon concentration. The
curves are probably asymptotic to a residual organic carbon concentration,
but over the range investigated an increase in mixed-liquor carbon concentra-
tion causes a decrease in effluent soluble organic carbon. Furthermore, the
relationship between effluent quality, sludge age, and carbon dose is clearly
non-linear. For example, to achieve an effluent quality of 12.5 mg/liter of
soluble organic carbon, the three options are: 100 mg/liter of carbon at a
sludge age of 20 days; 47 mg/liter of carbon at a sludge age of 60 days; 24
mg/liter of carbon at a sludge age of 150 days. If the relationship were
linear, the values calculated from a base case of 100 mg/liter at a 20-day
sludge age would be 33 mg/liter and 13 mg/liter at 60 days and 150 days,
respectively.
Apparently, the process loses effectiveness because of incomplete microbial
regeneration. Microbial regeneration of the spent carbon is probably not as
effective as using fresh carbon; some materials adsorbed by the carbon are
undoubtedly non-biodegradable, even after 150 days of contact with micro-
organisms in the pilot plant. The ability to retain significant effective-
ness even at 150 days is the key to cost effective high sludge age operation
with powdered activated carbon. Of course, there may be other reasons why
carbon loses effectiveness at high sludge age, such as production of cell
lysis products which are then adsorbed by the carbon.
349
-------
Effluent Variability
Variation in effluent quality over a 30-day (or longer) period is extremely
important. The EPA1 has set the daily maximum variability equivalent to the
99% probability value and the 30-day maximum variability to the 98% level.
For BATEA the daily maximum variability factors for TOC, COD, NH3~N, and
phenolics are proposed at 1.6, 2.0, 2.0, and 2.4, respectively. The 30-day
maximum values are 1.3, 1.6, 1.5, and 1.7, respectively.
Figures 4, 5, 6, and 7 show probability data for the 30-day operating periods
during Phase III. Table 6 shows the daily maximum (99% probability) and
30-day maximum (98% probability) variability factors calculated from these
figures for the plant fed with 25 mg/liter of Carbon Al. The EPA guideline
values are also given. The actual variability factor was calculated as the
99% (or 98%) probability value divided by the target quality value. In
general, the variability in effluent quality was higher than the guideline
values.
It is important to notethat the proposed guideline variability factors are
unrealistic.The data base used by EPA1 for their production was obtained
from limited pilot studies.-• In addition, BPTCA 30-day maximum (98% prob-
ability) values were used as the BATEA 30-day maximum values. Variability
factors will undoubtedly have to be amended before BATEA goals become BATEA
standards.
SUMMARY AND CONCLUSIONS
A viable alternative to granular activated carbon tertiary treatment of
refinery activated sludge effluent for meeting proposed 1983 BATEA effluent
quality standards has been demonstrated. The proposed process involves add-
ing powdered activated carbon to the aeration tank of the activated sludge
process, achieving cost effectiveness by operating at a very high sludge age
and a low carbon dose. Effective removal of oil and colloidal solids in
the pretreatment step is necessary for successful operation.
Effluent quality depends upon both the equilibrium mixed-liquor carbon
concentration and the surface area of the carbon. An experimental carbon
with a high surface area appears to be several times more effective than the
best commercial carbons in achieving an effluent quality standard. Pore
size of the activated carbon had no apparent effect upon effluent quality.
In general, the process can be used to meet only the long-term average
effluent quality proposed for BATEA. Daily maximum and 30-day maximum var-
iability goals, as presently defined cannot be met.
The proposed process also enhances nitrification at low temperatures and
dampens effects of increased hydraulic flow rate on the activated sludge
factors. Both phenomena will help to decrease effluent variability.
350
-------
REFERENCES
1. U.S. Environmental Protection Agency, (Draft) Development Document for
Effluent Limitations, Guidelines and New Source Performance Standards
for the Petroleum Refining Point Source Category, USEPA. Washington.
B.C., 20460 (April, 1974).
2. C. P. Derleth, U.S. Patent 1,617,014, February 8, 1927.
3. N. Statham, U.S. Patent 2,059,286, November 3, 1936.
4. F. L. Robertaccio, D. G. Button, G. Grulich, and H. L. Glotzer, "Treat-
ment of Organic Chemicals Plant Wastewater with the DuPont PACT
Process." Presented at AIChE National Meeting, Dallas, Texas, February
20-23, 1972.
5. A. D. Adams, "Improving Activated Sludge Treatment with Powdered Acti-
vated Carbon." Proc. 28th Annual Purdue Industrial Waste Conference,
1972.
6. F. L. Robertaccio. "Powdered Activated Carbon'Addition to Biological
Reactors." Proc. 6th Mid Atlantic Industrial Waste Conference,
University of Delaware, Newark 1973.
7. B. P. Flynn, "Finding a Home for the Carbon Aerator (Powdered) or Column
(Granular)." Proc. 31st Annual Purdue Industrial Waste Conference,
1976.
8. A. B. Scaramelli and F. A. DiGiano, "Upgrading the Activated Sludge
System by Addition of Powdered Activated Carbon." Water and Sewage
Works, 120: 9, 90, 1970.
9. 0. Hals and A. Benedek. "Simultaneous Biological Treatment and Acti-
vated Carbon Adsorption." Pres. 46th Annual Water Pollution Control
Federation Conference, Cleveland, 1973.
10. A. A. Kalinske, "Enhancement of Biological Oxidation of Organic Waste
Using Activated Carbon in Microbial Suspensions." Water and Sewage
Works. 115; 7, 62, 1972.
11. P. Koppe, et al, "The Biochemical Oxidation of a Slowly Degradable
Substance in the Presence of Activated Carbon: Biocarbon Unit."
Gesundeits Ingenieur (Ger) 95.: 247, 1974.
12. A. D. Adams, "Improving Activated Sludge Treatment with Powdered
Activated Carbon." Proc. 6th Mid Atlantic Industrial Waste Conference.
University of Delaware, Newark, 1973.
13. J. F. Ferguson, G. F. P. Keay, M. S. Merrill, A. H. Benedict, "Powdered
Activated Carbon-Biologica] Treatment: Low Detention Time Process."
Paper presented etc the Jist Annual Industrial Waste Cunleieuce. IUJ.UUB
University, Lafayette, Indiana. 1976.
351
-------
14. F. B. DeWalle and E. S. K. Chian, "Biological Regeneration of Powdered
Activated Carbon Added to Sludge Units." Water Research 11; 439, 1977.
15. F. D. DeWalle, E. S. K. Chian, E. M. Small, "Organic Matter Removal by
Powdered Activated Carbon Added to Activated Sludge." Journal Water
Pollution Control Fed. ^9_: 593. 1977.
16. Anon, "Chemenator" Chemical Engineering Ł4: 1, 35. 1977.
17. P. B. DeJohn, "Carbon from Lignite or Coal: Which is Better?" Chemical
Engineering. 82:9, 13. 1975.
18. D. G. Hutton and F. L. Robertaccio, U.S. Patent 3,904,518,
September 9, 1975.
19. P. B. DeJohn and A. D. Adams, "Treatment of Oil Refinery Wastewaters
with Powdered Activated Carbon." Pres. at the 30th Annual Purdue
Industrial Waste Conference. 1975.
20. J. A. Rizzo, "Use of Powdered Activated Carbon in an Activated Sludge
System." Proceedings of th&iOpen Forum on Management of Petroleum
Refinery Wastewaters, January 26-29, 1976, EPA, API, NPRA, University
of Tulsa, at the University of Tulsa, Tulsa, Oklahoma.
21. M. K. Stenstrom and C. G. Grieves, "Enhancement of Oil Refinery Acti-
vated Sludge by Additon of Powdered Activated Carbon." Pres. at 32nd
Annual Purdue Industrial Waste Conference, 1977.
22. C. G. Grieves, M. K. Stenstrom, J. D. Walk, and J. F. Grutsch, "Effluent
Quality Improvement by Powdered Activated Carbon in Refining Activated
Sludge Processes." Pres. at the 42nd Mid-year Refining Meeting, API.
Chicago, Illinois, May 9-12, 1977.
23. G. T. Thibault, K. D. Tracy, J. B. Wilkinson, "Evaluation of Powdered
Activated Carbon Treatment for Improving Activated Sludge Performance."
Pres. at the 42nd Mid-year Refining Meeting, API. Chicago, Illinois,
May 9-12, 1977.
f
24. L. W. Crame, "Pilot Studies on Enhancement of the Refinery Activated
Sludge Process." Pres. at the 42nd Mid-year Refining Meeting, API.
Chicago, Illinois, May 9-12, 1977.
25. B. R. Kim, V. L. Snoeyink, F. M. Saunders, "Influence of Activated
Sludge CRT on Adsorption.1! Environ. Engr. Div., ASCE, 102; 55. 1976.
26. J. F. Grutsch and R. C. Mallatt, "Optimize the Effluent System."
Hydrocarbon Processing. 55; 3: 105-112.
27. U.S. Environmental Protection Agency, Methods for Chemical Analysis
of Water and Wastes, USEPA, Washington, D.C., 1974.
352
-------
DISCUSSION
Piysuch Shah, Exxon Research and Engineering Co.; Would you please comment on the
effects of toxicant build up and on the performance of aged activated sludge units,
especially 100-150 days? Also, what is the maximum concentration that can be allowed
in the feed?
Grieves; Assume a very high sludge age of 150 days, a hydraulic residence time of 12
hours, 1 mg/liter of a toxicant (for example, chromium) in the feed, and 100% removal
of it by the activated sludge. At steady state, chromium concentration would build up to
300 mg/liter, which, in all likelihood, would be toxic to the microorganisms. However,
we have data to indicate that even at 150 days sludge age, chromium does not accumu-
late to more than 30-50 mg/liter in the sludge. We certainly have not observed any
effects of toxicant build-up — on the contrary, the 150-day sludge age reactor is the
most effective unit.
As for other toxicants — for example oil and grease and inert suspended solids
-- if they are not effectively removed by prefiltration, or air flotation, they could very
well accumulate to toxic or inhibitory concentrations in the mixed-liquor. As well as
being toxic or inhibitory to microorganisms, especially nitrifiers, oxygen transfer problems
will be encountered. High inert solids concentrations may also cause overloading
problems in the final clarifier.
Ed Sebesta, Brown & Root, Inc.; The data indicates that nitrification occurred during
some phases of the experiments while nitrification did not occur during other phases. Do
you have any comments about why this occurred?
Grieves: If you have ever operated a refinery wastewater treatment facility, you will
know that frequently there are excursions with nitrification. We achieved good nitrifi-
cation during the warm operating periods, phases I and II. During phase III operation, it
was relatively cool — we recorded a mixed-liquor temperature of 2°C on one occasion,
quite a severe winter for this part of Texas — and, as expected, nitrification in the
control activated sludge plant was poor. This is reflected in the probability plot (Figure
6) of the data. However, in the activated sludge pilot plants to which carbon was added,
almost complete nitrification was observed. This was unexpected because, as you know,
carbon does not normally adsorb ammonia.
Bob Smith, Carborundum Co.; Have you compared the cost effectiveness of the high
capacity Amoco carbon vs. the lower capacity carbons?
Grieves: No, we have not made this comparison yet. We have not decided whether to
go commercial with our carbon or not. However, if and when we do decide to
commercialize our product, you can rest assured that it will be cost effective with other
commercially available carbons.
353
-------
BIOGRAPHIES
Colin G. Grieves is a Research Engineer in the
Water Conservation group at Amoco Oil Company's
Research and Development department in Naperville
Illinois. He has M.S. and Ph.D. degrees in
Environmental Systems Engineering from Clemson
University, Clemson, South Carolina, and a B.Sc.
degree in Civil Engineering from the University
of Newcastle Upon Tyne, England. Previous
employment was in the Public Health Engineering
Division of Babtie Shaw & Morton, Consulting
Civil and Structural Engineering in Glasgow,
Scotland. Colin is the author of several papers
in the wastewater treatment field.
Michael K. Stenstrom is a Research Engineer in
the Water Conservation group at Amoco Oil
Company's Research and Development department in
Naperville, Illinois. He has a B.S. degree in
Electrical Engineering and M.S. and Ph.D. degrees
in Environmental Systems Engineering from Clemson
University, Clemson, South Carolina. Mike is the
author of several papers dealing with various
aspects of municipal and industrial wastewater
treatment.
354
-------
BIOGRAPHIES
Joe D. Walk is Active Carbon Project Director in
the Corporate Development Department of Standard
Oijl (Indiana), Chicago, Illinois. He has a B.S.
degree in Chemical Engineering from the University
of Texas. Previously he served as Process
Coordinator in air/water conservation, crude
running and product treating for Amoco Oil's ten
refineries. Prior assignments include Manager
of Technical department at Texas City refinery,
as well as positions in New York City, New Orleans,
and El Dorado, Arkansas, during his 31 years
service with Standard Oil.
James F. Grutsch is Coordinator-Environmental
Projects, Standard Oil Company (Indiana). He
holds undergraduate and graduate degrees in
chemistry from Indiana University. Prior to his
present assignment with Standard, Jim served
successfully 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 6 years at Indiana.
355
-------
TABLE 1 OPERATING CONDITIONS
Pilot Plant Conditions
Aeration Zone Volume
Settling Zone Volume
Nominal Flow Rate
Nominal Hydraulic Retention Time
Nominal Settling Time
Air Flow Rate
PH
Caustic Addition Rate
Phosphorous Added to Feed
Temperature
36.7
5.7
2.45
15.0
2.33
300
6-8.5
0.12-0.30
3
liters
liters
liters/hr
hours
hours
liters/hr
liters/hr
mg/ liter
Ambient (4-31°C)
Frequency
Daily
3 Times
a Week
Analytical Work
Analysis Performed^'» 28
Once a
Week
Influent and mixed-liquor pH, temperature,
influent flow rate, caustic addition rate.
Carbon addition and sludge wastage.
Influent and effluent total and volatile sus-
pended solids, soluble organic carbon, soluble
chemical oxygen demand, soluble ammonia nitro-
gen, and soluble phenolics. Mixed-liquor
suspended solids and mixed-liquor volatile
suspended solids. Sludge volume index.
Material balances to calculate quantity of
sludge to be wasted to maintain desired sludge
age.
356
-------
TABLE 2 PROPERTIES OF POWDERED ACTIVATED CARBONS
CO
Property
Carbon Designation
Experimental Amoco
High Surface Area -
Surface Area
BET, m2/g
Pore Volume, cc/g
> 15 A° Radius
< 15 A° Radius
Iodine Number
Methylene Blue Adsorption,
Phenol Number
Bulk Density, g/cc
Screen Analysis
Passes 100 Mesh, Wt.%
Passes 200 Mesh, Wt.%
Passes 325 Mesh, Wt.%
Al
Grade PX-21
3099
0.16
1.45
3349
mg/g 586
12.8
0.298
98.4
92.7
84.1
A2
Grade PX-23
3148
0.43
1.60
3375
550
12.6
0.228
99.1
93.4
80.8
Commercially Available Conventional Surface
Area Carbons
B
717
0.28
0.51
1790
100
34.1
0.610
99.2
86.7
60.6
C
514
0.38
0.11-0.42
920
83
22.9
0.576
100.0
94.4
68.3
D
532
0.03
0.25
888
50
23.8
0.484
100.0
97.9
91.8
Molasses Number
10
205
103
85
-------
TABLE 3
PHASES I AND II - EFFECT OF CARBON TYPE AND ADDITION RATE ON EFFLUENT QUALITY*
50% Probability Data During 30 Days of Steady-State Operation
Sludge Age = 20 Days
Concentration, mg/liter
Filtered Pilot Plant Effluent
Component Influent No Carbon Carbon Al Carbon B Carbon C Carbon D
Phase I: Carbon Addition Rate = 100 mg/liter
Equil. Mixed-Liquor Temp •» 31°C. Carbon Cone = 3200 mg/liter
soc
SCOD
NHs-N
Phenolics
Equil
SOC
COD
NH3-N
Phenolics
72.0
230
25.8
4.35
22.0
73
0.5
0.018
Phase II: Carbon
. Mixed-Liquor Temp
70.0
230
25.4
4.06
26.5
58
0.2
0.020
12.5
28.5
0.2
0.003
Addition Rate =
= 25°C, Carbon
9
17
0.2
0.001
17.5
48
0.5
0.010
18.5
44
0.5
0.010
23.0
65
0.5
0.017
200 mg/liter
Cone = 6400 mg/ liter
13.5
24
0.2
0.001
15.5
28
0.1
0.003
* BATEA effluent standards in mg/liter are: Soluble Organic Carbon (SOC) 15
Soluble COD (SCOD) 24
Ammonia Nitrogen (NH3-N) 6.3
Phenolics 0.02
358
-------
TABLE 4
PHASE III - EFFECT OF CARBON TYPE AND ADDITION RATE, SLUDGE AGE,
AND INFLUENT PRETREATMENT ON EFFLUENT QUALITY
50% Probability Data During 30 Days of Steady-State Operation
Equil. Mixed-Liquor Temp = 14°C, Carbon Cone. = 2400 mg/liter
Carbon
Influent
Pretreatment
Filtered Feed
Addition
Rate, mg/liter
Influent Concentration, mg/liter
SOC COD NH3-N Phenolics
Unfiltered
Filtered
Filtered
Filtered
Filtered
Filtered
Filtered
B
Al
Al
A2
73.5 294.5 19.3 3.95
Effluent Concentration, mg/liter
Sludge Age = 20 Days
32.0 103.5 12.1 0.027
29.0 83.0 14.5 0.027
Sludge Age = 60 Days
25.0 65.9
100 16.0 40.3
50 12.0 27.5
25 16.0 50.3
50 13.0 31.0
5.1
0.2
0.1
0.4
1.8
0.019
0.001
0.002
0.006
0.004
359
-------
TABLE 5
PHASE IV - EFFECT OF HIGH SLUDGE AGE, LOW CARBON ADDITION RATE,
AND DECREASED HYDRAULIC RETENTION TIME ON EFFLUENT QUALITY*
50% Probability Data During 30 Days of Steady-State Operation
.Equil. Mixed-Liquor Temp = 27°C
Type
B
Al
Al
Al
Carbon
Addition
Rate, mg/liter
-
25
25
25
10
Sludge
Age, days
60
60
60
60
150
Hydraulic
Retention
Time, hr
15
15
15
7.5
15
Equil. Mixed
Liquor Carbon
Cone, mg/liter
-
2400
2400
4800
2400
Effluent Cone,
mg/liter
SOC
29
22
18
17
16
COD
99
64
52
46
49
NH3-N
0.1
0.1
0.1
0.3
0.1
Phenolics
0.018
0.010
0.010
0.010
0.010
* Filtered influent contained 78 mg/liter SOC, 270 mg/liter COD, 29 mg/liter NH3-N, and 3.25 mg/liter
phenolics.
-------
TABLE 6
PHASE III - BAT GUIDELINE AND ACTUAL VARIABILITY FACTORS FOR PILOT
PLANT FED 25 mg/liter OF CARBON Al
BAT Guideline
Parameter
Soluble Organic Carbon
Soluble COD
NH3-N
Phenolics
Variability
Daily Max.
1.6
2.0
2.0
2.4
Factor
30 Day Max.
1.3
1.6
1.5
1.7
Actual
Variability Factor
Daily Max. 30
2.8
7.5
2.1
5
Day Max.
2.8
7.5
2.0
5
-------
FIGURE 1. SIMPLIFIED REFINERY BPT WASTEWATER
TREATMENT SYSTEM
Chemicals
Refinery
Waste-~~"
water
Gravity
Separator
Slop Oil
to
Treatment
Dissolved
Air
Flotation
Equalizat-
ion
Granular
Media
Filter
Final
-•••Effluent
Recycle
Sludge
to
Treatment
-------
FIGURE 2. SCHEMATIC OF ACTIVATED SLUDGE REACTOR
USED IN PILOT PROGRAM
CO
Air
.Baffle
Air— »•«==
36.7 Liters
Aeration
Zone
=
/
\/
\&
* \
.v y|
Side
Feed -*•<=
i—^FfflMpnf-
— 5.7 Liters
Settline
Zone
Sludge
Blanket
mSm
<
1
<
Enc
i
I
r
r
\
a^ Chemicals
Sampling
SS!""~Port
\ Dif fuser
Stone
Chemicals.
{r?r
Feed.
o
o
Top
Plexiglass
./Construction
Effluent
>A.ccess
Hole
-------
c
o
I
o
o
O
O
I/)
0)
LU
FIGURE 3. EFFECT OF MIXED-LIQUOR CARBON
CONCENTRATION ON EFFLUENT SOC
Legend
20 days sludge age
60
150
carbon
dose
mg/l 100
treatment objective
-150 days
1 2 3 A
Mixed-liquor carbon concentration
8
-------
o
o
FIGURE 4 - SOLUBLE ORGANIC CARBON - PHASE 3
CONTROL, NO CARBON, 60 DAY SRT-X
AMOCO PX-21, 25 M8/1, 60 DAY S8T- +
AMOCO PX-21, 50 U8/L, 60 DAY SRT-A
CARBON B, 100 H8/L. 60 DAY SRT-O
LU
X
o
2
U
•
o
0.013
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
CUMULATIVE PROBABILITY
0.9
0.987
365
-------
FIGURE 5
O
O
SOLUBLE CHEMICAL OXYGEN DEMAND
PHASE 3
CONTROL, NO CARBON, 60 DAY SRT-X
AMOCO PX-21, 25 U6/L, 60 DAY SRT- +
AMOCO PX-21, 50 UO/L, 60 DAY SRT-A
CARBON B, 100 HO/I, 60 DAY SRT-CP
CC
UJ
CD
•
O
•
O
•
O
•
CO
0.023
0.2 0.3 0.4 0.5 0.6 0.7 0.8
CUMULATIVE PROBABILITY
0.9
0.977
366
-------
FIGURE 6 - AMMONIA-NITROGEN - PHASE 3
CONTROL, HO CARBON, 60 DAY SRT-X
AMOCO PX-21, 25 IW/U, 60 DAY SRT- +
AMOCO PX-21, SO U6/L, 60 DAY SRT-A
CARSON B, 100 MS/I. 60 DAY SRT-O
CC.
LU
C!)
2
I
ro
0.013
O.t 0.2 0.3 0.4 O.J 0.6 0.7 0.6 0.9
CUMULATIVE PROBABILITY
0.987
367
-------
FIGURE 7 - PHENOLICS - PHASE 3
10
ro
CO
O
O
z
LU
•3Z
Q_
O
0.036
CONTROL, NO CARBON, 60 DAY SRT-X
AMOCO PX-21, 25 U6/L, 60 DAY 8RT- +
AUOCO PX-21, SO M8/L, 60 BAY SRT-A
CARBON B, 100 M8/L, 60 DAY 3RT - O
0.2 O.S 0.4 O.S 0.6 0.7 0.8
CUMULATIVE PROBABILITY
0.9
0.964
368
-------
"TREATMENT OF OIL REFINERY WASTEWATERS
WITH POWDERED ACTIVATED CARBON"
Paschal B. DeJohn
Manager, Purification Sales, ICI United States Inc. (Delaware)
James P. Black
Industry Coordinator, ICI United States Inc. (Delaware)
INTRODUCTION
The effectiveness of powdered carbon as an additive to improve activated
sludge treatment has been demonstrated in a variety of industrial and munici-
pal plants. This type of treatment has gained wide acceptance in the past
few years and is currently an essential part of treatment at 60-80 plants.
These plants range in size from 10,000 gpd package units located along the
Alaska Pipeline to the very sophisticated 40,000,000 gpd PACT treatment plant
at the DuPont Chambers Works in Deepwater, New Jersey. At least four petrol-
eum refiners currently use powdered carbon as an integral part of their waste
treatment scheme.
HOW THE PROCESS WORKS
The reason powdered carbon has gained such acceptance treating a wide
variety of waste streams is the extreme flexibility which can be employed in
its usage.
• The amount of carbon used can be varied to meet the
treatment requirements as they change.
• Higher COD or BOD removal than is usually obtainable
by conventional biological treatment can be achieved.
• The combination of activated carbon in a biological
system provides more effective treatment than either
of the processes would if used singularly.
Carbon aids the biological process two ways:
1. By direct adsorption of pollutants.
2. By providing a more favorable environment for the micro-
organisms to propogate.
Adsorption is an equilibrium phenomenon. In general, carbon preferen-
tially absorbs higher molecular weight compounds. Given a related series of
organic compounds; for example, alcohols, one finds that the lower molecular
weight alcohols (methanol, ethanol) are not appreciably absorbed by carbon
while the higher molecular weight alcohols are. Fortunately, compounds which
are poorly absorbed (weakly held by the carbon) are usually compounds which
are the most amenable to biological treatment.
369
-------
We can generally classify organic compounds into three broad categories
with respect to their adsorptability onto carbon.
1. Compounds which are readily adsorbed. These compounds are
usually "tightly held" by the carbon. And consequently,
they are not readily desorbed.
2. Compounds which are adsorbed with difficulty. These
compounds are desorbed easily.
3. Compounds which are poorly adsorbed.
Organic compounds can also be classified in terms of their susceptibili-
ty to biodegradation:
1. Compounds which are readily and rapidly biodegraded.
2. Compounds that are degraded slowly.
3. Compounds that are not biodegraded. Many of these can
function as toxicants in a biological system.
It is important for one to understand the interaction of carbon and the
microorganisms present in an activated sludge system. Exhibit 1 does this by
considering how a carbon-biological system handles each of the above classi-
fications of organics compounds. Relative adsorptivity and biodegradability
for organic compounds was taken from an EPA source (Reference 1).
The boxes in Exhibit 1 have been numbered from 1 through 9 and are
interpreted as follows:
Degree of
Box Biodegradability Adsorptability
1 Rapid Strong
2 Rapid Moderate
3 Rapid Weak
4 Slow Strong
5 Slow Moderate
6 Slow Weak
7 None Strong
8 None Moderate
9 None Weak
The classes of compounds which are represented by boxes 1, 2, and 3 would be
handled quite easily by the microorganisms in a carbon-biological system.
Those compounds which are represented by boxes 1, 4, and 7 would be removed
by direct adsorption on the carbon. Compounds which fall in box 1 (both
rapidly biodegradable and strongly adsorbed) are few in number. The only
example that we could find is o-cresol.
Boxes 4, 5, and 6 represent compounds which are slowly biodegradable.
These compounds probably would not be removed very effectively in a conven-
tional activated sludge system. In a carton-biological system, compounds in
370
-------
box 4 are removed by direct adsorption and are held very tightly by the
carbon. Compounds in boxes 5 and 6 are retained in the system by moderate or
weak adsorption. Because carbon is adsorbing these compounds, their concen-
tration in the liquid stream is reduced. The microorganisms are able to
degrade the organics in reduced concentrations, and as they do the equilli-
brium between the carbon and these organics in the waste is disturbed.
Because these compounds are not held very tightly by the carbon, they are
readily desorbed back into the system and a new equilibrium is established.
In this fashion, the carbon is acting as a storage area keeping the concen-
tration of slowly biodegradable organics at a level where they can be handled
by the microorganisms. Compounds which fall into the categories represented
by boxes 4, 5, and 6 are effectively handled in a carbon-biological system
because of synergistic effects. It is primarily these compounds that are
removed more effectively in a carbon-biological system as compared to either
process operating singularly.
Compounds represented by boxes 7, 8, and 9 are not biodegradable. And
in some cases, these compounds are actually toxic to microorganisms. In our
opinion, carbon performs its most beneficial action in this area. Compounds
which fall into the category represented by box 7 are removed by direct
adsorption and are held very tightly by the carbon. Compounds in box 8 are
removed by direct adsorption, and even though they are held very loosely by
the carbon, it is difficult to disturb the equillibrium. This is because
the concentration of these organics remaining in the waste stream are not
being degraded by the microorganisms. The compounds represented by box 9
are the only ones that cannot be handled very effectively in a carbon-biolo-
gical system. Fortunately, there are very few organics which are both non-
biodegradable and weakly adsorbed by carbon.
Examples of the different compounds are shown in the various boxes in
Exhibit 1.
One of the important effects of carbon in the system (not related dir-
ectly to adsorption) is the higher levels of biomass that can be used because
of the density and "weighting effect" of the carbon. (Both the use of great-
er sludge mass and the temporary retention of slowly degraded compounds by
the carbon gives more time for the compounds to be consumed biologically).
Carbon adsorbs the pollutants and oxygen, localizing them for bacterial
attack. Because the aerobic action is dependent upon the concentration of
the reactants, this localizing effect serves to drive the reaction further
towards completion resulting in improved BOD removal (Reference 2).
Many pollutants that are not biologically degraded in a conventional
activated sludge system would be if they were in contact with the biomass
for a longer period of time. When absorbed by the carbon, these molecules
settle into the sludge. Contact time is thereby, extended from hours to days.
This results in lower effluent COD's and TOC's. High density powdered carbons
Improve solids settling in the secondary clarifiers. This results in lower
effluent suspended solids and also a reduction in BOD. Under high organic
load conditions which normally would lead to sludge bulking, the dense carbon
371
-------
will act as a weighting agent keeping the sludge in the system. When dis-
persed biofloc results due to low organic loads, carbon serves as a seed for
floe formation preventing loss of solids. Under these conditions, phosphorous
and nitrogen removal are generally enhanced.
Powdered carbon improves treatment in activated sludge process because
of its adsorptive and physical properties. Powdered carbon can be added to
any convenient point in the activated sludge process to get it into the
aerator. Direct addition to the aerator, sludge return lines, influent
channels, or through the secondary clarifier are all possibilities. It is
not necessary to add carbon continuously in most cases. A dense, easily
wetted carbon can be added dry or in slurry form with water.
RESULTS
14 refineries have evaluated HYDRODARCO powdered activated carbons in
full scale activated sludge systems during the past three years. The first
treats a 2.2 MGD flow with an average BOD of 400 ppm in a 1.2 million gallon
aerator. Mixed liquor solids are maintained at 3600 ppm (2880 ppm volatile).
Waste activated sludge is digested aerobically, centrifuged, and hauled to
landfill. Despite a secondary clarifier overflow rate of only 423 gallons/
ft.^ and use of 22 ppm cationic polymer for secondary solids capture, efflaent
solids averaged in excess of 100 ppm. Toxic loads caused periodic loss of
aerator biosolids. Defoamer costs averaged $200/day for aerator foam control.
HYDRODARCO C, a high density, lignite based powdered carbon, was added
to the aerator over a four and one-half month period. Eventually, the equil-
librium carbon level reached 1800-2000 ppm. At the sludge solids concentra-
tion obtained and wasting rates employed, it was possible to maintain this
level with a daily average carbon dose of only 20 ppm.
Over the entire carbon test period, average BOD reduction equaled 82%
versus 23% during the post test control period (Figure 1). As carbon built
up in the system, BOD removals reached the 90-95% range, and the plant was
able to meet their 30 ppm BOD effluent standard. Effluent COD was reduced
from an average of 1180 ppm without carbon to 350 ppm with (Figure 2). Aver-
age effluent TOC decreased from 420 ppm to 100 ppm (Figure 3), and total
carbon decreased from 520 ppm to 180 ppm (Figure 4 ). The lower slope of the
carbon plots also indicate the decreased variability in effluent quality with
carbon present.
HYDRODARCO C had a dramatic effect on the reduction of oil through the
system (Figure 5 ) • The effluent concentration was reduced by 75% (average),
and the range was narrowed as well.
Both removal of the oil by the powdered carbon and the weighting effect
of carbon resulted in lower effluent solids (Figure 6). Prior to carbon
treatment, the plant used polymer at a dosage of 20 ppm, but still experienced
poor solids settling. When carbon was added to the system, solids settling
improved, and the polymer dosage was cut in half. Since effective solids
settling could not be achieved with the use of carbon or polymer alone, it
appears that the combination of the two was required to attain the desired
results. This, of course, represents an operating cost savings for the plant.
372
-------
Improved solids settling increased sludge thickening which allowed a 65%
reduction in sludge wasting. Again, savings on the operation of the centri-
fuges, including power and labor, occurred.
Use of HYDRODARCO C eliminated the need for aerator defoamer. Removing
the foaming agents from the wastewater by adsorbing them eliminated foam
problems in the receiving stream. Defoamers only suppress foam in the aera-
tor and do not prevent its reappearance in the effluent. Carbon can reduce
operating costs by allowing surface aerators to aerate and mix the activated
sludge rather than expand energy generating foam.
Both nitrogen (Figure 7 ) and phosphorous (Figure 8 ) removals were
improved with powdered carbon. Reason: carbon adsorbs compounds toxic to
nitrifiers and allows them to operate at normal levels. In this waste,
neither nitrogen nor phosphorous were limiting for bacteria growth. Increased
nitrogen removals are attributed to the fact that the dense carbon settled
the nitrifying organisms which normally would float out of the system. The
result is a longer solids retention time which is more favorable for nitrifi-
cation to occur. By the same token, improved solids settling is probably
the reason for decreased phosphorous levels. We suspect that the phosphorous
is precipitated with the carbon-biosolids floe and removed in the sludge
rather than degraded biologically.
While the exact reason for the bug kills prior to carbon was not known,
upsets were greatly reduced with carbon in the aerator. A possible explana-
tion is the effect carbon had on the removals of heavy metals such as zinc
(Figure 9). The ability of activated carbon to adsorb heavy metals from
wastewater has been established elsewhere (References 3, 4, and 5).
The second evaluation was conducted at a 12 MGD plant treating an
average 12 MGD flow. The TOG of the raw waste ranged from 100-1000 ppm,
averaging about 200 ppm. Major treatment problems included aerator foaming
caused by alkanolamines in the waste; high effluent TOG; oily, difficult-to-
handle sludge; and high effluent solids.
Effluent TOC's were maintained below 20 ppm during shock load periods.
This was well within the standard of 50 ppm. In a post test control phase,
a deterioration in effluent quality was observed as carbon was lost from the
system through sludge wasting.
A third evaluation was conducted at a 2.5 MGD plant treating a 550 ppm
COD refinery waste in a two stage, conventional activated sludge system.
Carbon was added to the second stage aerator over a six week period. A con-
stant daily carbon dose was maintained for each week and increased in
succeeding weeks. No sludge was wasted intentionally during this time.
Optimum treatment was found at relatively high influent dose of 200 ppm.
Effluent solids and COD removals increased 40%, and BOD removals, already
high, increased 10%.
373
-------
The major finding of this study was the increased removals of cyanide
with carbon in the aerator. In conjunction with CuSCfy treatment, the average
cyanide levels decreased from 1 ppm before carbon to 0.05 ppm. The precise
nature of this removal is not known and bears further investigation.
Results from a fourth study are summarized in Exhibit R for three
separate carbon addition periods. This plant is a current user of activated
carbon to treat 2.2 MGD flow. They have reported the following results
from carbon addition (Reference 6):
56% reduction of suspended solids.
36% reduction of COD.
76% reduction of BOD.
Foam problem elimination.
This improved plant performance is achieved at a carbon cost of
1.7-4.3C/1000 gallons treated.
A fifth study was made at a refinery which had a 6 MGD (8 MGD design)
activated sludge plant. Carbon dosage reached approximately 500 ppm in the
aerator before the study was terminated due to loss of biosolids. The loss
of biomass resulted from an inadvertent increase in sludge wasting. This
plant continued to use the same volumetric wasting rate when carbon was added
to the system and because of the sludge density increase due to carbon addi-
tion, the MLVSS dropped sharply from 2900 ppm before carbon to 1500 ppm at
the conclusion of the study. After carbon addition was stopped, the MLVSS
returned to 2500 ppm with no change in the wasting rate.
Although considerable improvement in BOD removal was achieved at this
plant (Exhibit S), the full benefit of carbon was not realized because of
high influent oil concentrations (100+ ppm).
SUMMARY & CONCLUSIONS
• In summary, it has been shown that refinery wastes can be success-
fully treated with powdered carbon in activated sludge.
• Powdered carbon can improve organic removals, aid solids settling
and sludge handling, and provide protection from toxic or shock loadings.
In the face of widely varying influent organic or hydraulic loads, carbon
levels effluent quality.
• High density carbons are preferred to minimize carryover from secon-
dary clarifiers and to increase sludge compaction. Such carbons also require
less makeup to maintain the desired aerator equillibrium level since carbon
is lost from the system only during sludge wasting.
• Normally, one would think that the use of carbon would increase
costs. However, savings on defearners, coagulants, powder and labor can often
decrease operating expenses.
374
-------
BIOGRAPHY
Mr. Paschal DeJohn is Manager of
Purification Sales and Project Leader in Activated
Carbon at ICI United States Inc.; holding this
position since 1972. Mr. DeJohn holds a B.S.
degree in chemistry from Westchester State College
and an M.B.A. degree from Widener College,
Pennsylvania. Previous to his present position Mr.
DeJohn was Water & Wasrewater Treatment District
Engineer with Drew Chemical, Parsippany, New
Jersey. He is 1977 Technical Conference Chairman
for WWEMI and the 1978 General Conference
Chairman for this ame organization. Mr. DeJohn
has been involved with water and wastewater treat-
ment for the past twelve years.
James P. Black is Industry Coordinator for
Water Purification at ICI United States Inc. He
has a B.S. degree in chemical engineering from
the University of Texas at Austin. Prior employ-
ment included ICI's Research & Development
Laboratory, and Corporate Planning Staff.
375
-------
DEJOHN PAPER DISCUSSION
Ed Sebesta, Brown & Root; Have you observed any situations where you
mentioned sot e compounds are loosely adsorbed or difficultly adsorbed and
then quite easy to be desorbed? Have you ever seen any situations or heard
of situations where because of changing influent situations you may suddenly
desorb an accumulation of adsorbed materials and effect the system in that
way?
DeJohnt Potentially that can occur. I would think something like that would
be more prone to happen in a granular carbon system. You can design a
granular carbon system around this however. In the PACT systems we've been
involved in, I'm not aware of any that have desorbed back an accumulation of
adsorbed material. But there's always the possibility that this can happen.
REFERENCES
1. EPA Contract 68-01-2926, April 1, 1975.
2. Adams, A.D., "Improving Activated Sludge Treatment with Powdered Activa-
ted Carbon - Textiles" presented at the 6th Mid-Atlantic Industrial
Waste Conference, University of Delaware, November'15, 1975.
3. Esmond, S. E. and Petrasek, A. C., "Removal of Heavy Metals by Waste-
water Treatment Plants," presented at the WWEMA Industrial Water and
Pollution Conference, Chicago, Illinois, March 14-16, 1973.
4. Sigworth, E. A. and Smith, S. B., "Adsorption of Inorganic Compounds by
Activated Carbon," Journal AWWA, Water Technology/Quality. June, 1972
(p. 306).
5. Linstedt, K. D. et. al, "Trace Element Removals in Advanced Wastewater
Treatment Processes," Journal WPCF, 43, No. 7, 1507 (July, 1971).
6. Rizzo, Joyce A., "Case History: Use of Powdered Activated Carbon in an
Activated Sludge System", presented at the Open Forum on Management of
Petroleum Refinery Wastewaters, Tulsa, Oklahoma, 1976.
376
-------
EXHIBIT 1
CARBON ADSORPTION
>;
>
5
0
0
0
5
Slow
None
(or toxic)
Strong
Moderate
Weak
TRIPHENYL
PHOSPHATE
377
-------
EXHIBIT R
No Carbon First Carbon Period
COD .„ ...-,
Influent «J ppm 457 ppm
Effluent 17° PPm 135 ppm
% Removed 63 70
BOD
Influent 152 PPm 2]3 ppm
Effluent 15 ppm 15 ppm
% Removed 9° y3
SUSPENDED SOLIDS
Effluent 115 50
COD
Influent 343 ppm 444 ppm
Effluent 266 ppm 183 ppm
% Removed 23 59
BOD
Influent 152 ppm 227 ppm
Effluent 30 ppm 14 ppm
% Removed 80 94
SUSPENDED SOLIDS
Effluent 162 72
COD
Influent 367 ppm 379 ppm
Effluent 166 ppm 112 ppm
% Removed 55 70
BOD
Influent 188 ppm 207 ppm
Effluent 12 ppm 3 ppm
% Removed 94 99
SUSPENDED SOLIDS
Effluent 79 42
378
-------
EXHIBIT S
Flow: 8 MGD design, 6 MGD actual.
Carbon Dose: 500 ppm in aerator.
MLVSS: 2900 ppm before carbon
1500 ppm during carbon
(A 50% loss of aerator solids resulting from continuing same
volumetric wasting rate of the more dense carbon sludge.)
2500 ppm after carbon
• BOD removal - 55% before carbon addition
70-80% during carbon addition
60% after carbon addition
t Influent oil concentrations were so high (100+ ppm) during
test that effects of carbon were overshadowed.
379
-------
§ % BOD
REMOVED
100
90
80
70
60
50
40
30
20
10
EFFECT OF POWDERED CARBON ON
BOD REMOVAL
HYDRODARCO C
FIG. 1
5 10 30 50 70 90 95
% OF VALUES LESS THAN
99
-------
co
oo
1000
900
800
700
EFFLUENT 600
COD, PPM
500
400
300
200
100
EFFECT OF POWDERED CARBON ON
EFFLUENT COD
HYDRODARCO C
I I
III 1111
I 1
5 10
30 50 70
90 95
99
OF VALUES LESS THAN
-------
do
1000
900
800
700
EiFFLUENT 600
TOC, PPM
500
400
300
200
100
EFFECT OF POWDERED CARBON ON
EFFLUENT TOTAL ORGANIC CARBON
I
BLANK
HYDRODARCO C
I I
I I I » I I I
I
I
5 10
30 50 70
90 95
99
FIG. 3
% OF VALUES LESS THAN
-------
•
1000
900
800
700
EFFLUENT TC, 60°
500
400
300
200
100
EFFECT OF POWDERED CARBON ON
EFFLUENT TOTAL CARBON
HYDRODARCO C
5 10
30 50 70
90 95
99
% OF VALUES LESS THAN
-------
100
90
80
70
60
EFFLUENT OIL,
*• rtr-kn/i 50
PPM
40
30
20
10
0
EFFECT OF POWDERED CARBON ON
EFFLUENT OIL
BLANK
II I I I I I I
HYDRODARCO C
» I I I
5 10 30 50 70 90 95
% OF VALUES LESS THAN
99
-------
00
in
EFFLUENT
SUSPENDED
SOLIDS, PPM 500
EFFECT OF POWDERED CARBON ON
EFFLUENT SUSPENDED SOLIDS
HYDRODARCO C
5 10
50 70
90 95
99
FIG. 6
% OF VALUES LESS THAN
-------
u>
oo
zo
1.8
1.6
1.4
EFFLUENT 1.2
NITROGEN,
PPM 1-0
0.8
0.6
0.4
0.2
EFFECT OF POWDERED CARBON ON
EFFLUENT NITROGEN
BLANK
•P
I I I I I I
HYDRODARCO C
I I
FIG. 7
5 10 30 50 70 90 95
% OF VALUES LESS THAN
99
-------
oo
-j
2.0
1.8
1.6
1.4
EFFLUENT 1.2
PHOSPHOROUS,
PPM 1.0
0.8
0.6
0.4
0.2
EFFECT OF POWDERED CARBON ON
EFFLUENT PHOSPHOROUS
I I
BLANK |
I
HYDRODARCO C
\ \ \ I I I I
I I
5 10
30 50 70
90 95
99
FIG. 8
% OF VALUES LESS THAN
-------
00
oo
1000
900
800
700
EFFLUENT 600
ZINC, PPM
500
400
300
200
100
EFFECT OF POWDERED CARBON ON
EFFLUENT ZINC
HYDRODARCO C
II » I I I I I I
I I
5 10
30 50 70
90 95
99
FIG. 9
% OF VALUES LESS THAN
-------
CASE HISTORY
THE USE OF POWDERED ACTIVATED CARBON
WITH A BIODISC-FILTRATION PROCESS
FOR TREATMENT OF REFINERY WASTES
J.F. Dehnert
Environmental Director, Avon Refinery, Lion Oil Co.
ABSTRACT
A description of the development of a supplemental petroleum waste
water treating plant utilizing a Rotating Biological Surface Unit and
Powdered Activated Carbon followed by clarification and filtration
from laboratory and pilot plant studies through construction, start up
and operation to meet July 1977 NPDES discharge requirements.
Starting in 1972 Pilot Plant studies were conducted to compare the
performance of activated sludge, trickling filter, RBS and activated
carbon absorption processes in treating the Avon Refinery Waste Water.
The primary objection was to meet the EPA guideline discharge limits
plus the California State limits on fish toxicity. After several months
of study the treatment scheme of a RBS Unit plus solids removal facili-
ties was selected to meet the Federal standards and powdered activated
carbon was selected to meet the toxicity limits.
INTRODUCTION
With the adoption of the 1972 Amendments to the Clean Water Act,
the Staff at the Avon Refinery near San Francisco, then operated by
Phillips Petroleum Co. embarked on an investigative program to determine
the waste water treatment necessary to meet the limitations which would
eventually be placed on the refinery discharge through the NPDES Program.
At that time and until January 1975, the refinery discharge was already
subject to limitations imposed by the California Regional Water Quality
Control Board on 5-day BOD, oil and grease, settleable solids, suspended
solids, coliform and fish toxicity. In addition, limitations were in
effect on receiving water quality with respect to pH, dissolved oxygen,
undissociated NH/jOH, chromium, lead, H2S, Fish Toxicity, floating oil,
discoloration or turbidity and odor. The existing waste water treatment
included sour water stripping, API gravity separation, dissolved air
flotation and pH equalizing surge ponds followed by a 108-acre bio-
oxidation pond. The company had also segregated the refinery sewers so
that as much as possible of uncontaminated storm run off could bypass
the process water treatment. This storm run off was combined with the
bio-oxidation pond effluent and discharged in an underwater diffuser in
389
-------
the main channel of an arm of San Francisco Bay which receives the
Central California Valley drainage. This treating process is illustrated
in Figure 1.
INVESTIGATION
At the start of the investigation, no specific Federal limitations
had been determined and, therefore, the studies were mainly concerned
with a comparison of generally accepted methods of waste water treatment
to determine which of these would be most effective in removing or
reducing the known pollutants in the Refinery effluent. A primary
objective was to determine a treatment scheme that would result in a
waste water discharge that would meet the more restrictive California
State fish toxicity limitations that were being proposed.
In early 1972, working with an engineering contractor, three pilot
plants were installed at the Refinery with a slip stream of the waste
water going to the bio-oxidation pond serving as the raw feed. These
units were an air agitated activated sludge unit, a 21-foot trickle
filter and a four-column granular activated carbon unit preceded by a
mixed media filter. These plants were operated from June 1972 through
1973 in parallel or in series under wide variations of operating
conditions such as hydraulic loading, recycle rates and suspended solids
concentrations.
During this period of operation, different sets of tentative EPA
guidelines were issued for the Petroleum Industry. In each case, the
effluent from the biological treating pilot plants failed to meet these
guidelines and the proposed California fish toxicity standards. From
the experimental data it appeared the only way these limitations could
be met was by the use of activated carbon at high regeneration rates as
a final treating step. It appeared that the waste water contained some
non-biodegradable or at least "refractory" organic material as indicated
by COD and TOC tests.
In the Spring of 1973, information was received describing the
rotating biological surface units which were being proposed for treating
industrial waste as an improved alternate to the other biological treat-
ment systems. There were several advantages claimed for this process
such as high biomass concentration; low volume, high density sludge
production and low power requirement.
Subsequently, arrangements were made to install a four stage pilot
rotating biological surface unit at the refinery and to compare its
performance with the other pilot plants.
The RBS test program consisted basically of three periods determined
partly by a difference in the quality of raw waste water and partly by the
type of operation of the unit. During the first period, a direct comparison
was made between the trickle filter, the activated sludge and the RBS units
with the same feed going to all three units. For the second test period,
390
-------
the feed rate to the RBS unit was reduced to a very low figure to
establish nitrifying bacteria in the bio-mass. During the third period
a series of hydraulic loading tests were performed where the rate was '
varied from 1/2 gpm to 18 gpm representing hydraulic loading of 0.2 to
6.8 gal/day/sq.ft. of surface area.
As a result of the pilot plant study, it was concluded that the
removal of organic pollutants by the RBS unit compared favorably with
the trickling filter and the activated sludge processes and, for some
waste water parameters, the RBS unit appeared to be superior. The pilot
plant operation verified most of the claims made by the manufacturer,
particularly with respect to energy requirements and ease of operation.
The removal of organics by the RBS unit was very similar to the
trickle filter and the activated sludge with each unit able to achieve
about the same percentage removal and final concentrations in the final
effluent at their optimum operation.
A portion of the test program was devoted to establishing nitri-
fication and determining the relationship between hydraulic loading and
the degree of conversion of ammonia to nitrate. This was accomplished
by operating the unit at a very low feed rate and adding sodium
bicarbonate to increase the alkalinity. At the low rate, it was possible
to lower the ammonia concentration from 15 to 20 mg/1 to less than
1 mg/1; however, as the feed rate was increased, nitrification decreased
and eventually stopped altogether. Contrary to what was expected,over
50% of the conversion of ammonia to nitrate took place in the first stage.
From the data obtained, it was concluded that, if nitrification is
desired in a commerical unit, it would have to be designed for about
one-half the hydraulic loading that would be required for organics
removal.
One of the most noticeable differences between the RBS effluent and
the effluent from the other bio systems was the suspended solids content.
Although the suspended solids did increase with feed rate, even at
relatively high hydraulic loading, the RBS effluent had lower suspended
solids than the best operation of the other processes. At low feed
rates, the RBS effluent after 30 minutes of settling, exhibited a
sparkling appearance that was achieved on the other processes only by
filtration or activated carbon treatment of the effluent.
Static bioassays were conducted weekly on samples of the various
pilot plant effluents using the APHA standard methods to determine the
96-hr, median toxicity (TLm). Although the RBS unit was not the answer
to the toxicity problem at the Avon Refinery, in general, this effluent
was less toxic than the effluents from either the trickling filter or the
activated sludge. Activated carbon absorption remained as the only waste
water treatment,that would produce a completely non-toxic water (100%
survival) from the waste water stream.
Activated carbon treatment data are presented in Tables I through
VIII.
391
-------
Table I presents early data showing the effect on fish toxicity
of treating various RBS effluents with powdered activated carbon.
Tables II and III present data on the effect of powdered carbon on
other parameters and indicate that toxicity is improved although other
parameters are not greatly affected.
Tables IV and V illustrate the effect of pH changes on toxicity and
possibly the ability of carbon to absorb the toxicants.
Tables of VI and VII show comparisons of two different powdered
activated carbons and indicate that selection of the proper carbon source
can make a very great difference in the ultimate success of carbon
treatment.
Table VIII is a part of a very large table of data obtained on a
granular carbon test conducted over a six month period. It is presented
to further illustrate that long after the carbon was "exhausted" with
respect to COD removal it would continue to produce a non toxic effluent
and that it could be rejuvenated by a hot water backwash.
By the time these pilot plant studies were complete, the final
guidelines had been issued by EPA and the N.P.D.E.S. permit for the
refinery had been issued by the Regional Water Quality Control Board.
This permit outlined not only the discharge limits but a compliance
schedule for submitting a conceptual plan, completion of Engineering,
start of construction and completion of construction. At this time a
thorough review was made of all the accumulated pilot plant data and the
conceptual plan developed. It appeared that of the many parameters of
water quality, COD, suspended solids and fish toxicity would control the
design of the treatment system. Included in the consideration was the
volume of water which varied considerably with the seasonal storm water
entering the process or oily sewers, since practically all of the annual
rainfall in this location occurs between the first of November and the
first of April. Several alternate plans were considered but all were
basically a supplemental biotreatment, solids removal and activated
carbon treatment.
During the long period of monitoring the raw waste quality it became
evident that treating requirements were also cyclic in that both COD and
toxicity increased during the winter months but during part of the year
the discharge would probably meet the 1977 limits without much, if any,
additional treatment. In our studies with granular activated carbon we
noted that in several instances long after the carbon was "exhausted"
with respect to removal of COD it would still produce a non toxic water.
From this information it appeared the biotreating system should be capable
of handling wide variation in waste loading and that a carbon system should
be designed to be used only when necessary. From Capital cost considera-
tions, possible ease of handling and the indication that relatively small
quantities would be required, the decision was made to' use powdered carbon
on a periodic and throw away basis rather than use a granular bed system.
392
-------
In the summer of 1975 the conceptual plan illustrated in Exhibit
II was put out for bids to Engineering-Construction Firms as a "turn key"
project. As a result Engineering was completed by December 1975, field
construction started in February 1976, and completed by January 1, 1977,
all well within the compliance schedule.
At this writing the treating facilities are still in the process of
starting up primarily because of an extraordinary length of time required
for biomass to develop on the RBS units and then delays in correcting
minor difficulties with certain mechanical equipment, instruments and
electrical control systems.
Our principle concern was the difficulty in establishing the biomass.
During the pilot plant phase we had started up three different pilot
plants charging similar waste water and in all cases a good growth was
established within 3 to 4 weeks. However, in the case of our commerical
unit after four weeks there was only a very slight indication of biogrowth
on the first stage. It was determined that low water temperature and
relatively low soluble BOD were responsible for the apparent lack of
bioactivity.
With increased temperature and the addition of higher strength
waste water, supplied with a portable pump for 10 days, we observed an
increase in the growth of the biomass extending through all three stages.
Up to this point only the RBS and the clarifiers were in operation with
the plant effluent returning to the feed surge ponds. However, with the
establishment of biomass, the filter, polymer injection system and sludge
digester were all put into operation. The carbon system was operated for
a short period primarily to test it mechanically.
Only limited data has been obtained at this time however, they
indicate the plant will perform satisfactorily and the waste water
discharge will be in compliance with the July 1, 1977 limitation.
BIOGRAPHY James F. Dehnert
James F. Dehnert is the Environmental
Director for the Avon Refinery of Lion OP Com-
pany at Martinez, California. He has a B.S.
degree in Chemical Engineering and a B.S. de-
gree in Chemistry from Washington State University.
He has been employed at this Refinery for thirty
years with various assignments in Techanical
Service, Economic Planning and Unit Operations.
He served as an Area Operation Supervisor be-
fore becoming involved in Environmental assign-
ments.
393
-------
EXHIBIT I
u>
VO
QOH-EFCS
COOUIMS TOV/tf'S.S
R.EFIUŁR.Y
OPERATIOM
Tout. WATER.
STR.IPPER.
OXIDATIONS
V PONJD
^
V CLEA.M CA-WX^L. 1
' DISCHARGE
m.
-------
EXHIBIT II
Cn
-------
TABLE 1
TREATMENT OF RBS EFFLUENT WITH
POWDERED ACTIVATED CARBON
Test
No. 1
Sample
RBS Feed
RBS Eff
RBS Eff +20 ppm PAC
TLm or Survival*
74
90
90%
No. 2
RBS Feed
RBS Eff
RBS Eff + 10 ppm PAC
80
90%
100%
No. 3
RBS Feed 80
2nd Stage RBS 92
2nd Stage RBS +10 ppm PAC
90%
No. 4
RBS Feed
RBS Eff
RBS Eff + 10 ppm PAC
< 35
< 75
60%
No- 5
RBS Feed
RBS Eff
RBS Eff + 10 ppm PAC
33
64
60%
No. 6
RBS Feed
RBS Eff
RBS Eff + 20 ppm PAC
" + 35 ppm PAC
" + 50 ppm PAC
33
>69
90%
90%
100%
*Survival in undiluted waste
396
-------
TABLE II
ACTIVATED SLUDGE EFFLUENT TREATED WITH PAC
Parameter
Toxicity (% Survival)
COD mg/1
NH3(N) mg/1
Oil mg/1
Naphthenic Acids mg/1
Cr(T) mg/1
Cu mg/1
Zn mg/1
Act Sludge Eff
0 (24 hr)
108
35
0.2
1.5
0.02
0.20
0.03
Act Sludge Eff
+100 ppm PAC
100 (96 hr)
84
28
0.1
0.6
0.02
0.25
0.02
TABLE III
ACTIVATED SLUDGE EFFLUENT TREATED WITH PAC
Carbon Dosage (ppm)
Parameter
Toxicity (% Survival)
COD mg/1
Phenol mg/1
Oil mg/1
Naphthenic Acid mg/1
50
100
150
0
150
4.8
0.1
3.1
10
130
4.9
0.1
4.3
100
120
4.7
0.1
3.5
100
120
5.2
0.1
3.1
397
-------
TABLE IV
EFFECT OF pH ON TOXICITY
RBS Effluent @ 7.2 pH
40% Survival
RBS Eff Lowered to 6.5 pH
0% Survival
RBS Eff Raised to 8.5 pH
90% Survival
TABLE V
EFFECT OF pH ON CARBON TREATMENT
RBS Feed 7.1 pH
36 TLm
RBS Feed @ 6.5 pH + 30 ppm PAC 65 TLm
RBS Feed @ 7.0 pH + 30 ppm PAC 61 TLm
RBS Feed @ 7.5 pH + 30 ppm PAC 80 TLm
TABLE VI
COMPARISON OF TWO CARBONS
Sample
RBS Effluent (as is)
RBS Eff +60 ppm PAC-A
+90 ppm PAC-A
Toxicity
58 TLm
85 TLm
93 TLm
TOC (mg/1)
68
57
57
RBS Eff + 15. ppm PAC-B
+30 ppm PAC-B
+60 ppm PAC-B
93 TLm
90% Survival
100% Survival
398
57
54
44
-------
TABLE VII
COMPARISON OP TWO CARBONS
PAC-A Added Continuously to RBS Pilot Plant and
Additional PAC-A or PAC-B Added to RBS Effluent
RBS Eff '+ 45 ppm PAC-A 40 TLm
RBS Eff +195 ppm PAC-A 86 TLm
RBS Eff +45 ppm PAC-A + 50 ppm PAC-B 93 TLm
RBS Eff +45 ppm PAC-A +75 ppm PAC-B 100% Survival
399
-------
20
TABLE VIII
GRANULAR CARBON TREATMENT OP TRICKLING FILTER EFFLUENT
COD mg/1 Toxi
Day
1
2
3
4
5
6
7
8
9
10
Feed
180
180
180
200
260
300
-
-
270
320
No.l
70
80
100
120
210
260
240
250
270
300
Column
No. 2
60
30
50
50
140
160
170
200
220
250
No. 3
30
30
30
30
80
110
100
140
170
210
No. 4
10
30
40
50
90
100
70
100
140
170
Feed No . 1
100
100
90
70
0
61TLm
Toxicity (Survival)
Column
No. 2 No. 3
No. 4
220
210
170
160
100
34TLm
100
100
100
30
45
50
55
60
70
75
100
180
Hot
190
150
150
Hot
170
130
160
170
Water
160
140
150
Water
170
130
160
170
Wash
120
140
140
Wash
160
120
130
150
120
130
140
130
110
130
130
120
100
140
130
71TLm
45TLm
53TLm
59TLm
45TLm
100
80
0
100
0
80
0
100
100
100
100
400
-------
SESSION VIII
ADD-ON GRANULAR ACTIVATED CARBON
Chairman
Nicholas D. Sylvester
Professor of Chemical Engineering
University of Tulsa, Tulsa, Oklahoma
Speakers
Fred M. Pfeffer
W. Harrison and L. Raphael ian
"Pilot-Scale Effect on Specific Organics Reduction
and Common Wastewater Parameters"
R. H. Zanitsch
R. T. Lynch
"Granular Carbon Reactivation-State of the Art"
L. W. Crame
"Activated Sludge Enhancement: A Viable
Alternative to Tertiary Carbon Adsorption?"
401
-------
BIOGRAPHY Nicholas D. Sylvester
Nicholas D. Sylvester is Professor and Chairman
of the Resources Engineering Division - Chemical and
Petroleum 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 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
investigator 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.
402
-------
ORGANICS REDUCTION THROUGH ADD-ON ACTIVATED CARBON AT PILOT SCALE
Fred M. Pfeffer
U. S. Environmental Protection Agency
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820
Wyman Harrison and Leo Raphaelian
Argonne National Laboratory
Argonne, Illinois 60439
ABSTRACT
The current wastewater BATEA model for the petroleum refining industry is
the treatment sequence: activated sludge, mixed-media filtration, activated
carbon. In an effort to develop data to assist in evaluating the model for
specific organic compounds, the EPA (Ada, Oklahoma) entered into an Interagency
Agreement with ERDA (Argonne National Laboratory) in January 1975. In cooper-
ation with API, a .25 GPM pilot test was conducted at the SOHIO Refinery in
Toledo, Ohio. Argonne followed with GC/MS analysis of samples collected across
the treatment system to identify specific organics which are treatable versus
those which pass-through (refractories).
The EPA's involvement included: the mobile pilot plant, refinery selec-
tion, conduct of the field study, sample preparation, and reporting. Argonne1 s
analytical results showing a small overall reduction in organics by mixed-media
filtration and a large reduction by carbon adsorption are discussed.
INTRODUCTION
In January 1975, the EPA (Robert S. Kerr Environmental Research Laboratory,
Ada, Oklahoma) entered into an Interagency Agreement with ERDA (Argonne Nation-
al Laboratory, Chicago) to develop data to assist in evaluating the performance
of the BATEA model in the Development Document of 1974 (1). Since that time
the BATEA regulations (and hence the BATEA model) have been remanded by a
ruling of the 10th Circuit Court on the petition for revision of the guidelines
by the API (2). However, the requirements for reconsideration and reissuing
of guidelines as stipulated in the ruling, together with the mandates in PL-
92-500 (Sec. 301. d.) (3), and the Settlement Agreement between EPA and NRDC
(4), are added incentive to complete the work funded through this Interagency
Agreement.
The proposed BATEA model was fixed bed carbon adsorption added onto the
BPT model, which is biological treatment followed by granular media filtra-
tion. The specific treatment train selected for study was activated sludge,
403
-------
mixed-media filtration, and granular activated carbon. The performance cri-
terion was to be the reduction of major organic compounds identified in the
influent to the biological treatment system. Pursuant to the agreement,
Argonne would perform qualitative organics analyses on samples provided by
EPA. Pilot-scale filtration and carbon adsorption would be applied to the final
effluent from a full-scale refinery treatment system. The results would serve
as guidance for determining the need for larger-scale study and would not be
used in predicting the performance of a full-scale add-on carbon system.
Refinery Selection
Considerable time was allocated to refinery selection, as there was suf-
ficient funding to study only one refinery. Repeated discussions and meetings
were held with members of the API's W-20 Task Group to arrive at a "represent-
ative" refinery. It was agreed to acquire permission from a Class B refinery
whose final effluent quality met BPT, with the possible exception of suspended
solids. Other criteria would include intake water quality and variability,
refinery turnaround plans, and final effluent quality, raw waste loading, and
hydraulic detention times typifying the activated sludge process at a Class B
refinery.
Agreement was reached in September 1976, to conduct the study at SOHIO's
Toledo refinery. This is a Class B refinery (crude topping and catalytic
cracking) with coking, having a crude capacity of 120,000 BPSD. The treatment
train at that time consisted of the API Separator, dissolved air flotation
(DAF), activated sludge (extended aeration) having 16-18 hours detention, and
final clarification. The final effluent quality routinely satisfied BPT re-
quirements with the exception of suspended solids. The refinery treatment
system returned to steady state in November 1976, following a 1-month turna-
round period.
The Pilot Study
A 30' EPA mobile trailer was transported to Toledo and positioned near
the final clarifier. Facilities aboard the trailer included 6" I.D. glass
columns for filtration and carbon adsorption (Figure 1), a TOC analyzer for
monitoring organic carbon breakthrough, pumping and distribution capability,
and sampling gear. The sampling equipment, pumps, and distribution lines were
fabricated and installed such that the only materials in contact with water
moving through the pilot treatment system were stainless steel, glass, Teflon,
and polypropylene (Figure 1). Sampling points aboard the trailer were: 1)
SOHIO's final clarifier effluent, 2) pilot mixed-media filter effluent, and
3) pilot carbon column effluent (Figure 2). The two remaining sample points
were SOHIO's intake water and DAF effluent (Figures 3 & 4). These five points
were sampled and iced on 4-hour intervals for 24-hour compositing over a con-
secutive 4-day period. During the study, there were no significant changes
in recorded flows through the full-scale treatment system, as measured by the
hourly biofeed pumping rates.
Two parallel down-flow mixed-media filters were utilized such that while
one was operating for 24 hours, the second, having been backwashed, was ready
for use the next day (Figure 5). Figure 2 shows the configuration of the
404
-------
filtering bed: anthrafilt, sand, and gravel. The sand used has an Effective
Size = 0.2 mm and a Uniformity Coefficient = 4,5. Backwashing was accomplished
by alternately pulsing with air and pumping carbon column effluent.
Two up-flow carbon columns (Figure 6) were packed as shown in Figure 2
and operated in series to achieve a total bed depth of 6 feet. A constant
flow rate of 0,25 gpm was maintained, giving a residence time in the carbon
bed of 36 minutes. The carbon used was Calgon's Adsorption Service Carbon.
Calgon's analyses of a sample from the lot used at Toledo gave these results:
Apparent Density (gcc): 0.51
Molasses Number: 282
Iodine Number: 821
Sieve Result (mesh): 8x40
Attention was given to decontaminating material coming in contact with
water samples. All glassware was cleaned by firing, maintaining 550°C for
1-hour. Sample bottle caps contained Teflon liners which had been cleaned by
Soxlet extraction with methylene chloride—the solvent later used in the
laboratory for extracting the organics from the water samples.
Each daily composited sample set was transported in ice chests to Detroit
for air shipment. The samples arrived at RSKERL in Ada within 9 hours of
final compositing in Toledo.
Laboratory Phase
Performance of the full-scale biosystem and the add-on filtration/carbon
train for the common wastewater parameters is shown in Tables 1 & 2. Some
values are reported as less-than (<), reflecting lower limits of detectability
as a function of the sampling and analytical protocol.
Following the field study, the remaining responsibility of EPA was the
preparation of the composited water samples for organics analysis by Argonne.
This involved a tedious liquid-liquid extraction sequence using methylene
chloride. Again, all glassware was fired for organics decontamination. A
major problem was emulsion formation, requiring emulsion breaking and phase
separation by various techniques. Each organic extract was dried by passing
through anhydrous sodium sulfate and the solvent was stripped, resulting in
1-ml of concentrated extract which was sealed in a glass ampul. A period of
9 man-hours was involved in preparing each sample to the ampul stage; there
were 20 samples requiring this preparation.
Gas-Chromatography/Mass-Spectrometry (GC/MS)
The samples supplied to Argonne for analysis consisted of 1 mililiter
methylene chloride solutions of the acid, base, and neutral fractions com-
posited over the 4-day sampling interval.
Analysis of the specific organics in these fractions was performed on a
Hewlett-Packard Gas Chromatograph/Mass Spectrometer equipped with a -unta syp."^:
405
-------
and such peripheral equipment as a Zeta plotter and hard copy unit. Capillary
columns were used in the gas chromatograph. These columns allow considerably
greater separation and resolution of the organic components in a sample than
do standard packed columns. Capillary columns also provide increased sensi-
tivity and drastically reduced background from column bleed in the mass spec-
tra. Also as opposed to typical GC/MS operation, no separator was used to
remove the carrier gas. The outlet of the capillary column was connected
directly to the source of the.mass spectrometer and, therefore, there could
be no discrimination in the amount of each component reaching the mass spec-
trometer. That is, assuming that the individual components in the mixture are
not lost in the column, the effluent of the column and the amount of these
components reaching the source of the mass spectrometer is a true representa-
tion of the quantities of compounds injected on the column. Finally, a Grob-
type injection system was used in place of the inlet splitters typically used
with capillary columns. The Grob system avoids the loss of large amounts of
samples and the discrimination, typically found in split systems, of compon-
ents of the mixture. It permits the analysis of minute concentrations of the
specific organics present.
Figure 7 is a capillary column GC/MS total ion chromatogram of the neutral
fraction of the dissolved-air-flotation effluent. It can be seen that there
are over one hundred peaks or components in this fraction and that many of the
components are present in minute quantities; that is, of the order of 200 ppt
of the original water sample, assuming 100% extraction efficiencies. It was
found that the organics in this neutral fraction of the DAF effluent were
predominantly n-alkanes, alkyl benzenes, alkyl naphthalenes and polynuclear-
aromatic hydrocarbons.
The activated sludge treatment system reduced the concentration of the
organics in the DAF effluent by nearly 98%, as shown in Figure 8. It can be
seen from the graphs that the peak height of several of the peaks in the FC
effluent is approximately one-twentieth those in the DAF effluent, indicating
there is approximately a twenty-fold reduction in pollutants by the activated
sludge process.
Further reductions in organics were accomplished by the multi-media fil-
ter arid the activated carbon column as shown in Figure 9. The concentration
of the largest peaks of the compounds refractory to the add-on treatment
system is of the order of 10 ppb. The percent reduction of the major classes
of organics by the multi-media filter and the activated carbon column is as
follows:
Compound . % Reduction
Alkanes 70-98
Alkyl Benzenes 35-90
Indenes 50-60
Indanes 76-96
Naphthalene 66
Alkyl Naphthalenes 65-90
Anthracene/Phenanthrene 86-93
Alkyl Anthracenes/Phenanthrenes 89-98
Ot-her PNAs 96-98
4Q6
-------
It can be seen that there is generally greater than fifty percent reduc-
tion in these classes of organic compounds.
Work has not yet been completed on the acid and base fractions. Results
will be available in a few months. The results of this study will be published
as EPA and Argonne reports and as such will be available to the public.
It is expected that sufficient funds will be forthcoming in the next
fiscal year to search and manipulate the data stored on disc for those consent-
decree organics that may be present in these samples.
REFERENCES
(1) Development Document for Effluent Limitations, Guidelines and New Source
Performance Standards for the Petroleum Refining Point Source Category,
EPA-440/l-74-014-a, April 1974.
(2) Ruling of the 10th Circuit Court, Denver, on the suit of EPA by API,
handed down August 11, 1976.
(3)> Public Law 92-500, 92nd Congress, S.2770, October 18, 1972.
(4) Settlement Agreement in the U.S. District Court for the District of
Columbia between the Natural Resources Defense Council and the U.S."
EPA, Civil Action No. 2153-73, June 7, 1976.
ACKNOWLEDGEMENT
The authors wish to acknowledge the assistance of the Calgon Corporation
relating to activated carbon and the efforts of the API's Water Quality Com-
mittee and W-20 Task Group in arriving at a suitable refinery. We wish to
thank Messrs. C. Tome, L.S. Van Loon, and J.H. Walters for assistance in the
wastewater sampling program and Mrs. C.S. Chow for help with the GC/MS
analyses. Most important, the study would not have been possible without the
cooperation of SOHIO personnel at the refinery in Toledo and in the Department
of Environmental Affairs in Cleveland.
DISCUSSION
Peter J. Foley, Mobil Oil; Would you comment on the contributions of the
filter and the carbon columns in reducing organics?
Raphael fan; Although I do not have the exact numbers at hand, it appears to me that the
multi-media filter had little or no effect on the concentration of organics whereas the
activated carbon removed appreciable amounts of organics.
L_AjV\cConomy, Calgon Corp.; The naphthalene removal was only 66% as compared
to more than 90% for other PNA's, why?
Ra^haeliaru These numbers are approximate figures based on the average of peak areas of
individual components. Generally, one can say that those organics that have a long
407
-------
alkyl chain such as an alkane or an alkylated PNA are adsorbed on the carbon whereas
parent PNA's are not. This is, of course, a crude approximation. We are still trying to
get better gas chromatograms because all of these results are dependent upon how well you
separate the compounds.
N. F. Seppi, Marathon Oil; Please comment on methylene chloride purification - also
what about decomposition products from methylene chloride under basic conditions?
Pfeffer; Regarding the purity of methylene chloride, we relied upon the glass
distillation procedures of the manufacturer (Burdick and Jackson). In addition,
Argonne received a blank extract obtained by taking a high purity water and
performing the acids, neutrals and basic extractions. This blank would also
account for laboratory contaminations. I cannot offer any information about
alkaline decomposition products.
Anonymous: Is there additional data from industry on the study at Toledo?
Pfeffer; I do not know. Both Exxon and SOHIO conducted parallel work to our
own, presumably into the realm of GC mass-spec. We would entertain comparing
notes with Exxon and SOHIO at some later date in order to validate what actu-
ally took place in Toledo.
Judith Thatcher. API: I noticed that the TOG of the influent water is very
close to the TOG of the final effluent. Have you done any identification in
the organics in the influent water to the refinery?
Pfeffer; We are looking at it, but haven't identified all the components yet.
Arthur J. Raymond. Sun Oil Co.; What phase was used on the capillary columns,
and did you notice that your highly branched-chains were degraded much faster
than the less branched? Also, was benzene degraded much faster than toluene
and xylene?
Raphaelian; I don't understand what you mean by degraded.
Raymond: Did they decompose faster or disappear or reduce? Not in the column
but in your system as you went from the influent water to the final clarifier.
If you had percent reductions, which compounds went faster than others?
Raphaelian; I am still putting all this data together. However, I can say
that it appears that the branched-chained alkanes, which were present in smaller
quantities than the straight-chained alkanes, were not removed as well by the
treatment system as the straight-chained alkanes. Because of the minute
quantities of pollutants present, I am presently doing single-ion monitoring
to try to get a better idea of the percent reduction across the treatment
system.
Raymond; What phase did you use on a capillary column? What coating?
Raphaelian; For the work presented in this talk, OV-101 was the liquid phase
and the columns were wall coated open tubular (WCOT) and not support coated
408
-------
open tubular (SCOT) capillary columns. It is difficult with OV-101 to get
symmetrical peaks with polar compounds, that is, peaks without tailing. I
used FFAP capillary columns for the acid and base fractions. By the way we
see a variety of alkylated phenols in the acid fraction. '
Ed Sebesta. Brown & Root! I noticed that for TSS there is no decrease across
the filter. Do you have any explanation for that?
Pfeffer: My only explanation is that considering the flow rate and sand spec-
ifications, the TSS coming from the final clarifier were such that the filter
was ineffective. Also, at the 10 mg/l level, differences are probably within the experi-
mental error of the test procedure.
BIOGRAPHIES
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.
Wyman Harrison holds the SB, SM, and a PhD degree in
Geophysics from the University of Chicago. He is
currently the Assistant Director of Applied Geoscience
and Engineering, Energy and Environmental Systems
Division at the Argonne National Laboratory, Argonne,
Illinois.
Leo Raphaelian holds the AB degree from Harvard
University and the MA and a PhD degree in Chemistry
from Yale University. He is currently Manager of
Environmental Sciences, Energy and Environmental
Systems Division at the Argonne National Laboratory,
Argonne, Illinois.
409
-------
Table 1. DAILY PERFORMANCE FOR COMMON WASTEWATER PARAMETERS
Oil & Grease
Cyanide
Phenol
COD
BOD
TOC
TSS
Oil & Grease
Cyanide
Phenol
COD
BOD
TOC
TSS
MG/L INTAKE MG/L DAF EFFLUENT MG/L FC EFFLUENT
Day 1 Day 2 Day 3 Day 4
<10 <10 <10 <10
<0.02 <0.02 <0.02 <0.02
0.03 <0.01 0.03 0.01
<15 18 <15 <15
<10 <10 14 <10
19 19 17 15
35 29 11 <10
Day 1 Day 2 Day 3 Day 4
22 33 21 22
0.19 0.25 0.31
320 260 520 450
122 172 154 154
82 127 108 96
39 56 72 60
31 56 37 30
Day 1 Day 2 Day 3 Day 4
<10 <10 <10 <10
0.16 0.12 0.20 0.10
0.02 0.01 0.04 0.02
49 50 51 44
<10 15 21 24
22 29 27 17
12 <10 <10 <10
MG/L FILTER EFFLUENT MG/L CARBON EFFLUENT
Day 1 Day 2 Day 3 Day 4
<10 <10 <10 <10
0.16 0.15 0.20 0.10
0.02 0.01 0.02 0.02
42 38 51 44
<10 11 22 27
19 26 23 18
<10 <10 12 12
Day 1 Day 2 Day 3 Day 4
<10 <10 <10 <10
<0.02 <0.02 <0.02 <0.02
<0.01 <0.01 <0.01 <0.01
<15 <15 <15 <15
<10 <10 <10 <10
10 12 11 <5
<10 <10 <10 <10
-------
Table 2. AVERAGE PERFORMANCE OVER 4-DAY STUDY PERIOD
FOR COMMON WASTEWATER PARAMETERS
Oil & Grease
Cyanide
Phenol
COD
BOD
TOC
TSS
MG/L INTAKE
<0.02
0.02
18
21
MG/L DAF
24
0.25
390
150
103
57
38
MG/L FC MG/L FILTER MG/L CARBON
0.14
0.02
48
17
24
0.15
0.02
44
17
22
<0.02
<0.01
411
-------
Is)
Figure 1. CARBON COLUMNS
Figure 3. SAMPLING POINT: PLANT INTAKE
-------
WASTE
00
.*'•'* '/«'«
;« «t-«'A;
•T-SAND
J*' - 4 ' «'
A-:4 *«*>.'
FINAL CLARIFIER
FIGURE 2 - PILOT TREATMENT FACILITY
-------
Figure 4. SAMPLING POINT: DAF EFFLUENT
-------
Figure 5. MIXED MEDIA FILTERS
Figure 6. CARBON COLUMNS
-------
Ill
DAF EFFLUENT
20
40
' ' '
eo
80
100 120
TIME (min)
Figure 7. Total Ion Chromatogram of DAF Effluent (Neutral Fraction, Four-Day Composite)
-------
OAF EFFLUENT
jii^^
i—i—i—i—i
T—i 1 1 1 r
T 1 1 1 1 r
V
FC EFFLUENT
(IOX)
0
20
40
60
TIME (min.)
80
i n
100 120
Figure 8. Total Ion Chromatograms of DAF Effluent and FC Effluent (Neutral Fraction
Four-Day Composite) '
-------
ACTIVATED-CARBON
EFFLUENT (IOX)
^^^^^
oo
FINAL CLARIFIER
EFFLUENT
MAA'
o
I I
20
I I
40
60
TIME (mm.)
III
80
100
120
Figure 9. Total Ion Chromatograms of the Activated-Carbon and Final-Clarifier Effluents (Neutral Fraction
Four-Day Composite) '
-------
"GRANULAR CARBON REACTIVATION: STATE-OF-THE-ART"
R. H. Zanitsch
Engineering Director, Calgon Environmental Systems Division
R. T. Lynch
Process Engineer, Calgon Corporation
Use of granular activated carbon for treatment of industrial waste-
water is receiving widespread acceptance. In the past several years, 100
adsorption systems have been installed in industrial plants. Applications
range from dye plant wastewater reuse to removal of toxic materials. Gran-
ular carbon is being used to treat flows as low as 1,000 gallons per day
to as high as 20,000,000 gallons per day in industrial waste applications.
It is being employed as a pretreatment step to remove toxic materials prior
to biological treatment, as the main treatment process and for tertiary
treatment of biological plant effluents.
In most industrial wastewater applications, cost of virgin carbon pro-
hibits using it on a throw-away basis. Chemical regeneration is feasible
in only a limited number of applications and regenerant disposal remains a
problem. Thermal reactivation is in most cases complete, efficient, and
economical whether it is performed on-site or on a contract basis at a
central reactivation facility.
The technology of reactivation with industrial waste carbons has de-
veloped in only the last ten years. There are now approximately twenty re-
activation systems installed in the United States which are reactivating
industrial wastewater carbons.
New thermal reactivation processes (such as fluidized beds and elect-
ric furnaces) are now being developed but no commercial experience with
industrial wastewater carbons has been developed in the United States.
For the purposes of this presentation, we will discuss our experience with
the design and operation of multiple hearth furnaces and rotary kilns as
they relate to industrial wastewater applications.
THE THERMAL REACTIVATION PROCESS
Granular carbon is usually wet when fed to the reactivation furnace.
Water concentration is a function of carbon size, water temperature dur-
ing the dewatering step, and the amount of adsorbate on the carbon. In
practice, moisture content varies from 40 to 50 percent on a wet spent
basis.
The reactivation process can be divided into three steps:
1. Evaporation of moisture on the carbon (Drying).
419
-------
2. Destructive distillation of organics resulting in pyrolysis of
a portion of the carbon from the organic materials (Baking).
3. Activation of the carbon by selectively burning carbon deposited
during the organic removal step (Activation).
During the drying step, carbon temperature is increased to approxi-
mately 212°F (100°C) and moisture evaporates into the gas phase. As moist-
ure evaporates it is also possible for highly volatile organics to be steam
distilled.
The second step is termed baking or pyrolysis of the adsorbate. Dur-
ing this step, carbon temperature increases to approximately 1200-1400°F
(649-760°C). A portion of the organic molecules are thermally cracked to
produce gaseous hydrocarbons which are driven off. The remaining lower
molecular weight organics are distilled. During this process, a carbon char
is deposited in the pore structure of the original activated carbon.
The final step is activation of the carbon - a chemical reactivation
whereby carbon char deposited during the baking step is combusted along
with a small amount of the original carbon. By this time, temperatures
are in the range of 1600-1800°F (871-982°C).
Since the fixed carbon and the granules are both carbon, the process
requires that fixed carbon be selectively gasified with minimum gasifi-
cation of the granular carbon. Steam is added to the furnace and oxygen
concentration is controlled to promote gasification of the fixed carbon
while minimizing burning of the original granular carbon.
REACTIVATION SYSTEM DESCRIPTION
The basic sequence for thermal reactivation is as follows: (See Ex-
hibit 1).
Spent carbon is removed from the adsorbers and transferred as a slurry
to a spent carbon storage tank. Spent carbon is then transferred to an
elevated furnace feed tank from which it is metered, at a controlled rate,
to a dewatering screw. The dewatering screw is an inclined screw conveyor
which serves the dual purpose of gravity draining slurry water from the
granular carbon and providing a water seal for the top of the furnace. A
timer operated valve is used to meter carbon to the dewatering screw.
Drained, but wet, spent carbon then gravity flows into the furnace where
it is dried, baked, and reactivated as discussed earlier. Reactivated
carbon exits the furnace by gravity and enters a quench tank. The quench
tank serves the dual purpose of wetting the reactivated carbon and provid-
ing a bottom seal for the furnace. The carbon is then transferred to a
reactivated carbon storage tank from which it is then returned to the ad-
sorbers as needed. In most industrial waste applications, an afterburner
and scrubber are provided for destroying organics and removing residual
particulates from the furnace off-gases.
420
-------
The spent carbon storage tank should be designed for five to ten days
storage of carbon in order to allow for routine furnace maintenance and un-
scheduled shut-downs. This is usually a lined carbon steel tank with a
cone bottom to facilitate carbon flow. The reactivated carbon storage tank
is usually sized in the same manner with the same materials of construction.
In the case of the reactivated carbon storage tank, facilities should be
provided for adding virgin makeup carbon to the system as required.
The furnace feed tank is usually sized for at least one shift of op-
eration. The feed tank insures a constant carbon feed to the furnace in-
dependent of the large storage system. This tank is usually a cone-bottom,
lined carbon steel tank.
Spent and reactivated carbon are transferred in slurry form using
either eductors, blowcases or slurry pumps. In the case of eductors and
pumps, dilution water must be provided in order to reduce slurry concen-
tration to less than one pound per gallon to minimize carbon abrasion and
line erosion. Eductors are generally applicable in non-corrosive services
where static head is not great. Pumps can be used satisfactorily in high-
head applications, but are subject to erosion and plugging. Eductors and
pumps require use of dilution water which necessitates installation of
water recycle systems. The blowcase is an efficient method of transferring
carbon. Both air and water have been used to pressurize blowcases. In the
case of a blowcase, carbon is transferred in a much denser slurry (three
pounds per gallon) and, therefore, care must be taken to maintain control
over line velocities to minimize abrasion and wear. Material for carbon
slurry lines should be compatible with the wastewater. As long as slurry
lines are flushed free of carbon after each transfer, galvanic corrosion
of carbon steel lines will not be a problem; however, if the wastewater is
corrosive, more exotic materials of construction should be used. All car-
bon slurry lines should be equipped with flush connections to facilitate
flushing and unplugging.
CAPITAL AND OPERATING COST ESTIMATES
Based on our experience with the design, installation, and operation
of multiple hearth furnaces and rotary kilns for reactivating industrial
waste carbons, we have estimated the installed cost of reactivation systems
to reactivate 5,000, 10,000, 30,000, and 60,000 pounds per day. (See Ex-
hibit 2); The capital cost curve shown in Exhibit 3 represents a total
installed cost including all equipment, site preparation, foundations, in-
stallation, startup, and indirects. We have assumed that necessary util-
ities and off-site facilities are available at the battery limits. As you
can see, we estimate the total installed cost of a 10,000 pound per day
reactivation system to be approximately $1.25 million plus or minus 20 per-
cent. The time required to design, procure, install, and startup a re-
activation system is usually estimated to be two years assuming a twelve-
Jnonth delivery time on the furnace and associated equipment.
We have also estimated direct operating costs for reactivating 5,000,
421
-------
10,000, 30,000, and 60,000 pounds per day of industrial wastewater carbons as
shown in Exhibit 4.
In order to develop these costs, the following elements were consider-
ed:
1. Labor was estimated to be one operator per shift at a rate of
$10/hour. An allowance of 25 percent of the labor cost for
supervision was also included.
2. Fuel was estimated at 8,000 BTU's per pound at a cost of $3/million
BTU's. This estimate includes afterburner operation and an allow-
ance for inefficiencies due to interruptions and reduced feed
rates. Approximately half of the fuel consumption is required for
the afterburner and idling.
3. Power costs for the reactivation system are minimal and were as-
sumed to cost $0.03/KWH.
4. Steam costs were based on an average demand of one pound of steam
per pound of carbon for reactivation at a cost of $4/1,000 pounds.
5. Maintenance costs for an industrial wastewater application can
range from 8 to 15 percent of the reactivation system cost per
year. For this estimate, we assumed a maintenance cost of 8 per-
cent per year.
6. Makeup carbon cbsts were based on an average carbon loss rate of
7 percent and a virgin carbon cost of $0.57/pound delivered.
Carbon losses can range from as low as 3 percent to greater
than 10 percent depending on design and operation of the system.
Most industrial waste systems operate in the 5 to 7 percent loss
range. Makeup carbon costs represent the highest individual cost
element in the direct operating cost estimate and, therefore, all
efforts should be made to minimize carbon losses through good de-
sign and operation.
7. A general plant overhead of 10 percent of the above cost was al-
lowed to cover such items as insurance, taxes, monitoring, ac-
counting, and administration.
As can be seen from Exhibit 4, the direct operating cost for a re-
activation system handling industrial wastewater carbons ranges from
$0.11 to $0.19/pound over the range investigated. This does not include
depreciation or amortization of investment. The economies of scale are
obvious. We feel these costs can r,Łnge plus or minus 20 percent, but in
general, reflect the cost to operate a reactivation system on industrial
waste applications.
422
-------
MULTIPLE HEARTH FURNACE
Exhibit 5 is a cross-sectional view of a multiple hearth furnace. The
furnace consists of a cylindrical refractory-lined steel shell containing
several refractory hearths and a central rotating shaft to which rabble arms
are attached. From four to eight hearths are used in carbon reactivation
furnaces. The center shaft and rabble arms are cooled by air supplied by
a centrifugal blower discharging air through a housing into the bottom of
the shaft. A sand seal at the top of the furnace and a sand or water seal
at the bottom are used to seal the furnace against introduction of extran-
eous air.
In operation, wet spent carbon is introduced through a chute into the
outside of the top hearth of the furnace. The rabble arms are equipped
with solid alloy rabble teeth which rake the carbon towards the center where
it drops to the hearth below. The teeth on the rabble arms are arranged to
move the carbon in a spiral path. The action is gentle to minimize at-
trition. The top hearth is termed an "in" hearth since carbon flow is in-
ward.
The second hearth is consequently an "out" hearth where the carbon is
moved outward by the rabble teeth. Out hearths have a series of holes
around the periphery of the hearth through which the carbon drops to the
next lower hearth.
In this manner, carbon passes through the furnace until it is finally
discharged through a chute in the bottom hearth into the water filled
quench tank. The chute extends under the water level in the quench tank
to provide a seal.
Drying is accomplished in the upper one-third of the furnace. Dis-
tillation and pyrolysis of the adsorbate occurs in the next one-third.
Activation of the carbon is completed in the bottom one-third of the furn-
ace.
Burners are mounted tangentially on the furnace shell in burner boxes.
Usually burners are placed on the bottom two or three hearths and on one
upper hearth below the lowest drying hearth. However, if desired, burners
can be mounted on any hearth including the drying hearths.
On small furnaces, two burners per fired hearth are used. On larger
furnaces, three burners are installed. The burners are of the nozzle-mix
type burning fuel oil or natural gas. Dual fuel burners are commonly
employed to burn gas when it is available and fuel oil at other times.
Steam addition ports are provided on the bottom two or three hearths
to add steam for control of the reactivation process.
The center shaft is driven through a variable speed drive at 0.5 to
2.5 rpm.
423
-------
A number of furnaces have been installed with integral or "0" hearth
afterburners. This is less costly from a capital cost standpoint than the
separate afterburner.
DIRECT FIRED ROTARY KILN
Exhibit 6 is a simplified sketch showing a direct-fired rotary kiln.
The kiln is a refractory lined steel shell enclosed on each end with re-
fractory lined stationary hoods. This sketch depicts a counter-current
operation where gas flow is opposite the carbon flow. Co-current operation
is also possible and one carbon reactivation kiln is currently operating
in this manner.
The kiln is mounted on two or three sets of trunions depending on the
length of the unit. The kiln is sloped from the feed to the discharge end
and one set of thrust rolls are used to maintain the kiln in position on
the trunions. Proper training and alignment of trunions is important to
minimize excessive wear of the trunions and tires.
The kiln is driven through a variable speed drive coupled to a speed
reducer and pinion gear which meshes with a bull or girt gear mounted on
the kiln shell. The kiln is equipped at each end with hoods. Rotary seals
are used to seal between the rotating kiln shell and the stationary hoods.
The hoods are refractory lined.
A feed screw or chute is used to feed wet carbon into the kiln. Flights
are usually employed to advance the damp carbon and to shower the carbon in
the feed end to obtain high heat transfer rates during the evaporation step.
Flights are also used in the first portion of the baking step up to a point
where the temperature reaches approximately 1200-1400°F (649-760°C). Ma-
terial of construction for the flights is a function of carbon corrosive-
ness and reactivation conditions in the kiln.
The hot reactivated carbon, at a temperature of 1600-1800°F (871-982°C),
discharges from the kiln and falls down the discharge chute into a water-
filled quench tank. The discharge chute extends under the water level in
the quench tank to form a seal to eliminate air leakage into the kiln.
A burner is mounted in the discharge hood to provide heat for the re-
activation process. Either fuel oil or gas may be burned. The burner
air-to-gas ratio is adjusted to minimize oxygen concentration in the kiln.
A steam addition port also is provided in the discharge hood to admit steam
into the kiln for control of the reactivation process.
The exhaust gases, at a temperature of 500-800°F (260-427°C), leave
the kiln through a duct connected to the feed hood. In most installations,
gases are passed through an afterburner for complete combustion of organics
and burning of carbon fines swept out of the kiln. New installations, as
is the case with the multiple hearth furnace, will probably require in-
stallation of a wet scrubber to meet air pollution codes in most areas of
the country.
424
-------
In a rotary kiln, reactivation process is controlled by varying the
kiln speed to provide adequate retention time and by adjusting the steam
rate and burner temperature at the settings required for the particular car-
bon to be reactivated. A steam rate of 0.6-1.2 pounds per pound of carbon
and a temperature of 1600-2000°F (871-1093°C) are ranges encountered in
practice.
OPERATING AND MAINTENANCE PROBLEMS
A number of unique operating and maintenance problems have been ex-
perienced in reactivating industrial wastewater carbons. These problems
include:
Corrosion
Slagging
Poor Reactivated Carbon Quality
High Carbon Losses
Feed Interruptions
Hearth Failures
Slurry Line Erosion and Corrosion
Corrosion
Selection of proper construction materials for carbon storage and
handling systems is very important. We have found lined carbon steel
tanks to be satisfactory, but proper selection and application of lining
material is extremely important. Lining material should be corrosion
and abrasion resistant. We recommend thorough corrosion coupon testing
prior to making a final selection. Erosion of lining material at carbon
outlet nozzles, followed by corrosion of the metal, has been a problem.
We have installed sacrificial wear plates or stainless steel cones on tanks
in order to minimize this problem. Dewatering screws and quench tanks are
generally constructed of 304 or 316L stainless steel. In general, these
materials are satisfactory for most applications. However, the dewatering
screw is exposed to the spent carbon slurry and, therefore, its material
must be compatible with the wastewater. Corrosion of rabble arms and teeth
in multiple hearth furnaces and lifting and drying flights in rotary kilns
can be a problem when handling chlorinated hydrocarbons and organic sulfur
compounds. Special attention must be given to material selection to mini-
mize this problem.
flagging
Formation of clinkers and slag in the furnace is generally a function
of sodium and/or organic phosphate content of the spent carbon. Slag
formation can be minimized by pretreatment of carbon and maintenance of
proper furnace conditions. Formation of slag can generally be attributed
to constituents in the water contained in the pores of the carbon as it
425
-------
enters the furnace. These chemicals react with alumina and silica in the
furnace refractory resulting in slag formation.
Hearth Failures
Hearth failures in multiple hearth furnaces can generally be attributed
to cyclic operation. Frequent feed interruptions, resulting in temperature
excursions on the upper hearths, will weaken hearths and ultimately lead to
failure. By minimizing the number of feed interruptions and maintaining
continuous furnace operation, upper hearth life can be maintained for three
to five years. Another problem leading to hearth failure is brick attack
by sodium compounds; which leads to the slagging problem discussed earlier.
Also, improper dewatering or improper operation of the dewatering screw,
which would result in excessive amounts of water entering the top hearth,
can result in thermal shock which leads to failure. In general, hearth life
is a function of the operating philosophy of the furnace. If frequent feed
interruptions due to improper carbon feed system design or cyclic operation
are encountered, poor hearth life can be expected.
Carbon Losses
As mentioned earlier, makeup carbon cost is the single most important
cost element for a reactivation facility. By properly designing the ad-
sorbers, the carbon transfer and handling systems and the carbon storage
and reactivation systems, losses can be controlled at the 5-7 percent level.
Within the reactivation furnace itself, carbon losses should not exceed
1-3 percent. Carbon which is lost is due to oxidation during the activation
step. This can be controlled by maintaining oxygen levels in the activation
zone at 0-2 percent, or roughly that required for destruction of organics
without sacrificing carbon. Most carbon losses in a granular carbon system
occur due to backwashing of carbon in adsorbers, abrasion in slurry lines,
spillage, and carryover in overflow lines. These losses can all be mini-
mized by proper design of the basic system. Care must be given to overflow
rates, backwash rates, and slurry line velocities, and frequent checks must
be made to see that good housekeeping and operating techniques are being
followed.
Slurry Line Erosion and Corrosion
As mentioned earlier, construction material of spent carbon slurry
lines should be compatible with the wastewater in order to minimize cor-
rosion. If spent carbon or reactivated carbon is allowed to accumulate in
a carbon steel slurry line, galvanic corrosion can be expected. Therefore,
flushing of all slurry lines after each transfer is recommended. If the
wastewater is extremely corrosive, we recommend lined steel or stainless
steel slurry piping be considered.
Erosion of slurry lines can be attributed to excessive transfer
426
-------
velocities. We recommend a slurry line velocity of 3-5 feet per second
which is sufficient to prevent settling and minimize abrasion. Also
slurry lines should be as direct as possible with a minimum number of
bends. We recommend that long-radius bends be used to minimize abrasion.
Also, we recommend that all bends be accessible for periodic inspection
and replacement. Flush connections should be provided at frequent inter-
vals on all slurry lines in case a line becomes plugged.
FURNACE SELECTION CRITERIA
Choice of a reactivation furnace depends on many factors. A thorough
analysis of each type of equipment plus reactivation characteristics of
the carbon are necessary to make a final decision on which piece of equip-
ment to use.
.Both multiple hearth furnaces and kilns are being employed to react-
ivate granular activated carbon used in industrial wastewater treatment.
The quality of reactivated carbon that can be achieved is the same for both
units.
Major parameters influencing the selection of a reactivation furnace
are as follows:
1. Capital Cost - Total installed costs for either a multiple hearth
furnace or a rotary kiln are approximately the same. The purchase
price is generally higher for a multiple hearth furnace. However,
the installation costs are lower which tends to make installed
costs the same. Site preparation, foundations, and structural
costs are higher for a rotary kiln because of the greater area re-
quired to install a kiln.
2. Area Requirements - Area requirements are much greater for a rotary
kiln than for a multiple hearth furnace. The kiln also requires
more foundations and structural steel for walkways than a multiple
hearth furnace. The multiple hearth furnace is higher than a kiln
which means more structural steel is required to support the car-
bon feed equipment.
3. Fuel Consumption - Fuel consumption is higher in a rotary kiln be-
cause of higher heat losses. In a multiple hearth furnace, in-
sulation is used behind the wall brick to minimize heat loss.
This is not possible in a rotary kiln. Surface area is also higher
in a rotary kiln than a multiple hearth furnace of equivalent ca-
pacity. Fuel consumption for each will be in the following ranges
depending on capacity and operating rate as a fraction of rated
capacity.
427
-------
BTU/LB Carbon*
Multiple Hearth Furnace 2500-4500
Rotary Kiln 3500-8000
*Does not include afterburner fuel requirements.
4. Capacity Turndown - Capacity turndown ratio is defined as the per-
cent of rated capacity at which the furnace can be operated while
producing good reactivated carbon with reasonable carbon loss.
Capacity turndown for the equipment being evaluated in this paper
are as follows:
Multiple Hearth Furnace - 33 Percent
Rotary Kiln - 50 Percent
The multiple hearth furnace can be operated at a lower fractional
capacity because of the greater degree of control that can be ob-
tained in various zones of the furnace. In a rotary kiln, with
only one burner and one steam addition point, kiln speed is the
major parameter that can be varied to operate at lower capacities.
5. Degree of Control - Better reactivation process control can be
achieved in a multiple hearth furnace because the furnace is di-
vided into distinct zones according to the number of hearths in
the furnace. Each hearth can be equipped with burners, steam
addition, and air addition which can be controlled independently.
Thus, it is possible to control temperature and vary the atmosphere
in each hearth to optimize carbon reactivation.
In a rotary kiln, the steam port, and burner can only be mounted in
the firing end of the kiln. With this arrangement, the degree of
control that can be achieved is less than in a multiple hearth
furnace. In a properly sized kiln, this is not a distinct dis-
advantage and good carbon reactivation can be achieved. However,
as discussed previously, capacity turndown is not as great in a
kiln.
6. Corrosion and Slag - Many industrial waste streams contain in-
organic impurities which can cause corrosion and slag formation in
the reactivation furnace. These impurities are mostly chloride
and sulfur salts of calcium and sodium. The multiple hearth furn-
ace has more exposed alloy parts than a kiln and is, therefore,
more susceptible to corrosion. Rabble teeth and arms are expen-
sive, long delivery castings as opposed to the alloy flights in a
kiln which are fabricated from readily available plates. Also,
considerable corrosion of flights can occur in a kiln before re-
placement is required.
Slag buildup in a multiple hearth furnace will require periodic
428
-------
shutdowns to remove accumulated material. Slag in a rotary
kiln will be discharged, with the reactivated carbon, into the
quench tank where it can be removed without shutting down the
process.
7- Maintenance - Experience with reactivating industrial wastewater
carbons indicates higher maintenance costs in a multiple hearth
furnace. The factors responsible are:
a. Corrosion and slag formation resulting in shutdowns for
repairs.
b. Rabble teeth and arms are more expensive to replace than
alloy flights used in a rotary kiln.
c. Multiple hearth furnaces are more difficult to work on.
It takes more man-hours to rebuild a hearth than to re-
place brick in a kiln. Because of these factors, down-
time to affect repairs is longer in the multiple hearth
furnace.
d. More instrument components are required with a multiple
hearth furnace.
8. Effect of Feed Outages - The upper hearths in a multiple hearth
furnace can be damaged from temperature cycling caused by inter-
ruptions in furnace feed. Periodic planned shutdowns can be con-
ducted without hearth damage.
Feed outages are usually not a major problem in a rotary kiln. The
refractory is much less effected by temperature cycling in a kiln
than the hearth refractory in a multiple hearth furnace.
9. Operating Factors - Operating factors for kilns and multiple hearth
furnaces are as follows:
Rotary Kiln 85-95 Percent
Multiple Hearth Furnace 75-90 Percent
Multiple hearth furnaces must be shutdown more often to clean slag
and replace rabble teeth. When a furnace is down for repairs, the
work requires more man-hours to complete than similar work on a
rotary kiln.
Based on the above parameters, the multiple hearth furnace offers
the following advantages over a rotary kiln:
• Better control of temperature and atmosphere
• Lower fuel consumption
• Greater capacity turndown
429
-------
• Less area required
• Lower carbon losses from carryover and attrition
The rotary kiln advantages are:
• Less corrosion and slag formation
• Less downtime
• Lower maintenance costs and easier maintenance
• Less effect from feed outages
• Easier to operate
SUMMARY AND CONCLUSIONS
Granular activated carbon has been demonstrated to be effective in
treatment of a wide variety of industrial wastewaters. Both multiple hearth
furnace and rotary kilns can satisfactorily reactivate spent carbons used in
industrial wastewaters provided adequate consideration is given to selection
of materials, sizing of equipment, and operating philosophy. Experience
gained over the last ten years indicates that corrosion, slagging, poor re-
activation quality, carbon losses and line erosion can all be minimized
through good design. Although the same types of problems exist in in-
dustrial purification and municipal water and waste treatment applications
using granular activated carbon, they are magnified in industrial waste-
water applications where wastewater quality, and thus carbon exhaustion
rates, are more variable and substantially more corrosive. However, our
experience with reactivating over 100,000,000 pounds of spent carbon for
more than 75 different industrial wastewater applications, indicates that a
high quality product can be produced on a reliable, economical basis.
430
-------
BIOGRAPHIES
Roger H. Zanitsch is Engineering Director
of the Calgon Environmental Systems Division
for Calgon Corporation. He joined Calgon as a
Project Engineer in 1969 and was later named
Project Manager of the Environmental Engineer-
ing Department. Zanitsch received his BS degree
in Civil Engineering from the University of Cin-
cinnati and an MS degree in Environmental Engin-
eering from the same school. Zanitsch is a mem-
ber of the Water Pollution Control Federation.
Richard T. Lynch is a senior engineer in the
Process Engineering Group of Calgon Corporation's
Engineering Department'. He has a B.S. degree in
Chemical Engineering from the University of Florida.
He has been a project manager for the design of
several carbon adsorption reaction systems treating
industrial waste streams. He is a member of the
American Institute of Chemical Engineers and a
registered professional engineer in Florida.
431
-------
EXHIBIT 1
TYPICAL REACTIVATION SYSTEM
FLOW DIAGRAM
Exhaust
Gas
Spent Carbon
From Adsorbers
Spent
Carbon
Storage
5-10
Days.
Blowcase
React
Carbon
Storage
5-10
Days
Reactivated Carbon
To Adsorbers
432
-------
EXHIBIT 2
INSTALLED COSTS OF REACTIVATION SYSTEMS
CAPITAL COSTS
Capital Cost Estimate
($ Million)
Purchased Equipment
Installation*
Reactivation Rate - 1,000 Lbs/Day
5,000
10.000
30.000
60.000
0.24
0.61
0.85
0.36
0.91
1.27
0.77
1.93
2.70
1.20
3.00
4.20
*Installation costs include foundations, structural equipment setting,
electrical, instrumentation, site preparation, engineering contractor,
overhead and profit, and indirects.
Fuel - 8,000 BTU/Lb @ $3/106
BTU
Power @ 30/KWH
Steam - 1.0 Lb/Lb @ $4/1,000
Lbs
Labor @ Supervision
Makeup Carbon - 7% @ 57c/Lb
Maintenance @ 8% Capital
General Plant Overhead
Total Operating Cost
($l,000/Yr.)
OPERATING COSTS
Reactivation Rate - 1,000 Lbs/Day
5,000 10.000 30.000 60.000
* -
45
10
10
110
70
70
in
90
20
15
110
145
100
50
265
50
45
110
435
215
110
525
80
90
110
875
335
200
345
Operating Cost C/Lb Carbon 18.8
530
14.5
1,230
12.2
2,215
11.0
433
-------
EXHIBIT - 3
10,000
9
8
7
6
o
o
o
H-
V)
0 1,000
o 9
u
CAPITAL COST ESTIMATE
FOR REACTIVATION SYSTEMS
I I I I I
JS
(A
Z
O
I-
8
7
6f-
5 —
A -mi
3 -
2 —
too
I I I I I I
I
I I L I I I
I I I I
Z 34-56789 2
1,000 10,000
CARBON REACTIVATION RATE -
6789
100,000
S 6 7 B 9
LBS/DAY
-------
EXHIBIT - 4
DIRECT OPERATING COST
FOR REACTIVATION SYSTEMS
I.OOO
5 6789
10,000
CARBON REACTIVATION RATE - LBS/OAY
56789
100,000
5 6 7 8 9
-------
EXHIBIT 5
•Carbon- In
Hearth
i
61
Diameter
uu
Rabble Arm
Rabble Teeth
Two Burners and
Steam Inlets at
Hearth 4, 5 and 6,
not shown
Carbon Out
FIGURE 5 - CROSS SECTIONAL VIEW OF MULTIPLE HEARTH FURNACE
USED AT POMONA WATER RECLAMATION PLANT
436
-------
EXHIBIT 6
c
Dewatered
^
Carbon
Feed
Screw
V
J
Feed
Hood
^^^
—
_
J
_
Gas Discharge
S~~ Duct
_.. Feed End
|K" Seal |
./-Flights /-Refracto
1 _Ł - X
\i/iyifiM77*rH*i [til mi unit
7—*' Kiln Shell-^ '
^
ry
k
/
c
1
X,
/•
!•!*
2
•*
Discharge
End Seal
>^-Firing Hood
4-*- Steam
y
p t Fuel
nt
X^Nose
Ring
s^_ Discharge
Chute
437
-------
DISCUSSION
L. L. Krohn. Union Oil Co.; Have you made an analysis of the off-gases
from that furnace?
Roger Zanitsch; Yes, we have. We've looked at in excess of 100 different
industrial waste carbons. We've analyzed the off-gases before and after
afterburning on many of these, to determine what temperature and residence
time is needed for organic destruction.
L. L. Krohn. Union Oil Co.: Do you have any feel for the particulate matter
coming out?
Roger Zanitsch; We have systems operating that are designed with a 2 second
residence time which is primarily there to consume the particulates. If
you have a half second residence time which is certainly sufficient to des-
troy most of the organics present, you'll still have some carbon fines that
will need to be scrubbed. We feel that the 1-2 second residence time at
1800eF can destroy the carbon fines as well as the organics.
L. L. Krohn, Union Oil Co.: Considering the new source review - can we hope
to build this system?
Roger Zanitsch; Reactivation furnaces are now in operation and many are
being designed for industrial waste applications. Technology exists to
handle essentially all air pollution control requirements at a reasonable
cost.
«&
Mac McGinnis. Shirco_t Inc.: We have made some of the economics that you're
talking about for our electric regeneration furnace and compared them with
similar economics as you have presented here from multiple hearth and other
approaches and just a couple of comments - a couple of factors that we have
included that you haven't mentioned are in the area of utilities, scrubber
water which on small capacity units may be a fairly significant contribution
of operating costs; and the other factor youfmentioned quality of the pro-
duct, laboratory labor, lab time to confirm that the product is indeed of
the desired quality, can be a fairly significant contribution.
Roger Zanitsch; I'm glad you brought this up. In the analysis that I show-
ed, the operating cost included a 10% general plant service allowance on
the total operating cost to cover overhead items such as accounting, qual-
ity control, etc. As far as scrubber water cost and disposal, it can be a
factor. In those installations where we have scrubbers, we've recycled
water through the pretreatment system to remove the carbon fines. Frankly,
we haven't found this to be a significant cost factor.
438
-------
DISCUSSION
Mac McGinnis. Shirco. .Inc.: Well, a half cent here and half cent there it
begins to add up. The other general comment is you've indicated that there
is considerable data on regeneration costs in multiple hearth furnaces in
particular and you've showed us some trend lines in terms of direct operat-
ing costs. Can you comment on any specific data, you know, accumulated over
a period of time that indicates an actual cost figure for some specific ap-
plication?
Roger Zanitsch; The numbers which I presented are based on our experience
in operating both small and large furnaces.
Mac McGinnis, Shirco, Inc.: One last comment - would you say then that the
actual data would fall within that plus or minus 20% about your nominal
curve?
Roger Zanitsch; On industrial waste applications, yes. In process applic-
ations, such as the decolorization of sugar solutions, operating costs are
substantially lower since they have a constant feed and a very predictable
product.
Colin Grieves, Amoco Oil Co.; First, would you care to comment on some of
the new technology which you eluded to? And second, would you like to say
anything about regeneration of powdered activated carbon?
Roger Zanitsch; As far as the new technologies are concerned, I was person-
ally thinking of the electric furnace and the fluidized bed furnaces. The
Japanese have several different types of furnaces. Most of the experience
with the newer furnaces has been in either pilot-scale or on the commercial
scale, but in considerably less corrosive application than you have in in-
dustrial wastes. In industrial waste applications, the big awakening has
been in the areas of corrosion, maintenance costs, and feed interruptions.
The new technologies have not been demonstrated in this type of service.
As the new technology develops, it's going to take some time to gain the
experience necessary to apply these new furnaces in the industrial waste
effort. As far as powdered carbon activation, I don't really feel qualified
to discuss it on the basis that I would only be expressing my opinions
since no commercial experience has been developed.
439
-------
"ACTIVATED SLUDGE ENHANCEMENT: A VIABLE
ALTERNATIVE TO TERTIARY CARBON ADSORPTION"
Leonard W. Crame
Senior Chemical Engineer, Texaco Inc.
INTRODUCTION
In view of the possibility of more stringent 1983 BATEA (Best
Available Technology Economically Achievable) effluent guidelines,1>2»3.4
petroleum refiners are faced with the dilemma of an insufficient data base
to determine the proper approach for making cost-effective improvements.
The EPA previously proposed granular activated carbon adsorption after acti-
vated sludge treatment as BATEA technology; however, the current emphasis is
to consider both effluent quality and the cost effectiveness of attaining the
desired results. Two proposed approaches to BATEA technology are (1) in-
creasing the sludge age (or mean cell residence time) of the activated sludge
biomass to develop a more diverse population capable of assimilating biore-
fractory organics or (2) adding powdered carbon directly to activated sludge
aeration basins. Both alternatives to tertiary carbon adsorption would re-
quire little capital investment and would lower operating costs.
Grutsch and Mallatt5'6'7'8'9'10'11 have proposed that the best
refinery end-of-pipe treatment for soluble organic removal should include pH
control, equalization, optimized dissolved air flotation (DAF), and high
sludge age (20-50 days) activated sludge treatment. High sludge ages (SA)
require mixed liquor solids levels above conventional levels (5-10 days SA).
These higher levels increase solids flux and must be considered in secondary
clarifier solids loadings. Also high effluent TSS, despite less frequent
sludge wasting, can result in a loss of mixed liquor solids.
Grutsch and Mallatt emphasize that optimized chemically-assisted
DAF pretreatment (or comparable pretreatment) reduces the colloid charge
(zeta potential) to maximize particle agglomeration for efficient flotation,
and reduces the organic load on the activated sludge unit (ASU). Removing
colloids normally present in raw refinery wastewater allows better biofloc-
culation and lower effluent total suspended solids (TSS) since most refinery
colloids and biosolids have repelling negative charges. The microbial popu-
lation could then acclimate to the biorefractory organics by producing
enzymes which reduce these to simpler biodegradable substrates. Current
reports from within the petroleum industry seem to indicate some benefits
for increasing SA. Other investigators^ have reported that high SA (low
food/microorganism ratio) produces poor sludge settleability.
440
-------
As a result of pilot studies at the Du Pont Chambers Works, Button
and Robertaccio-U were issued a U.S. patent!* for the Du Pont PACT process 15
The PACT process basically involves the addition of powdered carbon (or
fuller's earth, etc.) to an ASU, usually in a range of 50-400 mg/1 based on
influent flow. Du Pont has reported 16, 17, 18 a number of advantages of the
PACT process which include:
(1) color removal,
(2) stability against shock loadings,
(3) improved BOD removal,
(4) improved refractory organic removal,
(5) resistance to toxic substances,
(6) improvement in hydraulic capacity,
(7) improved nitrification (mainly in municipal wastes) ,
(8) foam suppression, and
(9) improved sludge settling and increased clarifier capacity.
A disadvantage of the PACT process is that the system can become very expen-
sive if powdered carbon addition rates become high (hundreds of mg/1) , even
though powdered carbon is cheaper than granular carbon.
DeJohn and Adamsl9»20,21 have developed a considerable amount of
pilot study data on activated sludge-powdered carbon systems. They report
significant enhancement in studies involving refinery and petrochemical
wastewaters. DeJohn and Adams explain the powdered carbon enhancement
mechanism as localization and concentration of oxygen and pollutant as the
result of adsorption on carbon surfaces, resulting in a more complete bio-
oxidation. The adsorption of biorefractory organics allows a longer resi-
dence time for these components in the system. Other researchers", 23 have
found similar improvements using activated sludge-powdered carbon systems
and propose analogous enhancement mechanisms.
has reported a case history of a full-scale activated
sludge-powdered carbon demonstration run at the Corpus Christi, Texas, Sun
Oil refinery. Results included better system stability, reduction of foaming,
resistance to upset conditions, lower effluent suspended solids and clearer
effluent, and improved organic removal. These improvements were achieved
by maintaining only a 450-mg/l powdered carbon reactor concentration with a
10rttg/l powdered carbon dosing requirement. The shortcoming of this inves-
tigation was that a parallel control could not be run simultaneously and most
improvements reported could possibly have been attributed to better
clarification.
The merits of powdered carbon enhancement have been further
confused with the more recent development of several types of powdered
carbons with significantly different properties.
441
-------
SCOPE OF WORK
1. Objective
The objective of this study was to determine if the relatively
simple process changes of increased sludge age or the addition of powdered
activated carbon in conventional refinery activated sludge systems can sig-
nificantly enhance the removal of organic wastewater contaminants to achieve
or approach the level of proposed BATEA (1983) technology more cost effec-
tively than the addition of granular activated carbon contactors to BPCTCA
(1977) technology.
2. Procedures
In Part I of this study, five completely-mixed (15 gal) ASU's were
operated in parallel with identical 18-hr retention times and 300-gpd/sq ft
clarifier rise rates. A sixth ASU was run as a second-stage unit with the
same 18-hr retention time (Figure 1). All biological reactors were located
in a temperature controlled room in an attempt to dampen influent wastewater
temperature variations and control biological reactions at about 85 F.
ASU's A and F served as controls, simulating conventional refinery
units with a 0.3 Ib TOC/lb MLVSS-day organic loading. Separate controls were
run to determine the effect of optimized pretreatment on activated sludge
treatment and tertiary carbon adsorption. Equalized (24-hr)«*and pH-control-
led refinery wastewater was pretreated by dual-media (sand-anthracite) fil-
tration (4.6 gpm/sq ft) and a chemically assisted DAF unit (1.5 gpm/sq ft)
prior to control ASU's A and F, respectively. The optimized DAF pretreatment
neutralized the negatively charged colloids, thus facilitating their removal
and producing a bio-unit feed that contained essentially only soluble
organics. Sodium phosphate (monobasic) was added to the filter and DAF unit
effluents for a minimum TOG:phosphorus ratio of 100:2 to assure a proper
nutrient level. Effluents from ASU's A and F were continuously filtered
through a dual-media (sand-anthracite) tertiary filter for TSS removal before
passing through a series of four granular activated carbon contactors to
simulate proposed 1983 BATEA technology.
ASU's B, C, and D treated optimized DAF effluent with sludge wast-
ing calculated for a 50-day biological SA. A commercially available, conven-
tional-surf ace-area, powdered carbon25 (designated PC-C) was added daily to
ASU C to maintain a 500-mg/l reactor operating level. Similarly, PC-H, a
high-surface-area powdered carbon,26 was added to ASU D to investigate its
enhancement capabilities.
ASU E was also operated at a 50-day SA treating ASU B effluent to
determine if there was any benefit to ASU staging.
In Part II of this study, the second-stage ASU E was placed in
parallel with other ASU's treating the DAF unit effluent as shown in
Figure 1 and redesignated as ASU G. PC-H was added to ASU G to maintain
a 2500-mg/l operating level while powdered carbon levels were increased in
ASU C and D to 1000 mg/1.
442
-------
Effluents were collected daily, Monday through Friday, as grab
samples. Grab samples were taken in lieu of composites for convenience since
the pilot unit treatment scheme contained significant equalization. Samples
from the equalization basin, biological reactors, and carbon columns were
filtered with Gelman type A/E glass fiber filters to give the soluble con-
taminant (TOG, COD, S=) level. Glass fiber filters were used instead of
0.45-micron filters because the solids retained on glass fiber filters
define the TSS measurement. Samples were analyzed immediately after collec-
tion or were preserved until analyzed using accepted preservation methods.27
All effluent data were compared after plotting values on log-normal
probability papers. Single straight line data fits were determined by calcu-
lating 50th and 90th percentile values. The 50th percentile values equaled
the antilog of the mean of the log values of data sets. The 90th percentile
values were calculated assuming a single-tailed log-normal data distribution
-f~.
In N90 = In NSO + 1.282 In 84
where
S(j is the standard deviation.
Engineering judgment was used to determine which data sets being compared
appeared different and required additional statistical analyses to confirm
significance of median differences. Median data values were compared to
determine if they were from the same population using a paired t-test2°
assuming a log-normal distribution as follows:
compute t = , -=
/Łd2 - (Id)2/n
V n(n-l)
n
3"= mean of differences =]Tdi/n where d^ = In x^-ln y±
1
for i - l,2,3,.,.n of n data pairs
S
-------
H20) and 20-40 mg/1 Dearborn 431 cationic polymer. Chemical dosages were
initially optimized using zeta potential titrations in conjunction with jar
tests; however, only jar tests were continued since zeta potential calcula-
tions became tedious and inconsistent due to the high wastewater specific
conductivity (usually 4000-8000 micromhos/cm). This salinity was primarily
due to the brackish intake waters to the refinery and may have impeded
chemical coagulation at lower chemical dosages as reported by Grutsch and
Mallatt.8
The superiority of the optimized DAF unit (operating at 1.5 gpm/sq
ft) as a pretreatment system over simple sand filtration is clearly evident
in the three weeks of run data represented in Figure 2.
DAF unit effluent 50th percentile TOG (158 mg/1) was 55 percent
less than the 50th percentile TOG (352 mg/1) in the equalization basin
influent to the DAF unit, while the sand filter gave only a 18-percent reduc-
tion. Since, at best, only a 31-percent TOG reduction could be achieved by
vacuum filtration of equalization basin samples with glass fiber filters
(which define TSS), the true effectiveness (55 percent reduction) of colloid
and oil coagulation and removal in the chemically assisted DAF unit can be
seen. The DAF unit effluent contained essentially only soluble organic
contaminants.
The continuous dual-media filter (operating at 4.6 gpm/sq ft) could
only manage a TOG reduction of about one-third of the DAF unit. There was
no indication that a shorter run time would improve the filter effluent
significantly. It appeared that due to the nature of the solids, chemical
addition to the filter feed would have been required for an improved system.
The purpose of the filter, however, was to produce a biological reactor feed
with characteristics comparable to DAF treatment without chemicals. It was
observed, on occasions, that DAF pretreatment was very poor when chemical
feed pumps failed.
Figure 2 illustrates COD removal by both pretreatment systems
employed and again demonstrates the effectiveness of optimized DAF treatment.
As in subsequent graphs, the data points are not shown to avoid congestion of
data.
Fiftieth percentile oil and grease values during Part I of this
study were 101, 70, and 16 mg/1 for the equalization basin, filter effluent
and DAF unit effluent, respectively. The equalization basin 50th percentile
TSS level of 78.0 mg/1 was reduced to 57.5 mg/1 by the filter and to 19.0
mg/1 by the DAF unit. A portion of the TSS in the DAF unit effluent was due
to biological growth rather than influent solids. It is possible that part
of the organics reduction through the DAF unit was the result of biological
activity which could also occur in full-scale systems.
2. Part I - Activated Sludge Performance
TOG and COD (filtered) effluent variability plots for pilot-scale
18-hr retention ASU's are compared in Figure 3. These results are from the
initial 3-week data run.
444
-------
Control ASU F produced a better effluent than control ASU A both
operating at equal 0.3 Ib TOC/lb MLVSS-day (F/M) loadings. MLVSS levels in
ASU A averaged 1,148 mg/1, about twice that of ASU F due to a twofold
increase in feed TOC. Considering both TOG and COD removal, ASU B, C, E,
and F did not show any significant overall difference in performance.'
Fiftieth percentile TOC values ranged from 53-58 mg/1 while COD values were
97-116 mg/1 as shown in Figure 3. Sludge age was not a controlling perfor-
mance variable as ASU B (50-day SA) and ASU F (about 10-day SA) differed
greatly in solids retention time with average MLSS levels of 1,621 mg/1 and
816 mg/1, respectively. Chemically assisted pretreatment for removal of
colloids and oil had the most significant effect on organics removal. The
high-surface-area powdered carbon (designated PC-H) significantly enhanced
organic removal in ASU D, with a 50-day SA and a 500-mg/l PC-H operating
level. Enhancement was not evident in ASU C containing the conventional-
surface-area powdered activated carbon (designated PC-C). Powdered carbon
addition increased the average ASU C MLSS level to 1,885 mg/1 with ASU D
averaging 1,976 mg/1.
Since a marginal enhancement occurred with the addition of PC-H at
a 500-mg/l level, the scope of this investigation was expanded to evaluate
powdered activated carbon addition at a 1000-mg/l level and only PC-H at
approximately 2500 mg/1. This would give a greater overview of the enhance-
ment capabilities of powdered carbon, especially the highly active PC-H. ASU
E was taken out of service since it was only succeeding in lysing biological
cells as a second stage following ASU B. The reactor was placed in parallel
with other units being fed by the DAF unit and redesignated ASU G. The SA
was maintained at 50 days and PC-H was built up to a reactor level of about
2500 mg/1 for Part II of this study.
3. Part I - Granular Carbon Adsorption
Granular carbon Series A, treating ASU A effluent, exhausted two
130-gram carbon beds during 17 days of a 3.4-gpm/sq ft hydraulic loading in
Part I. A 20-mg/l soluble (filtered) TOC and a 44-mg/l soluble COD effluent
(50th percentile, Figure 4) was produced with 0.10 and 0.09-g TOC/g carbon
accumulative loadings at exhaustion. Carbon Series F, treating ASU F efflu-
ent, reduced the 50th percentile effluent soluble TOC and COD to 23 mg/1 and
40 mg/1, respectively. Because of the relatively few data points used to
establish Figure 4, there is little significance in the difference between
carbon series A and F 50th percentile values. A single carbon bed was ex-
hausted to a 0.12-g TOC/g carbon loading. TOC loadings of carbon columns in
Series A were comparable to an average of 0.11 g TOC/g carbon reported for
the granular carbon during the four previous exhaustions prior to each
regeneration.
Granular carbon effluents were of substantially better quality than
all biological unit effluents. The 50th percentile soluble TOC and COD re-
ductions in carbon Series A were 44 mg/1 arid 84 mg/1, respectively, whereas
carbon Series F accounted for a 50th percentile 35-mg/l soluble TOC and 65-
""8/1 soluble COD reduction.
445
-------
4. Part II - Pretreatment
Pretreatment by filtration and chemically assisted DAF treatment
continued as in Part I of this study. Again, using dosages of 40-mg/l filter
alum and 20-40 mg/1 Dearborn 431, optimized DAF pretreatment reduced the
equalization basin TOG and COD by more than 50 percent as shown in Figure 5.
Filtration could only remove gross quantities of oil and solids without pre-
liminary chemical coagulation. Although equalization basin, filter, and DAF
unit effluent 50th percentile TOC and COD concentrations were approximately
equal to those experienced in Part I of this study, there existed a greater
degree of variability in Part II. A contributing variability factor was the
rainfall dilution of refinery wastewater streams as an average of 0.21 in./
day of rain fell during Part II compared with 0.06 in./day during Part I.
5. Part II - Activated Sludge Performance
The effluent quality for Part II, basis filtered TOC and COD, is
given in effluent frequency distributions, Figure 6, for the six-week run
period. Control ASU A (F/M = 0.3), without optimized pretreatment, continued
producing the most inferior effluent and experienced three upsets due to the
development of a filamentous bulking sludge. The unit was restarted on each
upset occasion with new seed and allowed'to acclimate for a few days before
effluent data were used for comparison with parallel systems. ASU B, C, and
F, as in Part I, produced nearly equivalent effluents in terms of filtered
TOC and COD with neither high SA (50 days) nor 1000 mg/1 PC-C enhancing bio-
logical treatment. PC-H added to ASU D and G at levels of 1000 and 2500
mg/1, respectively, reduced TOC and COD substantially. Compared with high SA
control ASU B, 50th percentile TOC was reduced an additional 10 mg/1 and 22
mg/1 in reactors D and G, respectively. COD 50th percentile reductions below
reactor B were 22 mg/1 for ASU D (1000 mg/1 PC-H) and 39 mg/1 for ASU G
(2500 mg/1 PC-H). The ASU G run time was abbreviated, however, due to the
time required for acclimation at the higher PC-H level. As in Part I of this
study, it was observed that as powdered carbon levels were suddenly increased
in ASU C, D, and G, performance was exceptionally good for a short period of
time.
Phenols feed concentrations were higher in Part II of this investi-
gation as 90th percentile values reached 18 mg/1, compared with 8.5 mg/1 in
Part I. ASU D (see Table 1) and G, containing PC-H, provided the best
phenols removal with 50th percentile phenols levels of 0.05 mg/1 and 0.04
mg/1, respectively. This was slightly lower than the high SA control ASU B
(0.06 mg/1) and low SA control ASU F (0.07 mg/1). Although the lack of opti-
mized pretreatment produced higher 50th percentile phenols levels (0.11 mg/1)
in ASU A, even poorer reductions were experienced with ASU C as in Part I.
Similar results were obtained in Part I. An occasional high phenols value
was measured in the effluents of ASU D and G but not with the consistency
or magnitude of ASU C.
Effluent oil and grease values, included in Table 1,'illustrate the
significance of removing most of the oil and grease before biological treat-
ment. The 18 mg/1 50th percentile oil and grease effluent level of ASU A
greatly exceeded the concentration of 5 mg/1, or less, discharged from
446
-------
reactors receiving optimized pretreatment. The addition of PC-H to ASU D and
G gave slight oil and grease improvement with SA alone not being an
enhancement factor.
Effluent TSS levels were high in ASU A at a 50th percentile level
of 86 mg/1. TSS increased to more than 150 mg/1 when filamentous sludge
bulking occurred. However, ASU's with DAF pretreatment produced a more
settleable sludge. The high SA control ASU B had a significantly higher
effluent TSS level than the lower SA control ASU F, but the high SA could
be maintained. ASU C and D experienced effluent TSS levels less than control
ASU B despite higher reactor solids due to powdered carbon. No effluent TSS
increase was observed in ASU G due to the higher (2500 mg/1) PC-H level.
Ammonia nitrogen removal in system A was less than the 80 percent
achieved in the systems with optimized pretreatment. The organic loading was
higher and the sludge age was less than in other systems. The factors con-
trolling the degree of nitrogen removal were not investigated. Nitrification
during Part II was not as complete as that obtained in Part I.
Sludge Characteristics and production rates are summarized in
Table 2 for all ASU's. As expected, ASU A had the highest measured oxygen
uptake averaging 0.16 mg oxygen/1-min due to a higher influent organic con-
centration. Oxygen consumption averaged 0.10-0.12 mg oxygen/1-min in other
ASU mixed liquors, but a relationship of increased oxygen demand and enhanced
biological treatment did not exist. The sludge volume index (SVI), a measure
of sludge compactability, significantly improved with SA and powdered carbon
addition. Sludge settling velocities were exceptionally high with the worst
rate (ASU A) being 0.17 ft/min corresponding to a 1830-gpd/sq ft clarifier
rise rate. Other mixed liquors settled with zone settling velocities of
0.34-0.39 ft/min. The average MLSS concentration of 745 mg/1 in ASU F was
too low for zone settling to occur. One of the most surprising results of
powdered carbon addition was that less biomass was produced than in control
systems. ASU G produced an average of 0.08 Ib biomass/lb COD removed com-
pared with control rates of 0.22 for ASU A and 0.19 for ASU F. PC-H was more
effective than PC-C at reducing biomass production rates at the same SA. The
total sludge production of activated sludge powdered carbon systems was not
much higher than controls, due to lower biological sludge production rates.
Powdered carbon inventories and makeup requirements for ASU's are
summarized in Table 3 for Parts I and II of this study. PC-C losses were
slightly higher than PC-H but still reasonably close to 2 percent per day.
Since biological sludge was wasted at a rate of 2 percent per day in high SA
reactors, it is a fairly good assumption that powdered carbon lost In efflu-
ents was in the same proportion to biological sludge as in the mixed liquor.
Thus both biological and powdered carbon SA may be assumed to be equal for
simplification of powdered carbon daily makeup requirements. The powdered
carbons must be wetted to prevent loss of floating carbon in the clarifier.
This was accomplished by boiling the carbon slurry in this study. Vacuum
degassing could also be used.
Another observation made during Part II was that activated sludge-^
powdered carbon systems significantly reduced aeration basin foamir.0 .c
447
-------
with control systems. Foaming in ASU aeration basins was not a problem but
did occur occasionally.
6. Part II - Granular Carbon Adsorption
A single granular carbon bed was exhausted from carbon Series A
with an accumulative organic loading of 0.15 g TOC/g carbon during Part II.
The data include three short runs. The first two carbon beds required back-
washing almost daily due to high TSS levels which could not be continuously
removed by dual-media filtration. Fiftieth percentile effluent soluble TOC
was 30 mg/1 (see Figure 7) for a reduction of 38 mg/1 from ASU A. ASU A 50th
percentile soluble COD was reduced by 84 mg/1 to 79 mg/1. Although phenols
levels were generally low (50th percentile of 0.04 mg/1) a few very high
effluent phenols levels were detected in carbon Series A giving a 90th
percentile phenols value of 4.8 mg/1 (Table 1). Phenols must have been
adsorbed, concentrated, and then eluted in slugs from the carbon beds to
achieve such a high level. Effluent oil and grease levels remained low with
50th percentile values less than 3 mg/1.
Carbon Series F exhausted a single carbon bed to an accumulative
organic loading of 0.13 g TOC/g carbon while surpassing the performance of
carbon Series A. The 50th percentile soluble TOC was significantly lower at
18 mg/1 for a 28 mg/1 reduction (Figure 7) from ASU F. Fiftieth percentile
soluble effluent COD was 64 mg/1 for a 44 mg/1 reduction. Phenols levels
were extremely low at 0.02 mg/1 (50th percentile) and no sudden loss of
adsorbed phenols was detected during most of the Part II data run (Table 1).
Oil and grease effluent levels (50th percentile) were again less than 3 mg/1.
The lower dashed lines in Figure 7 represent the performance of ASU
G, the best of the activated sludge-powdered carbon reactors. ASU G produced
an effluent superior to carbon Series A and approached the quality of carbon
Series F. The powdered carbon enhancement removed about 85 percent of the
soluble TOC adsorbed on carbon Series F and about 60 percent of the COD based
on 50th percentile effluent values. The 2,500 mg/1 PC-H operating level in
ASU G significantly reduced effluent color to a level comparable with
granular carbon effluent color.
7. Economics
Although unequal in overall performance, a high SA activated
sludge-powdered carbon system (ASU G, 72 mg/1 COD) approached the level of
granular carbon adsorption (carbon Series F, 64 mg/1 COD) to within 8 mg/1
COD at the 50th percentile point. Both systems would require extensive pre-
treatment and tertiary suspended solids removal. All other process compon-
ents being essentially equal, daily carbon usage costs were estimated for
theoretical plant flows of 1-5 MM gpd.
The cost of virgin powdered carbon (PC-H or PC-C) was estimated at
$0.30/lb and it was assumed that wasted carbon would be thrown away. To
calculate the equivalent powdered carbon dosage required for an 18-hr reten-
tion ASU it was assumed a 50-day SA would be maintained, giving an average
2 percent powdered carbon makeup. This was the equivalent of a 37.1 mg/1
448
-------
powdered carbon addition rate based on influent flow. A powdered carbon
feeder and storage facilities were included in powdered carbon costs using
Du Pont economicsJ-o and applying the 0.6 rule. It was assumed that the
powdered carbon feeder could handle a 50 mg/1 addition rate.
Regenerated granular carbon adsorption costs were estimated, using
Brown and Root, Inc. economics,29 and converted to 1977 dollars. Daily gran-
ular carbon costs were estimated using 17.2 percent of the fixed investment
for operating and maintenance cost and 17.7 percent for depreciation. The
total daily costs for powdered carbon were estimated using the same per-
centage allowances.
Daily estimated carbon costs are shown in Figure 8 for theoretical
flows of 1-5 MM gpd by scaling up ASU G and carbon Series F carbon require-
ments. The cost effectiveness of the relatively simple process change of
adding powdered activated carbon to the activated sludge process can be
clearly seen. Estimated daily cost savings would range from $987/day at 1 MM
gpd flow to $2750/day at 5 MM gpd using high-surface-area powdered carbon
(PC-H) addition rather than granular carbon adsorption. The incremental cost
would be about $14.73 per pound of COD at 1 MM pgd (see Figure 9).
DISCUSSION
1. Increasing Sludge Age (SA)
Contrary to conventional activated sludge design techniques, the
increased SA did not result in sludge deflocculation, higher SVI, and high
effluent TSS. With the exception of a few days, the high SA control ASU B
easily achieved a high SA as a result of good pretreatment as proposed by
Grutsch and Mallatt.10,11 However, no enhanced performance was measured at
the high SA nor in the two-stage system with both reactors at a high SA in
Part I. Possibly more emphasis should be placed on the benefits of optimized
pretreatment than on increased SA as parallel activated sludge systems at
about the same SA (F/M = 0.3) were vastly different in performance at
bontrasting degrees of pretreatment.
Increasing SA gave a reduction in biological solids production as
the conventional F/M control ASU F produced 0.19 Ib VSS/lb COD removed com-
pared with 0.16 for high SA control ASU B. This sludge production was not
quite as significant as it would have been if ASU F had operated at 5-10 days
SA where many conventional ASU's operate instead of about 14 days. Any re-
duction of biological solids production would help lower sludge treatment and
disposal costs.
The high effluent 50th percentile ammonia level of ASU A (11.5
mg/1) in Part II was probably the result of upset conditions which resulted
from filamentous sludge bulking causing the loss of biomass. The calculated
10-day SA of ASU A was only slightly less than ASU F (14 days) which Produced
a 50th percentile ammonia level of 2.3 mg/1 in Part II. ^hougV ** to
SA generally does improve nitrification, no conclusions could be drawn as to
its effect in this study. The conventional ASU F had already produced an
effluent that was about 90 percent nitrified.
449
-------
2. Activated Carbon Enhancement Mechanism
A 0.59-g TOC/g carbon loading was calculated in ASU G while operat-
ing with a 2500 mg/1 PC-H level. This extremely high "apparent" TOG loading,
as explained by Flynn,16 may be the result of continuous adsorption of slowly
biodegraded organics which are "biologically regenerated" from the carbon
many times over the biological and carbon SA. "Apparent" TOG loadings there-
fore increase with higher SA, optimizing the use of powdered carbon, until
the carbon becomes loaded with completely biorefractory organics. This
explanation of an "apparent" loading or enhancement mechanism appears logi-
cal; however, oxygen uptake and biological sludge production data presented
here negate biological regeneration in this study. ASU G not only had com-
parable oxygen uptake measurements with control ASU B, but it produced about
50 percent less biological solids. This implies that the actual enhancement
may have been predominantly due to adsorption on the high PC-H surface area.
Considering that PC-H had approximately five times the surface area of con-
ventional powdered carbons, such as PC-C, the expression of the TOG loading
as 0.12 g TOG/500 sq m of surface area would be more reasonable.
DeJohn-'-^ > 30 explains that granular carbon columns are sometimes
undersized because the designer uses virgin carbon and assumes that regen-
erated carbon will have the same activity. The thermal regeneration process
will enlarge some carbon pores reducing the surface area and decreasing the
adsorption of small molecules which are not so strongly adsorbed on larger
pores. Assuming that many small molecules require small powdered carbon
pores for moderately strong adsorption, PC-H may have been more effective
than PC-C because of pore size distribution, provided that the normally
biorefractory refinery organics were small molecules.
The mechanism of powdered carbon enhancement of the activaged
sludge process was not defined in this study and needs further investigation
in Phase II. Target SA's of the activated sludge-powdered carbon systems
were 50 days. Ideally, systems should be operated for periods of several
SA's to insure that equilibrium conditions have been reached and that the low
(2 percent) daily powdered carbon makeup rate will continue to give consis-
tent results.
The selection of the best powdered carbon for a particular acti-
vated sludge enhancement is not a simple task since powdered carbons vary in
their adsorptivity. Carbon isotherms performed on a refinery wastewater
would exhibit a wide variability and require a statistical analysis to select
the best powdered carbon. Isotherms would have to be performed on the acti-
vated sludge effluent (as in Phase II of this study) to determine enhancement
strictly due to adsorption.
The powdered activated carbon (PC-H) utilized with very good
enhancement results is not, as of yet, commercially available. Because of
the relatively high cost of granular carbon adsorption, other powdered
carbons at similar and higher operating levels would probably offer a signi-
ficant improvement in activated sludge performance and remain more cost
effective than granular carbon.
450
-------
3. Granular Carbon Adsorption
Granular carbon adsorption data indicate that the quality of end-
of-pipe refinery wastewater treatment depends on optimization of each treat-
ment step from primary to tertiary treatment. The use of equilibrium or re-
generated granular carbon in pilot studies will provide a more realistic data
base, recognizing that economics would favor regeneration for many potential
users.
The classical approach31'32'33'34 for handling carbon adsorption
data to establish breakthrough curves was virtually useless in this study
because it assumes the carbon column influent has a single adsorbate. In the
calculation of accumulated TOG loadings at apparent breakthrough of carbon
columns there were several instances where organics were eluted in slugs from
carbons in both series. At times, phenols were two orders of magnitude high-
er than normal. This phenomenon is a very real problem and must be consid-
ered when establishing stringent effluent discharge guidelines for industry.
Even the best available technology, disregarding economics, has its
limitations.
SUMMARY
The EPA 1983 guidelines for the petroleum refining industry have
assumed that 1977-type technology must be upgraded by the addition of costly
systems, such as granular activated carbon adsorption. The results of this
API study indicate that, should the EPA adhere to the granular carbon techno-
logy originally proposed, it may be possible to achieve this level of treat-
ment technology by the much more cost-effective method of adding powdered
activated carbon to the 1977 activated sludge system.
Process modifications including optimized pretreatment and the
addition of a high-surface-area powdered activated carbon can be used to pro-
duce an effluent which is comparable in quality to that obtained by granular
carbon adsorption. Increasing activated sludge age from the conventional
mode of operation (about 10 days) to about 50 days did not give a significant
system improvement; however, in conjunction with powdered carbon addition,
high sludge age allowed higher equilibrium reactor concentrations (2500 mg/1)
at low (2 percent) carbon makeup rates. This benefit has been demonstrated
with the high-surface-area carbon and it is possible that it can also be
obtained with increased levels of conventional powdered carbon. The cost-
effectiveness of any powdered carbon will depend on the wastewater charac-
teristics and powdered carbon adsorptivity, which was greater for the high-
surface-area carbon (2462 sq m/g) than for the conventional-surface-area
carbon (550 sq m/g) investigated here. Even granular carbon adsorption was
found to have limitations as slugs of phenols were eluted, on occasion,
into the effluent.
ACKNOWLED6B86STS"
This study was funded in part by the American Petroleum Institute,
Divison of Refining, CREC Liquid Waste Subcommittee.
451
-------
REFERENCES
1. Environmental Protection Agency, "Petroleum Refining Point Source
Category Effluent Guidelines and Standards," Federal Register, Vol 38
(240) 34542 (December 14, 1973).
2. Environmental Protection Agency, "Petroleum Refinery Point Source
Category Effluent Guidelines and Standards," Federal Register, Vol 39
(91) 16560 (May 9, 1975).
3. Environmental Protection Agency, "Petroleum Refining Point Source
Category Effluent Guidelines and Standards," Federal Register, Vol 40
(98) 21951 (May 20, 1975).
4. Jones Associates, "Effluent Limitations in the Petroleum Refining
Industry," Vol IB, Prepared for the Office of General Counsel, American
Petroleum Institute (January, 1976)
5. J. F. Grutsch and R. C. Mallatt, "Optimize the Effluent System - Part 1:
Activated Sludge Process," Hydrocarbon Processing, Vol 55 (3) 105 (1976).
6. J. F. Grutsch and R. C. Mallatt, "Optimize the Effluent System - Part 2:
Intermediate Treatment," Hydrocarbon Processing, Vol 55 (4) 213 (1976).
7. J. F. Grutsch and R. C. Mallatt, "Optimize the Effluent System - Part 3:
Electrochemistry of Destabilization," Hydrocarbon Processing, Vol 55 (5)
221 (1976).
8. J. F. Grutsch and R. C. Mallatt, "Optimize the Effluent System - Part 4:
Approach to Chemical Treatment," Hydrocarbon Processing, Vol 55 (6) 115
(1976).
9. J. F. Grutsch and R. C. Mallatt, "Optimize the Effluent System - Part 5:
Multimedia Filters," Hydrocarbon Processing. Vol 55 (7) 113 (1976).
10. J. F. Grutsch and R. C. Mallatt, "Optimize the Effluent System - Part 6:
Biochemistry of Activated Sludge Process," Hydrocarbon Processing,
Vol 55 (8) 137 (1976).
11. J. F. Grutsch and R. C. Mallatt, "A New Perspective on the Role of the
Activated Sludge Process and Ancillary Facilities," Presented at Joint
EPA-API-NPRA-TU Open Forum on Management of Petroleum Refinery Waste-
waters, Tulsa, Oklahoma (January 26-29, 1976).
452
-------
12. D. L. Ford and W. W. Eckenfelder, Jr., "Effect of Process Variables on
Sludge Floe Formation and Settling Characteristics," Journal Water
Pollution Control Federation, Vol 39 (11) 1850 (1969)T~~
13. G. Grulich, D. G. Button, F. L. Robertaccio, and H. L. Glotzer, "Treat-
ment of Organic Chemicals Plant Wastewater with the Du Pont PACT
Process," Presented at AIChE National Meeting (February, 1972).
14. D. G. Button and F. L. Robertaccio, U. S. Patent 3,904,518 (September 9,
15. E. I. Du Pont DeNemours and Company, "Du Pont PACT Process," Technical
Bulletin.
16. B. P. Flynn and L. T. Barry, "Finding a Home for the Carbon: Aerator
(Powdered) or Column (Granular)," Proceedings of the 31st Annual Purdue
Waste Conference (May 5, 1976).
17. B. P. Flynn "A Methodology for Comparing Powdered Activated Carbons for
Activated Sludge," Presented at 168th National Meeting, ACS, Div. of
Petroleum Chemistry, Symposium on Disposal of Wastes from Petroleum and
Petrochemical Refineries (September 13, 1974).
18. B. P. Flynn, F. L. Robertaccio, and L. T. Barry, "Truth or Consequences:
Biological Fouling and Other Considerations in the Powdered Activated
Carbon - Activated Sludge System," Presented at 31st Annual Purdue Waste
Conference (May 5, 1976).
19. P. B. DeJohn and A. D. Adams, "Treatment of Oil Refining Wastewaters
with Granular and Powdered Activated Carbon," Proceedings of 30th Annual
Purdue Industrial Waste Conference (May 6, 1975).
20. A. D. Adams, "Powdered Carbon: Is It Really That Good?," Water and
Wastes Engineering, Vol 11 (3) B-8 (1974).
21. P. B. DeJohn and A. D. Adams, "Activated Carbon Improves Wastewater
Treatment," Hydrocarbon Processing, Vol 54 (10) 104 (1975).
22. A. B. Scaramelli and F. A. DiGiano, "Upgrading the Activated Sludge
System by Addition of Powdered Carbon," Water and Sewage Works, Vol 120
(9) 90 (1973).
23. A. E. Perrotti and C. A. Rodman, "Enhancement of Biological Waste Treat-
ment by Activated Carbon," Chemical Engineering Progress, Vol 69 (11)
63 (1973).
24. J. A. Rizzo, "Case History: Use of Powdered Activated Carbon in an
Activated Sludge System," Presented at Joint EPA-API-NPRA-TU Open Forum
on Management of Petroleum Refinery Wastewaters (January 26-29, 1976).
25. ICI United States Inc., "Powdered Hydrodarco Activated Carbons Improve
Activated Sludge Treatment," Product Bulletin PC-4 (October, 1974).
453
-------
26. Amoco Research Corporation, Amoco Active Carbon Grade PX-21 Product
Information Sheet (May, 1976).
27. APHA, Standard Methods for the Examination of Water and Wastewater.
13th Ed., New York, New York (1971).
28. H. E. Klugh, Statistics; The Essentials for Research, 2nd Ed., John
Wiley & Sons, Inc., New York, New York (1974).
29. Brown and Root, Inc., "Economics of Refinery Wastewater Treatment,"
American Petroleum Institute Publication No. 4199 (1973).
30. P. B. DeJohn, "Carbon from Lignite or Coal: Which is Better?,"
Chemical Engineering, Vol 82 (9) 113 (1975).
31. Metcalf and Eddy, Inc., Wastewater Engineering, McGraw-Hill, New York
New York (1972).
32. W. W. Eckenfelder, Jr., Industrial Water Pollution Control, McGraw-
Hill, New York, New York (1966).
33. H. J. Fornwalt and R. A. Hutchins, "Purifying Liquids with Activated
Carbon," Chemical Engineering, Vol 73 (8) 1976 (1966).
34. J. L. Rizzo and A. R. Shepherd, "Treating Industrial Wastewater With
Activated Carbon," Chemical Engineering. Vol 84 (1) 95 (1977).
BIOGRAPHY Leonard W. Crame
Leonard W. Crame is a Senior Chemical
Engineer in the Air and Water Conservation
Section of Texaco's Pror Arthur, Texas, Research
Laboratories. He has a B.S. degree in Engineer-
ing Technology (Chemical) and an M.S. in Thermal
and Environmental Engineering from Southern
Illinois University at Carbondale. Len has been
involved in several refinery wastewater treat-
ment pilot studies and conceptual designs since
Joining Texaco in 1973. He is a member of
the Texas \Afater Pollution Control Association
and has recently authored several papers on
wastewater treatment.
454
-------
DISCUSSION
E. A. Buckley. Lion Oil Co.; What were the levels of alum and polyelectro-
lyte used in the DAF pretreatment to the system?
Len Crame; In the dissolved air flotation unit we were using, we found
by the use of the zeta meter and jar tests that it required 40 milligrams
per liter of filter alum, 20 milligrams to 40 milligrams per liter of catonic
polymer Dearborn 431. It was not the intent of this study to try to zero in
on the best and most economical chemical dosage but mainly to get the soluble
feed for the bio units. We found from experience that wide fluctuations in
the feed characteristics did not affect these two chemical doses.
Ed Sebesta, Brown & Root; In the slide (Figure 2) comparing effluent and
COD concentrations from the various pretreatment systems, were the samples
filtered or unfiltered?
Len Crame; They were filtered COD's for our bio effluents and carbon
effluents. In Figure 2, I did not identify them, but it is total COD. It
does include suspended solids because we were looking for the contribution
of solids in this case.
I also would like to make the comment that I do pretty much agree
with everyone else's presentations as far as what work has been done with
carbon on the enhancement mechanism and we will continue to look at this
throughout the second phase of our pilot study. I think that you have to be
very careful in using powdered activated carbon. In a short term study I
agree with the other gentlemen (Amoco) (DuPont), that when you first put in
activated carbon you have to allow time for this matrix to form, which we
did. And you don't get the same settling effect as when you initially add
carbon. When you allow the system to come to an equilibrium and the bio-mass
starts adhering to the powdered carbon, it does greatly improve the sludge
settling characteristics, but it takes a little time. I believe that when
you initially add powdered carbon your results are going to be very good
because you're going to get a tremendous amount of adsorption. We followed
this and have seen it. I'm very hesitant about including data right after
you start running an enhanced bio-system. You will see a sharp decrease in
the effluent organic levels. You have got to wait until an equilibrium is
reached.
J. Dewell. Phillips Petroleum Co.: In your cost comparison between enhanced
activated sludge and the granular carbon, I wasn't sure if the enhanced
activated sludge assumed that the conventional activated sludge was already
in place or not. Would these comparisons still be valid on a grass-roots
.treating system?
Len Crame; We were assuming that activated sludge was already in place and
actually we were only comparing the cost of carbon contactors and regenera-
tion equipment against the additional equipment you have to put in to add
powdered carbon. We were not including filters. We would believe the filter
would have to be a part of both treatment systems and would have no ettect
on this comparison.
455
-------
J. Dewell: Do you have any feel for what would be most cost effective on a
grass-roots basis, assuming no treatment system at all exists?
Len Crame: I think it is a safe argument that activated sludge is probably
the best system for driving the effluent organic levels down for 1977 and
1983 but not necessarily the most cost effective. At this time I do not see
any other system that perhaps can be enhanced cost effectively with powdered
carbon for 1983.
J. Dewell: I was referring to a situation where one does not have activated
sludge at this time and is meeting 1977 standards so one is going to come to
1983 without any activated sludge.
Len Crame; I think this must be determined on an individual basis. As you
know when you calculate out guidelines for '77 or '83 you will find that in
some cases you are stuck with very tight guidelines for a certain parameter
and I don't think it is appropriate to say which type of treatment would be
best. We would definitely not put in any powdered-carbon enhanced system
until we piloted it and you would be taking a risk if you did. All treat-
ment systems are unique, including activated sludge systems and enhanced
biological systems. We think that powdered carbon addition has a lot of
merit, but still you should determine it on a case-by-case basis.
F. L. Robertaccio, DuPont: I think that the easiest way to look at it is
that the activated sludge system in this case is common to both the powdered-
carbon addition and the granular-carbon addition system so the difference in
cost here would have added to it the cost of the activated sludge system if
you were starting out with a brand new plant. You can use that as a first
estimate, but what we have found at the plant I talked about yesterday is
that with a grass-roots plant you have additional savings that you can accrue
to take full benefit of the system. We talked about having smaller secondary
clarifiers, higher upflow rates through the clarifiers, smaller dewatering
equipment; and having no secondary solids disposal if you go through regenera-
tion. So our experiences have been that with the grass-roots system you can
put in powdered carbon systems with regeneration for the same capital costs
and essentially the same operating costs as a conventional secondary waste-
water treatment system. If you want further reference on this there was a
paper I referred to yesterday that had details of those cost estimates.
J. E. Rucker, API; Please comment on why your COD values were greater than
those we looked at earlier this morning from the Argonne work?
Len Crame; The refinery where we were located is a very complex refinery
and I am quite sure that the refractory COD that remains is going to differ
from plant to plant. We did try to exclude everything from the chemical
plant, but I am not surprised really that we have a different refractory
COD level and I don't think you can compare the refractory COD's out of
these carbon systems from plant to plant and find a great consistency as far
as concentration goes.
456
-------
Jeffrey Chen. Dravo; What would you propose to use to treat the sludge
generated from the pretreatment unit? Will the cost associated with the
treatment be cost effective when compared to the improvement of the following
bio system?
Len Crame; We were assuming that for our best case here that when we were
comparing granular carbon with powdered carbon you would have a primary
sludge treatment and disposal problem in both cases so that it really
doesn't affect our economics here. Sludge disposal is another problem and
again it does depend upon the availability of land and other considerations
and it is just something totally different; but actually we're comparing the
two systems here and assuming that primary sludge is going to be a problem
in both cases. You would have to do a cost effectiveness study on the pre-
treatment and sludge disposal cost vs the benefit obtained from it. But
from an operational standpoint, once you get the colloids and oil out it is
much easier to operate the activated sludge process, since the oil and solids
interfere with flocculation and sludge settling.
Tom McConomy, Calgon Corp.; During the period you were operating the granu-
lar carbon columns, was the carbon changed or was the same carbon used during
the entire test?
Len Crame; As you will see in the paper we had four carbon columns and we
would measure TOC at intermediate points and whenever we found a breakthrough
on the first carbon column, we would shift the carbon columns and put a fresh
regenerated column on the tail-end of the system. This is why we are confi-
dent that the final column effluent is representative of what carbon adsorp-
tion can do with the regenerated carbon. We did try to determine how much
we had in those columns and we were running about 0.12 to 0.15 pounds TOC
per pound of carbon, which I think is fairly typical. But because of the
activated sludge enhancement you get with powdered carbon, where I won't
necessarily say "biological regeneration" occurs, you are effectively regen-
erating it somehow by desoprtion, or whatever, within the process and that
is what makes it economical.
457
-------
FILTER
PH
CONTROL
FEED
00
ACTIVATED
SLUDGE
CONV.
0.3 F/M
HIGH SA
HIGH SA
+
PC-C
HIGH SA
+
PC-H
CONV.
0.3 F/M
FILTER
CARBON COLUMNS
HIGH SA
HIGH SA
+
PC-H
FILTER
CARBON COLUMNS
FIGURE 1 - TREATMENT SCHEMES
-------
u
o
H
w
3
600
500
400
.- 300
200
w
100
1000
900
800
700
j 600
§500
- 400
300
w
200
W
100
1—I I I I
T T
592^
^DESIGNED FOR LESS THAN OPTIMUM PRETREATMENT
I | I I I I I I I I I
5 10 20 30 40 50 60 70 80 90 95
PERCENT OF TIME LESS THAN INDICATED VALUE
FIGURE 2 - PART I - TOC AND COD REMOVAL BY PRETREATMENT
459
-------
H -
a u
wo
Fn Q
fe W
W Od
W
H
hJ
M
100
90
80
70
60
50
40
30
300
50TH PERCENTILE VALUES, MG/L
TOG COD
A,
B,
C,
D,
E,
F,
FILTER - LOW SA
DAF
DAF
DAF
DAF
DAF
HIGH SA
HIGH SA-500 MG/L PC-C
HIGH SA-500 MG/L PC-H
HIGH SA-STAGE 2
LOW SA
64
56
55
48
53
58
128
108
97
89
116
105
PERCENT OF TIME LESS THAN INDICATED VALUE
FIGURE 3 - PART I - ACTIVATED SLUDGE TOG AND COD DISTRIBUTIONS
-------
200
8100
Q 90
w 80
w 70
B 60
Ł 50
Ł 30
&
20
50
40
• 30
.J
•—.
§ 20
I1?
a f
& «
w
23
i i I I
1
95
10 20 30 40 50 60 70 80 90
PERCENT OF TIME LESS THAN INDICATED VALUE
FIGURE 4 - PART i - GRANULAR CARBON COLUMN TOC AND COD DISTRIBUTIONS
93
461
-------
700
600
500
400
>j
g 300-
u
2 200
W
100-
90-
80-
70-
60-
334
*DESIGNED FOR LESS THAN OPTIMUM
j I I II I I I
PERCENT OF TIME LESS THAN INDICATED VALUE
FIGURE 5 - PART II - TOC AND COD REMOVAL BY PRETREATMENT
462
-------
100
90
>-) 80
5 70
s 60
o 50
o
w
a
H
3
30
20
10
300
cs
g
,200
§
o
Ł100
rf 90
E 80
Ł 70
H 60
w
50TH PERCENTILE VALUES, MG/L
A, FILTER - LOW SA
B,
C,
D,
F,
G,
DAF
DAF
DAF
DAF
DAF
HIGH SA
HIGH SA-1000 MG/L PC-C
HIGH SA-1000 MG/L PC-H
LOW SA
HIGH SA-2500 MG/L PC-H
TOG
68"
44
41
34
46
22
COD
T63
111
119
89
108
72
j-^l—sb—30 4o s'o b'u ;i—s* 5*—rf—rt
PERCENT OF TIME LESS THAN INDICATED VALUE
FIGURE 6 - PART II - ACTIVATED SLUDGE TOG AND COD DISTRIBUTIONS
463
-------
100
90
80
70
^ 60
3 50
Bo" 40
w o
30
w
w
S 20
10
200
Bo
go 100
.-j 90
fe w 8C
70
60
50
H
40
30
2
I I I
I I I I I
T5—JB—SIT
PERCENT OF TIME LESS THAN INDICATED VALUE
FIGURE 7 - PART II - GRANULAR CARBON TOG AND COD DISTRIBUTIONS
464
-------
w
3500
3000
2500
2000
w 1500
o
u
H
O
H
1000
500
0
\
GRANULAR CARBON
POWDERED CARBON
12345
FLOW MM GPD
FIGURE 8 - COMPARISON OF ESTIMATED CARBON COSTS
465
-------
50
40
30
o >-<
o w
20
10
0
36
0.61
,B cc5b
44
$3.19
LB COD
POWDERED
CARBON
GRANULAR
CARBON
$14.73
LB COD
INCREMENTAL
IMPROVEMENT
FIGURE 9 - ESTIMATED EFFECTIVE CARBON COST AT 1 MM GPD
-------
PHENOLS
TABLE 1- PART II - EFFLUENT SUMMARY
(ALL VALUES MG/L)
OIL & GREASE
EFFLUENT
SAMPLE
EQ BASIN
FILTER
DAF UNIT
ASU A
ASU B
ASU C
ASU D
ASU G
ASU F
(HBON COL.
(•KIES A)
ORBON COL.
(PERIES F)
PERCENTILE
50TH 90TH
7.3 18.0
—
—
0.11
0.06
0.15
0.05
0.04
0.07
0.04
0.02
—
—
0.16
0.14
0.38
0.20
0.13
0.17
4.8
0.08
PERCENTILE
50TH 90TH
108
69
14
18
5
3
<3
<3
4
<3
<3
191
130
19
38
7
9
5
3
7
6
4
PERCENTILE
50TH 90TH
64.0 119
34.0
13.0
86.0
27.5
23.0
18.0
22.0
8.4
—
—
74.0
21.0
149
41.0
77.0
57.0
44.0
28.0
—
—
PERCENTILE
50TH 90TH
20.9 27.2
—
—
11.5 20.0
3.9 9.2
3.3 5.4
3.1 4.8
3.1 4.4
2.3 5.4
— —
—
-------
TABLE 2 -PART II - SLUDGE DATA
AVERAGE VALUE ACTIVATED SLUDGE UNIT
NOMINAL LOADING
ACTUAL LOADING
MLSS, MG/L
MLVSS, MG/L
PC, MG/L
-P» V \JCC
CTi '" VŁ>::i
OO
OXYGEN UPTAKE, LB 02/LB COD REM
MG/L-MIN
SVI, ML/G
SETTLING VELOCITY, FT/MIN
BIOMASS PRODUCTION RATEd
TOTAL PRODUCTION RATE^
A
F/M =0.3
F/M = 0.3
1,487
1,302
0
88
0.40
0.16
95
0.17
0.22
0.25
B
50-DAY SA
39-DAY SA
1,892
1,562
0
83
0.68
0.12
64
0.34
0.16
0.19
C
50-DAY SA
42-DAY SA
2,728
2,269
l,000a
83
0.71
0.12
41
0.38
0.12
0.17
D
50-DAY SA
44-DAY SA
2,720
2,416
l,000b
89
0.49
0.11
43
0.38
0.11
0.14
F
F/M = 0.3
F/M =0.3
745
689
0
92
0.61
0.11
91
N/AC
0.19
0.21
G
50-DAY SA
56-DAY SA
4,096
3,898
2,500b
95
0.47
0.10
30
0.39
0.08
0.09
aCONVENTIONAL-SURFACE-AREA.
bHIGH-SURFACE-AREA.
CDISCRETE SETTLING
dLB VSS/LB COD REMOVED.
eLB TSS/LB COD REMOVED (INCLUDES CARBON)
-------
TABLE 3-POWDERED CARBON (PC) REQUIREMENTS
ASU-PART
C-I
D-I
C-II
D-II
G-II
PC
LEVEL
(MG/L)
500
500
1,000
1,000
2,500
PC
TYPE
PC-C
PC-H
PC-C
PC-H
PC-H
PC
INVENTORY
(G)
28.4
28.4
56.7
56.7
141.9
AVG PC
LOSS
(G/DAY)
0.68
0.56
1.50
1.12
2.21
PC
MAKEUP
(%)
2.4
2.0
2.6
2.0
1.6
469
-------
SESSION IX
COSTS/BENEFITS
Chairman
Thomas L. Hurst
Corporate Director of Safety and Environmental Services
Kerr-McGee Corporation, Oklahoma City, Oklahoma
Speakers
W. Wesley Eckenfelder, Jr.
"Overview of Costs/Benefits"
Lial F. Tischler
"Treatment Cost-Effectiveness as a Function of
Effluent Quality"
Carl E. Adams, Jr. and John H. Koon
"The Economics of Managing Refinery Sludges"
Melville Gray
"Compliance Monitoring Costs for the Priority
Pollutants"
Robert F. Babcock, Leo J. Duffy and Gilbert G. Jones
"Analytical Costs in the Problem Pollutants"
470
-------
BIOGRAPHY Thomas L. Hurst
Thomas L. Hurst is Corporate Director
of Safety and Environmental Services for Kerr-
McGee Corporation in Oklahoma City. He has
a B.S. degree from North Carolina State Univer-
sity; M.S., University of Washington; and Ph.D.
(Engineering), University of Illinois; and is a
registered professional engineer in Oklahoma.
He is a member of AIME, American Petroleum
Institute, American Mining Congress, American
Chemical Society, Air Pollution Control Asso-
ciation, and the Water Pollution Control Fede-
ration.
471
-------
"OVERVIEW OF COSTS/BENEFITS"
W. Wesley Eckenf elder, Jr.
Distinguished Professor of Envrionmental and
Water Resources Engineering, Vanderbilt University
Andrew Edwards
Project Engineer, Associated Water and Air Resources Engineers
INTRODUCTION
It is difficult to interpret cost/benefits in the classic sense when
considering present water pollution control limitations for industrial
effluents. Two cases will be considered, the first involving a water
quality limiting discharge, and the second related to present effluent
guideline limitations.
A water quality limiting discharge will usually involve effluent re-
quirements more stringent than that imposed by effluent guidelines and in
many cases require wastewater treatment beyond conventional secondary
treatment processes. This discussion will relate to parameters pertinent
to the oxygen balance in the receiving water and the cost effectiveness of
wastewater treatment relative to this impact.
WATER QUALITY
Several factors should be considered in the water quality limiting case.
A mixed wastewater may contain many organics of varying biodegradability.
The overall rate will be the sum of the individual rates. For example a
recent study involving treatment in the activated sludge process using a
mixture of glucose, phenol and sulfonilic acid showed individual removal
rates of 0.072 mg/mg/day, 0.049 mg/mg/day and 0.015 mg/mg/day respectively
with an overall removal rate of 0.130 mg/mg/day. The removal rate coefficient
K can be computed from Equation (1)
S - S S
_ „ _e m
* K ^ UJ
X t
v o
The biodegradation rate coefficient, K, as defined by Equation (1) reflects
the overall rate for the wastewater in question. Removal of the more
readily degradable organics through a treatment facility, i.e. the glucose
and the phenol in the above example, will cause a reduction in the de-
oxygenation rate coefficient (K) of the residual organics discharged to the
receiving stream. For example, a raw wastewater might exhibit a BOD rate
472
-------
coefficient of 0.25/day, while after treatment and removal of the readily
degradable organics, the resulting BOD rate coefficient might be 0.08/day.
Hypothetically, discharging the same pounds of BOD will cause less of an
oxygen deficit if the deoxygenation rate is low as opposed to a high K rate,
as shown in Figure 1. This illustrates the fallacy of only considering the'
Ibs. of BOD discharged without regard to the resulting deoxygenate rates in
the receiving stream. Therefore, relative to the impact on the receiving
stream, both the quantity of BOD as well as the biodegradation rate of the
residual BOD should be considered. The effect of biodegradability on
activated sludge plant costs are shown in Figure 2. It becomes apparent
that for cost effective design, higher concentrations of effluent soluble
BOD can be discharged for wastewaters of low biodegradability without as
significant an effect on the oxygen balance in the receiving stream.
There is an increasing emphasis today on nitrification in a wastewater
treatment facility, particularly as it relates to the oxygen balance in
the receiving water. Removal of carbonaceous organics may move the nitri-
fication oxidation upstream closer to the wastewater discharge resulting in
greater depletions of oxygen. This phenomena is accentuated when the
wastewater treatment plant is nitrifying and thereby discharging increased
numbers of nitrifying organics to the stream. Temperature has a major
effect on the nitrification process, both in the treatment plant and the
receiving stream. A cost effective wastewater treatment plant might be
designed to produce nitrification during the summer months when oxygen
depletion into the receiving stream would be greatest. Nitrification would
not be significant during the winter months when the nitrification rate in
the stream and the resulting numbers of nitrifyers from the plant would be
minimal. The effect of temperature on nitrification design and resulting
costs is shown in Table 1.
TABLE 1 "EFFECT OF TEMPERATURE ON NITRIFICATION DESIGN AND COSTS"
RAW WASTE LOAD 10 mg/1
150 mg/1 BOD
20 mg/1 NH3-N
EFFLUENT QUALITY 15 mg/1 BOD
2 mg/1 NH3-N
TEMPERATURE, °C 10 25
SLUDGE AGE, DAYS 14 5
AERATION BASIN VOLUME, 2.0 1-0
MIL. GAL.
AERATION, HP 200 170
CONSTRUCTION COST $696,000 $481,000
Significant capital and operating costs for constfu^io\a^P™^ctor **"
quired to achieve nitrification under cold conditions
which nee'ds to be considered in the oxygen balance in the receiving stream
is the fact that many substances that depress the reaeration coetticient,
473
-------
such as surface active agents present in raw wastewaters, are removed in
wastewater treatment. The net effect is to increase the reaeration coefficient
in the receiving water with increasing degrees of wastewater treatment. This
has the net effect of permitting higher organic loads without further oxygen
depletion. Because of the nature of the biodegradation process, both in the
wastewater treatment plant and the receiving stream, cost effective design
and operation should lead to a two-tiered standard in those parts of the
country where cold weather temperatures effect the biological oxidation
process, both in the treatment facility and in the receiving water.
EFFLUENT GUIDELINES
The second case considers those plants subject to effluent guideline
limitations. It should be recognized that most industrial plants are or
will be coming into compliance with Best Practicable Control Technology
(BPT) regulations effective July 1, 1977. This would imply that most
industrial plants discharging organic wastewaters will have installed
biological wastewater treatment, and that any consideration of additional
reduction in pollutional loads should consider the existence of a bio-
logical wastewater treatment facility at that time. The original effluent
limitations relating to Best Available Technology Economically Achievable
(BATEA) generally considered some in-plant reductions in wastewater volume
and principally add-on end-of-pipe treatment units such as filtration and
carbon adsorption. Cases that have been evaluated by the writers would
indicate that other approaches to effluent quality improvement may be con-
siderably more cost effective than indiscriminant add-on treatment facilities.
These approaches are:
1) In-plant changes to eliminate or reduce pollutional loads.
2) Installation of treatment systems for process modification at
specific discharge sources to eliminate, reduce or modify the
wastewater characteristics to render them more compatible with
existing wastewater treatment facilities.
3) Add-on tertiary*treatment units to the existing wastewater
treatment facility.
A detailed study was conducted for the Effluent Standards and Water Quality
Information Advisory Committee (EPA) in order to define the cost effective-
ness of pollution reduction by in-plant changes with existing treatment
facilities as compared to additional end-of-pipe wastewater treatment as
defined by the then BATEA criteria. This study considered an activated
sludge plant in place at the time improved effluent quality was to be
considered. The alternatives considered are coagulation, filtration,
carbon adsorption and in-plant changes to reduce wastewater flow and strength.
The in-plant changes include equipment revision and additions, unit shut
downs, scrubber replacement, segregation, collection and incineration, raw
material substitutions, reprocessing and miscellaneous small projects.. ,
Table 2 summarizes the results of this study. Cost relationships for COD
removal for the various options are shown in Figure 3. It is apparent that
474
-------
little benefit in effluent quality is gained by adding carbon adsorption to the
activated sludge plant effluent over in-plant changes with biological and
chemical treatment. As can be seen, additional end-of-pipe treatment,
including filtration and carbon adsorption, resulted in an increased removal
of 23,930 pounds of COD per day at an annual additional cost of $4.1 million
dollars per year. In^contrast, in-plant changes with minimal additional
treatment, i.e., chemical coagulation and filtration, resulted in a COD
reduction of 19,260 pounds per day at a cost of $1.3 million dollars per year.
It is readily apparent from Figure 3 that a cost effective analysis would
mitigate against the application of carbon adsorption for the minimal
improvement in effluent quality achieved.
In most cases, effluent treatment facilities have been designed to treat
total wastewater discharges to levels consistent with effluent guidelines
limitations established by EPA for specific industrial categories and sub-
categories. In many cases, removal of particular constituents which are in-
hibitory to the biological treatment process or possess a very low degradation
rate,by treatment of these constituents at their source can result in a
marked improvement in performance and an increase in capacity of existing
biological wastewater treatment facilities. Table 3 illustrates a case in
which one wastewater stream markedly reduced the overall biodegradation rate
in the biological wastewater treatment facility.
TABLE 3 "REACTION RATE COEFFICIENTS WITH AND WITHOUT
CARBON PRETREATMENT OF A PESTICIDE WASTEWATER"
Non-Carbon Treated Carbon Treated
T °C K I/day K I/day
28°C 2.25 23.1
8°C 0.81 6.5
Removal of this constituent by separate carbon adsorption treatment rendered
the total wastewater stream considerably more degradable. This would sub-
stantially reduce the effluent pollutant levels from the biological treatment
facility or permit higher organic loadings through the facility with
resultant reductions in effluent discharges. An evaluation of both the bio-
degradability and the effect of wastewater constituents on biodegradation on
specific sources within the industrial facility should in many cases lead to
marked improvement in both wastewater treatment plant operation and effluent
variability. This is particularly true where new products are to be intro-
duced into the plant which may affect the overall biodegradation character-
istics.
A few years ago post-tertiary treatment from the petroleum and chemicals
industries generally considered filtration followed by granular carbon
adsorption columns. Recent developments in the field show promise for other
processes and process modifications which are considerably more cost
effective for further pollutant reduction. These include the application ot
475
-------
polymers or coagulants prior to the final clarifier for improved suspended
solids reduction. This approach involves minimal capital expenditures. The
application of powdered activated carbon in conjunction with the biological
treatment process has shown considerable promise in recent plant and pilot
plant studies. In fact, several papers at this seminar have reported on the
results of such studies. Modification to post filtration systems using
modified media, such as compacted clay or coal, to encourage and enhance bio-
logical action has shown the capability of reducing effluent soluble BOD
levels to less than 5 mg/1 and suspended solids to less than 10 mg/1. In
this case post filtration serves the dual purpose of removal of suspended
solids and further biological oxidation, since oxygen is limiting BOD
reductions in the order of 10-20 mg/1 are feasible. In some cases, depending
on oxygen limitation, nitrification might also be achieved in this process.
Most of the data available to the writers at this time involves application
of this process to domestic wastewater. Further experimental studies would
be required to indicate the feasibility of applying such processes to
tertiary treatment of industrial wastewaters.
BIOGRAPHY
W. Wesley Eckenfelder holds a B. E. from
Mahattan College; a M.S. from Pennsylvania State
University; and a M.E. from New York University.
Professor Eckenfeleder has served on the faculty
of Manhattan College, the University of Texas at
Austin, the University of Delft in Holland, and
is now Distinguished Professor of Environmental
and Water Resources Engineering at Vanderbilt
University. Wes is also Board Chairman of
AWARE, Inc. and has written over 200 publications
and 11 books on water quality management. Profes-
sor Eckenfelder has received many awards.
Andrew W. Edwards is a Project Engineer
with the Operatinal Services Division for Associ-
ated' Water and Air Resources Engineers, Inc. (AWARE)
of Nashville, Tennessee. He holds the following
degrees: B.E. Chemical Engineering, Vanderbilt
University and M.S. Chemical Engineering,
Vanderbilt University. He is a professional
engineer in the State of Tennessee.
476
-------
POINT OF DISCHARGE
O
(T
S
(fl
H-
o
= 0.25/day
MILES DOWNSTREAM
FIG. 1. EFFECT OF DEOXYGENATION RATE ON OXYGEN SAG CURVE
477
-------
3,000
o 2,500
o
o
o
2,000
o 1,500
o:
CO
0 1,000
500
BASIS 10 mgd
BOD 150 mg/l
(ACTIVATED SLUDGE ONLY)
Low K K= I day-I
10 20 30
EFFLUENT SOLUBLE BOD, mg/l
40
FIG. 2. EFFECT OF BIODEGRADABILITY ON ACTIVATED SLUDGE COSTS
478
-------
(T)
a:
<4
_J
o
Q
-. 3
Option
A
F
G
H
o
o
•
Technology
Activated Sludge and Coagulation
A and Inplant Changes
A and Inplant Changes and Filtration
A and Filtration and Carbon Adsorption
REDUCTION IN COD
(1000 Ib/day)
FIG. 3 ECONOMIC ALTERNATIVES FOR COD REDUCTION
479
-------
TABLE 2 "EFFLUENT QUALITY COST EFFECTIVE ALTERNATIVES "
Graph
Code
BPT
A
F
G
H
Pollution
Reduction
Scheme
Activated
Sludge
Activated
Sludge 5
Coagulation
All in-plant
changes 5
Activated
Sludge 5
Coagulation
All in-plant
changes 6
Activated
Sludge fi
Coagulation
§ Filtration
Activated
Sludge f,
Coagulation
6 Filtration
f, Carbon
Adsorption
Influent to
Treatment
Flow Sol. BOD
(MGD) (1000 Ib/day)
11.1 55.7
11.1 55.7
B.3 37.1
8.3 37.1
11.1 55.7
Effluent from
Treatment
Sol. BODr Tot. BOD COD TSS
> ~ nnnn iTwH-nr^ . .->
3.50 6.30 41.78 8,51
3.50 5,00 36,33 4.62
2.35 3.50 25.43 3,46
2.35 2.81 22.52 1.38
2.27 2.57 17.85 0.92
Annual Costs*
Capital Operating Total
Millions of Dollars
0.21 0.22 0.43
0.71 0.22 0.93
0.93 0.40 1.33
1.20 2.90 4.10
Costs presented are those above the cost of the installed BPT facility. All costs are 197S dollars.
-------
"TREATMENT COST-EFFECTIVENESS AS A FUNCTION OF EFFLUENT QUALITY"
Lial F. Tischler
Austin Office Manager, Engineering-Science, Inc.
INTRODUCTION
The cost of required wastewater treatment is a major concern to industry
and the petroleum refining industry is no exception. By July 1, 1977, all
petroleum refineries in the United States are required to have, as a minimum,
treatment technology that will provide an effluent of the quality which the
U.S. Environmental Protection Agency (EPA) has defined as "Best Practicable
Control Technology Currently Available" (BPCTCA). Five years later, unless
the Congress acts to modify Public Law 92-500, the next level of treatment
technology, designated as "Best Available Technology Economically Achievable"
(BATEA), will be required. In the interim period, effluent limitations for
certain toxic constituents in wastewaters, probably including petroleum
refining effluents, will be promulgated by EPA under the auspices of Section
307 of Public Law 92-500. Thus, the petroleum refining industry will be
required to implement even more costly methods of wastewater control than
those which have been applied to date.
This paper addresses the costs associated with the end-of-pipe treatment
processes which are most frequently associated with petroleum refinery secondary
and tertiary wastewater treatment. The objective is to provide a frame of
reference for evaluating treatment costs as a function of increasingly stringent
effluent quality limitations. As such, the information presented herein is
not all-inclusive in terms of either the costs of reaching a given effluent
quality or the end-of-pipe treatment processes considered. Because each
refinery has different wastewater problems, it is impossible to accurately
assess the effectiveness and cost of in-plant controls for reducing or elimi-
nating raw waste load. Therefore, these costs are not included in this analysis.
Furthermore, to simplify the evaluation and allow comparison of these costs
with costs in other industrial categories and between individual petroleum
refineries, it is assumed that the costs are based on treatment units follow-
ing primary and secondary oil/solids separation. Finally, a cost comparison
of various treatment alternatives at a given level of effluent quality is not
the subject of this paper, but it should be recognized that in a specific case
at a particular refinery, there may be several options for biological treat-
ment and effluent polishing which should be considered in order to select the
least-cost alternative.
All costs shown in this paper are adjusted to January 1, 1977. An annual
inflation rate of six percent was used to perform this adjustment and it is
recognized that this may be too conservative in some cases. To determine the
annual cost of capital, an interest rate of 10 percent amortized over 15 years
is used. No attempt was made to adjust the costs presented in this paper for
other considerations such as geographical location, the availability of existing
!and, and similar factors. Reports which have been prepared on the costs of
wastewater treatment in the petroleum refining industry have addressed these
481
-------
considerations in depth (Ref. 1, 3) and, while their importance is recognized,
their exclusion from this analysis will not affect attainment of the paper's
stated objective which is to provide a framework for evaluating cost as a
function of effluent quality. It must be emphasized, however, that the costs
presented herein are to be used only to provide order of magnitude comparisons
of one level of treatment with another and may vary substantially at given
petroleum refineries.
The remainder of this paper is organized into two main sections: the
first presenting limited cost information from petroleum refining and petro-
chemical waste treatment facilities which are considered representative for a
cost comparison of this type; and the second being an analysis of the costs
associated with a model refinery which is used as an example to illustrate the
increasing costs as the effluent quality limitations become more stringent.
The cost comparisons presented in this paper are presented on a unitized basis
using both the flow of the wastewater being subjected to treatment and the
removal of pollutants associated with each of the unit processes. This is a
useful way to evaluate cost data inasmuch as it demonstrates unit process
cost-effectiveness not only in terms of the quantity of wastewater treated,
but also in terms of the removal of pollutants which it is designed to treat.
The three levels of end-of-pipe control technology which are addressed in
this paper are biological treatment, as represented primarily by the activated
sludge process for petroleum refinery wastewaters, effluent polishing for
suspended solids removal which is represented by granular-media filtration,
and tertiary effluent polishing which is represented by activated carbon
adsorption in columns or beds. In addition, cost data are presented for the
model refinery for removal of selected potentially toxic pollutants which may
occur in petroleum refinery wastewaters.
INDUSTRIAL CASE HISTORIES
Industrial case history data for three levels of effluent control are
presented in the following sections: biological treatment, granular-media
filtration of biologically treated effluents, and activated carbon treatment
of selected industrial wastewaters. These cost data, which are primarily for
petroleum refining applications, are presented to give the reader an idea of
the range of costs which might be encountered for treating petroleum refining
wastewaters to each of the effluent quality levels which can be assumed to be
approximated by the end-of-pipe treatment technology described above.
Biological Treatment
Table 1 presents cost data for six petroleum refinery case histories
utilizing biological treatment. As described previously, these costs repre-
sent the biological treatment module only and in all cases the treatment
process has been preceded by an air flotation unit for secondary oil/solids
removal. The only solids handling costs included in Table 1 are for aerobic
digestion and thickening of waste activated sludge.
482
-------
Two basic types of biological treatment systems are shown; completely
mixed activated sludge and rotating biological surface (RBS). Other bio-
logical processes which might achieve similar levels of effluent quality,
the most notable of which for petroleum refining wastewaters is the aerated
lagoon, are not included because either insufficient data are available or it
would be difficult to make a proper comparison between the process and the two
processes shown in Table 1. The aerated lagoon falls into the second category
because it is extremely land-intensive and also cannot be assured of con-
tinually meeting the level of effluent suspended solids provided by the
activated sludge and RBS systems. This is not to say that the aerated lagoon
would not provide the necessary effluent quality and be the most cost-effec-
tive system for biological treatment at some petroleum refineries.
•The cost data for the biological treatment processes shown in Table 1 are
largely self-explanatory. However, several interesting conclusions can be
drawn from reviewing these data. First, there is suprisingly little variation
in the cost, in $/1000 gal, of biological treatment of petroleum refining
wastes considering the rather significant size and raw waste load differences
between the petroleum refineries. There is an obvious economy-of-scale as a
function of flow for both the activated sludge and RBS systems. The influence
of organic raw waste load on both the capital cost and operation and mainten-
ance (O&M) costs of biological waste treatment is demonstrated by these data.
It must be recognized that there are several factors which influence the
cost-effectiveness of biological treatment for a particular petroleum refining
or other industrial waste. One factor is the size of the facility in terms of
both quantity of wastewater treated and raw waste load. There is substantial
economy-of-scale in biological waste treatment systems at both higher flow
rates and higher organic loadings. Another important factor in determining
the costs of biological waste treatment is the biological kinetics for the
particular wastewater being treated. A wastewater with a high organic removal
rate requires a relatively smaller aeration basin but may also require higher
aeration capacity than a wastewater with a lower removal rate. These factors
have obvious influences on both capital and operating costs. Thus, it must be
recognized that the unit costs shown in Table 1 are subject to both increases
and decreases, depending upon the specific characteristics of a particular
petroleum refinery wastewater and the design of the biological treatment
facility.
Effluent Suspended Solids Removal
The next step in end-of-pipe treatment usually considered following
biological treatment is enhanced suspended solids removal. In some cases,
biological treatment systems for petroleum refinery wastewaters have con-
sistently demonstrated effluent suspended solids concentrations in the range
of 15 to 10 mg/1. This, however, is the exception rather than the rule.
At most refineries utilizing activated sludge plants, an average effluent
total suspended solids (TSS) concentration in the range of 30 to 40 mg/1 can
be expected. Tertiary effluent polishing consisting of either polishing ponds
or granular-media filtration is common for decreasing effluent suspended
solids concentrations. The polishing ponds cannot be considered as an option
which will work in all cases; therefore, only the granular-media filter systems
483
-------
are considered in this cost analysis. The filters are designed mainly to
remove TSS, however, organic materials, as measured by five-day BOD and COD,
representing the organic content of the solids removed by the filters, are
also removed. In addition to this improvement in effluent suspended solids
content and organic content, polishing filters can be considered a requirement
prior to further effluent polishing by activated carbon adsorption or more
exotic processes such as reverse osmosis.
Table 2 shows case history cost data for two different industrial
installations of biological effluent polishing. These costs represent com-
parisons of different types of granular-media filtration systems used at the
two industrial plants. The unitized capital costs reflect a large economy-of-
scale engendered by the great difference in size between the two facilities in
terms of both the quantity of wastewater treated and the pounds of suspended
solids removed. The annual O&M costs also reflect the economy-of-scale, but
to a lesser extent than the capital costs. The unitized costs are reasonably
low in terms of quantity of wastewater treated and pounds of suspended solids
removed, at least for the larger filtration system. However the effectiveness
of a filtration system for removing organic material is quite poor, as is
reflected by the high cost per pound of five-day BOD removed in the filtration
system. Thus, it can be concluded that the filters are a good unit process
for removing suspended solids if suspended solids are considered a serious
pollutant. On the other hand, the cost-effectiveness of the filters in
removing organic materials is very poor, as might be expected, and an alterna-
tive would be to design the biological system to remove more of the organics,
if this is feasible, rather than using the filtration system to remove organics.
Tertiary Organics Removal
The remanded 1983 guidelines for the petroleum refining industry were
based on the use of activated carbon adsorption to remove additional organic
constituents from biologically treated petroleum refining effluents. The
intention was to remove those organic compounds which pass through the bio-
logical treatment system, as measured by COD and total organic carbon (TOC),
and which might either exert a long-term oxygen demand or create chronic
toxicity in the receiving waters. Recently, it has been suggested that the
primary application of activated carbon might be as pretreatment for selected
wastewater streams within a plant to reduce refractory or potentially toxic
organic compounds at the source. Notwithstanding this possible future change
in application for activated carbon, there is still emphasis by the EPA on
the use of carbon adsorption technology as part of BATEA effluent treatment.
Unfortunately, no full-scale operating data are currently available for carbon
adsorption technology in the petroleum refining industry. Table 3 provides
some unitized cost information for the use of activated carbon treatment
technology in several applications including the only two full-scale plants in
the petroleum refining industry. None of the unit costs presented in Table 3
are directly comparable. They represent different types of treatment systems
and different wastewaters. The two petroleum refining systems include the
continuous carbon columns at the BP petroleum refinery in Marcus Hook, Penn-
sylvania which treat refinery wastewaters that have been pretreated by gravity
separation and granular-media filtration. Biological pretreatment is not
included in this system. The second petroleum refinery system is at the ARCO
484
-------
refinery in California. This system operates on an intermittent basis without
biological pretreatment and consists of fixed beds which are used for organics
removal from mixed storm and process wastewaters. The third set of unit costs
shown in Table 3 are estimates prepared for a tertiary carbon system following
biological treatment of mixed industrial wastewaters, including petroleum
refinery wastes. The costs presented for the above three facilities exclude
the costs of pretreatment systems. The final two sets of unitized costs shown
in Table 3 are for the use of powdered activated carbon in activated sludge
aeration basins. Obviously, the level of effluent quality attainable with
these latter two systems is not comparable with that of the former three, but
since this application is receiving considerable attention, it was included in
this Table.
Given the basis of the data provided for the activated carbon systems in
Table 3, only general comparisons can be made. However, it is easy to see
that the cost-effectiveness in terms of both quantity of watewater treated and
pounds of organic materials removed is much poorer for the activated carbon
systems than it is for biological treatment. This is not unexpected consider-
ing the fact that the activated carbon systems are both energy and labor
intensive and the carbon itself is quite expensive. In the tertiary polishing
application, the quantity of five-day BOD and COD removed is relatively small
compared to that removed in the biological treatment phase which tends to make
the unit costs of removal much higher. This phenomena is expected and, as
will be demonstrated in the next section, makes the cost of this level of
effluent treatment quite high in terms of cost per unit of pollutant removed.
MODEL REFINERY COST ANALYSIS
The preceding sections illustrated how the unit cost of waste treatment
increases substantially as the effluent quality limitations become increas-
ingly more stringent over a rather narrow range in terms of the commonly used
parameters. Since these costs reflect a wide range of waste characteristics,
specific plant characteristics, and other similar factors which result in the
rather wide ranges demonstrated by these data, it is useful to prepare an
example analysis for a specific petroleum refinery to observe how treatment
cost-effectiveness relates to effluent quality. This is accomplished by
creating a hypothetical petroleum refinery, referred to hereinafter as the
"Model Refinery," of a given size and with specified wastewater characteristics
and then preparing cost estimates for increasingly stringent levels of end-of-
pipe control. It is reemphasized that the costs presented in this paper
include only end-of-pipe treatment and exclude costs associated with in-plant
control and the pretreatment system prior to biological treatment. The costs
of solids handling, which are always significant, are not included in this
analysis.
The Model Refinery selected for the cost-effectiveness analysis is the
Subcategory B, median refinery which was used in a study for the National
Commission on Water Quality (Ref. 3). The characteristics of this hypothe-
tical refinery are shown in Table 4. It is emphasized that this refinery is
strictly hypothetical and does not represent any actual refinery.
485
-------
The wastewater characteristics at varying levels of treatment for this
Model Refinery are shown in Table 5. The values indicated in this Table can
be considered annual average concentrations and flows. This Table shows the
expected levels of effluent quality which will be obtained by the application
of biological treatment, tertiary filtration, and tertiary carbon adsorption.
These concentrations are used as the basis for the cost-effectiveness analysis.
In Table 6, the incremental mass of pollutant removed by each of the three
end-of-pipe treatment processes being considered is shown. It is obvious that
biological treatment removes the majority of the organic materials from the
petroleum refinery wastewater and the additional levels of treatment, includ-
ing tertiary filtration and carbon adsorption, remove relatively small quan-
tities. This impacts heavily on their cost-effectiveness.
The capital costs, annual O&M costs, and annual energy costs for each of
the three incremental levels of end-of-pipe treatment for the Model Refinery
wastewater were calculated using cost data from a recent study conducted for
the State of Texas. These costs are shown in Table 7. The annual O&M costs
shown in Table 7 do not include the cost of energy, which is shown separately.
These costs can be considered to be planning-level cost estimates for the Gulf
Coast area.
The data on the size and performance of the Model Refinery wastewater
treatment plant presented in Tables 5 and 6 can be used with the cost data in
Table 7 to analyze the cost-effectiveness for each of the three levels of end-
of-pipe treatment. These unit costs, presented in terms of quantity of waste-
water treated and pounds of specified pollutant removed, are presented in
Table 8. To obtain the total annual cost for a given end-of-pipe treatment
system and a selected basis, i.e., flow, BOD, etc., the two figures shown as
capital and annual O&M costs can be summed. To facilitate comparison of the
cost data shown in Table 8, several Figures have been prepared. Figure 1
shows the total annual cost of end-of-pipe wastewater treatment as a function
of annual average effluent five-day BOD concentration. The unit costs in
terms of both flow and five-day BOD removal are shown in this Figure. In
terms of unitized costs as a function of flow, the cost increase between
activated sludge and granular-media filtration is approximately 40 percent.
The same increment, when viewed in terms of additional five-day BOD removal,
represents an approximate sixfold increase in unit cost. As might be expected
by the low incremental removals of BOD shown in Table 6 and the high cost of
treatment shown in Table 7, the activated carbon adsorption unit results in an
almost exponential increase in cost for a 10 mg/1 improvement in effluent
five-day BOD concentration. However, since the activated"carbon unit is
primarily designed to remove materials which are not easily biologically
degraded, its effectiveness is better viewed in terms of COD removal, as shown
in Figure 2. Once again, the unit cost of COD removal increases very rapidly
as treatment beyond the biological step is implemented. A threefold decrease
in effluent COD concentration between biological treatment and activated
carbon levels of effluent control technology results in an approximate tenfold
increase in total annual cost, expressed as $/lb COD removed. Figure 3 ex-
presses the cost data shown in Figures 1 and 2 in a slightly different manner,
showing the cumulative total annual cost expressed as $/lb BOD or COD removed
as a function of percent removal of BOD and COD from the pretreated refinery
wastewater. Once again, this Figure shows the almost exponential increase in
/•
486
-------
cost associated with the removal of a small increment of BOD going from
biological effluent to the activated carbon system, whereas it shows an almost
linear increase in the cost of COD removal as a function of percent removal.
Figure 4 presents the total annual costs of end-of-pipe waste treatment
at the Model Refinery from a different perspective than that presented in the
preceding Figures and Tables. This Figure illustrates the actual cumulative
total annual costs for end-of-pipe waste treatment for the Model Refinery and
shows the distribution of these costs between O&M and the cost of capital for
each type of facility. This is simply a graphical representation of the costs
shown in Table 7 but illustrates effectively some interesting relationships
not shown by the unitized costs. The addition of granular-media filtration to
an activated sludge system results in about a 35 percent increase in total
annual cost for the end-of-pipe treatment system excluding primary and second-
ary oil/solids removal. Most of this cost is associated with the capital
costs of the filter system as the increase in annual operation and maintenance
costs due to the filter is only about 10 percent of the same costs for the
biological system alone. When the carbon adsorption step is added for
removal of refractory organics, however, the O&M costs actually exceed the
annual costs for amortizing the capital investment in the activated carbon
units. This is brought about principally by the energy costs associated with
carbon regeneration and the cost of purchased carbon for makeup of system
losses. The increase in total annual costs due to the addition of the carbon
system is about 390 percent of the cost of the biological system alone but is
a far smaller percentage increase than if the costs are considered on a unit-
ized basis, as illustrated in Figures 1 through 3. This serves to reemphasize
the point made earlier: the capital costs of implementing the carbon technology
for polishing of biological effluents are not significantly greater than the
cost for biological treatment plus filtration and are less than the overall
costs of meeting BPCTCA effluent quality. However, the cost of this level of
treatment per unit of pollutant removed is many times that of the biological
system and the petroleum refineries will pay dearly for every mg/1 of COD
removed by this method. This raises very strongly the question of whether or
not the uniform application of carbon adsorption technology to polishing
biological effluents is a necessary and cost-effective requirement for meeting
the quality standards set for our nation's waters.
The preceding cost evaluations have dealt with the end-of-pipe treatment
systems which will be used in petroleum refineries to remove organic substances
measured as BOD, COD, and suspended solids. An extremely important considera-
tion over the next several years will be the impact on the petroleum refining
industry of the toxic effluent limitations which are currently being prepared
by EPA in accordance with Section 307 of Public Law 92-500. In a consent
decree from the Circuit Court of Appeals in Washington, D.C., the EPA agreed
to promulgate toxic effluent limitations for selected constituents from a list
of some 65 different classes of compounds. As of the date of this paper, some
109 different chemical compounds are being studied in detail to determine
whether or not toxic limitations should be promulgated under Section 307.
Since these limitations have not yet been promulgated, an evaluation of the
cost-effectiveness of the associated treatment technology is not possible at
this time. However, as an example, several constituents which may be found in
petroleum refinery wastewaters can be used to evaluate cost-effectiveness of
treatment processes designed for specific constituent removal.
487
-------
Cadmium, cyanides, and chromium can be found in some petroleum refinery
wastewaters and can be toxic to certain organisms in the receiving waters if
they are discharged at high enough concentrations. There is, however, con-
siderable controversy as to the toxic concentrations of many of these materials
and the selection of these for an example cost-effectiveness analysis is not
to be construed as suggesting that the limitations selected for the performance
of treatment processes to remove these constituents represent their toxic con-
centrations in receiving waters.
Cadmium is not present in most crude oils; therefore, the source of
cadmium in petroleum refinery wastewaters is either intake water or cadmium
addition during processing (Ref. 3). Cadmium in petroleum refinery waste-
waters can generally be traced to intake water, corrosion products, the
addition of cadmium compounds for distillate desulfurizing, or as a lube oil
additive to prevent oxidation. Because of the diffuse nature of the cadmium
sources in petroleum refineries, it is not generally practical to attempt to
segregate cadmium-bearing waters for separate treatment. Thus, for the Model
Refinery used in the cost analysis, treatment for removal of cadmium involves
treatment of the entire process wastewater effluent. Lime precipitation and
filtration is probably the most dependable method of removing cadmium. Cad-
mium forms an insoluble and highly stable hydroxide precipitate at an alkaline
pH and, in the absence of appreciable complexing agents, precipitation and
filtration provide effective removal. The completeness of the reaction is a
function of pH and lime addition to pH 10 is required to meet the concentra-
tion limit selected for this evaluation, which is 0.05 mg/1. The cadmium
removal process selected requires provisions for lime addition, rapid mix,
flocculation, sedimentation, filtration, and reneutralization to pH between 6
and 9. Table 9 shows the total annual costs, the amortized capital costs, and
the annual 0&M_costs for removing cadmium from the Model Refinery wastewater
stream of 1.0 MGD. The total annual cost of $0.40 per 1000 gallons is approxi-
mately two-thirds due to O&M costs, most of which are associated with the
purchase of lime, and one-third due to the amortized capital costs. No cost
per pound of cadmium removed can be calculated for this process because of
insufficient information on cadmium concentrations in petroleum refining
effluents. However, it is known that these concentrations are generally quite
low, and thus the cost per pound of cadmium removed is likely to be extremely
high.
Simple and complex cyanides are generated in cracking and coking
operations as reaction products and enter the effluent from overhead receivers
and through washing operations (Ref. 3). Most or all of the cyanide genera-
tion in a refinery occurs in these two operations, so the cyanide-bearing
water can generally be isolated from other waste streanis within the refinery
complex and treated separately. Alkaline chlorination is the most proven
treatment technology for removing cyanides from wastewaters although its
application in the petroleum refining industry has not been extensively tested
(Ref. 3). Cyanides can be reduced to less than 0.025 mg/1 with appropriate pH
control, chlorine dosage, and residence times. Biological treatment of re-
finery wastewaters is quite effective in reducing cyanides to concentrations
below 0.5 mg/1, however, for the purposes of this cost-effectiveness analysis
it is assumed that alkaline chlorination will be required to produce a final
effluent concentration of approximately 0.025 mg/1 total cyanides. In the
488
-------
Model Refinery this will be accomplished by segregating the sour water streams
after stripping, from the fluid catalytic cracker and coker and subjecting
them to alkaline chlorinatlon. The total estimated quantity of the wastewaters
requiring treatment in the model refinery is 150,000 gpd, containing approxi-
mately 100 mg/1 of total cyanides. The annual costs associated with treatment
of these waste streams in the Model Refinery to remove cyanides are shown
in Table 9. The capital costs of the cyanide removal system are quite low,
since the only equipment items required are a chlorinator, reaction tank, and
a caustic feed system. However, the annual O&M costs per 1000 gallons of
waste treated are extremely high, resulting in a total annual cost of $2.24
per 1000 gallons. In terms of total annual cost per pound of cyanide removed,
the cost-effectiveness of this unit operation is even poorer, resulting in a
cost of $2.68 per pound of cyanide removed on an annual basis. The majority
of the costs for this process are associated with the neutralization chemical,
sodium hydroxide, and the chlorine. Another potential drawback of this pro-
cess which should be mentioned is the possible formation of chlorinated hydro-
carbons which, of themselves, may prove to be toxic materials.
The final specific constituent considered for removal is chromium. The
major source of chromium within petroleum refineries is the chromate-based
corrosion inhibitors used in cooling towers. These inhibitors find their way
to the wastewater treatment plant in cooling tower blowdown. Chromium is
included in the effluent limitations for the petroleum refining industry which
have been promulgated by EPA, but it is also being considered under the toxic
effluent limitations being reviewed at this time. In petroleum refineries,
the effluent chromium limitations can usually be met without resorting to
separate treatment of cooling tower blowdown. However, in many other types of
industrial plants, particularly in the organic chemicals industry, cooling
blowdown constitutes a very high volume and is not treated in the biological
waste treatment plant. In these cases, it is often necessary to remove
chromium from the cooling tower blowdown before it is discharged to the
receiving waters and several treatment methods have seen wide application.
The types of treatment technology used for removing chromium from cooling
tower blowdown include chemical reduction/precipitation, ion exchange, and
electrolytic reduction/precipitation. Each of these treatment processes has
advantages and disadvantages. The electrolytic process is finding increasing
applications for treating cooling tower blowdown and is used in this example
cost analysis for the Model Refinery. It is assumed for the purposes of this
example that the Model Refinery generates approximately 190,000 gpd of cooling
tower blowdown which is treated by the electrolytic reduction/precipitation
method. The costs of applying this treatment procedure are shown in Table 9.
The total annual cost for chromium removal by this method for this size
facility is approximately $0.66 per 1000 gallons, but is $4.06 per pound of
chrome removed. This assumes an influent chromate concentration in the
cooling tower blowdown of approximately 20 mg/1 and an effluent concentration
from the unit of approximately 0.5 mg/1 of chromate.
SUMMARY
rjhls paper has presented an evaluation of the cost-effectiveness of
several important processes in petroleum refinery wastewater treatment using
both actual cost data from full-scale facilities and an example cost analysis
489
-------
for a hypothetical petroleum refinery. The salient points of this evaluation
can be enumerated as follows:
(1) Cost-effectiveness, expressed on a unit cost basis as a function of
effluent quality, should be based upon the constituent or con-
stituents of the wastewater which are to be removed by the waste
treatment process being evaluated, if the most meaningful compari-
sons are to be made.
(2) In terms of removal of organic materials from petroleum refinery
effluents, biological treatment is by far the most cost-effective.
Suprisingly, multimedia filtration is the second most cost-effec-
tive, and tertiary effluent polishing with carbon adsorption is the
least cost-effective of the likely alternatives for this applica-
tion.
(3) The poor effectiveness of carbon adsorption technology when used on
a biologically treated effluent for the purpose of removing trace
refractory materials raises a serious question as to whether this
is the most suitable application for this treatment process. In
the absence of known specifically toxic organic constituents,
questions should be raised about the efficacy of removing trace
residual organics, as measured by COD or TOG, simply for the purpose
of reducing the effluent concentrations unless there is a con-
commitant improvement in receiving water quality. This particular
point was made quite effectively by Mr. Joe Moore at last year's
Open Forum. At that time he indicated that the cost-effectiveness
of BATEA technology should be carefully considered before requiring
nationwide implementation (Ref. 6).
(4) The cost-effectiveness of unit processes designed for the removal of
specific constituents in petroleum refining waste streams will
probably be quite poor in terms of unit cost per quantity of waste-
water treated or constituent removed. This is not unexpected,
however, and, if a constituent is truly toxic at the effluent
concentrations being considered, the unit cost of removing the
constituent has no real meaning since Public Law 92-500 precludes
discharge of toxic materials in toxic quantities.
(5) Expressing the performance of various waste treatment unit processes
in terms of unit costs per quantity of flow treated or mass of
constituent removed is an effective basis for cost comparisons
between alternative unit processes. It is a convenient method for
putting costs of treatment processes designed to perform a given
service on a consistent basis and allows a direct comparison of the
respective costs of each option.
490
-------
REFERENCES
1. Brown & Root, Inc. "Economics of Refinery Wastewater Treatment,"
American Petroleum Institute, Publication Number 4199, Washington
D.C. (August 1973).
2. Gulp, G.L. and A.J. Schukrow. "What Lies Ahead for PAG?," Water and
Wastes Engineering, pages 67-74 (February 1977).
3. Engineering-Science, Inc. "Petroleum Refining Industry - Technology
and Costs of Wastewater Control," National Commission on Water Quality,
Washington, D.C. (June 1975).
4. Engineering-Science, Inc. Selected technical reports (1970-76).
5. McCrodden, B.S. "Treatment of Refinery Wastewater Using Filtration
and Carbon Adsorption," Paper presented at a Technology Transfer
Seminar jointly sponsored by Environment Canada, The Pollution Control
Association of Ontario, and the Canadian Society of Chemical Engineers
(October 24, 1974).
6. Moore, J.G. "The Role of the National Commission on Water Quality
(NCWQ)," Proceedings of Open Forum on Management of Petroleum Refinery
Wastewaters, EPA, API, NPRA, and the University of Tulsa, Tulsa,
pages 39-50 (1976).
7. Prosche, M.A. "Activated Carbon Treatment of Combined Storm and Process
Waters," Proceedings of Open Forum on Management of Petroleum Refinery
Wastewaters, EPA, API, NPRA, and the University of Tulsa, Tulsa,
pages 399-410 (1976).
8. Rizzo, J.A. "Case. History: Use of Powdered Activated Carbon in an
Activated Sludge System," Proceedings of Open Forum on Management of
Petroleum Refinery Wastewaters, EPA, API, NPRA, and the University of
Tulsa, Tulsa, pages 359-374 (1976).
BIOGRAPHY Lial F. Tischler
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 Tpps. Prior to joining Engineering-Science,
> 'tlte was Director of Systems Engineering at
Texas Water Development Board.
491
-------
CUMULATIVE TOTAL ANNUAL COST OF END-OF-PIPE TREATMENT
N)
N
sŁ
*r"^f ^^
m ^
31
m>
-< o
o
m
m
m
jp m
CD
o
in
>
I
m
m o
m
O |M
o o
r>
IN)
P
KI
01
O
O
p
in
o
FLOW $/IOOO GAL
01
o
o
ro
01
CD
ro
b
o
•ACTIVATED CARBON-
GRANULAR MEDIA FILTRATION
•ACTIVATED SLUDGE
8
-i
ro
b
o
OJ
b
o
b
o
o
.
b
o
00
b
o
-Q3AOW3M aoa
-------
1.75
z
LJ
i
1.50
1.25
iL
o
UJ
o
O
> 1.00
o:
«» 0.75
0.50
0.25
ZERO COST IS
AT39Omg/l
COD, AFTER
PRETREATMENT
25
50
75
100
125
150
ANNUAL AVERAGE EFFLUENT COD CONCENTRATION
mq/l
Fig.2COST OF WASTE WATER TREATMENT FOR
MODEL REFINERY AS A FUNCTION OF
EFFLUENT COD
493
-------
8.00 i—i
ID
o
7.00
o
O o
LU
LJ
o:
6.00
5.00
CO
O
o
.n
•w-
4.00
LU
> 3.00
2.00
1.00
70 80 90 100
PERCENT REDUCTION IN DESIGNATED CONSTITUENT-
ZERO IS AIR FLOTATION UNIT EFFLUENT QUALITY
Fig.3 UNITIZED TOTAL ANNUAL COST OF
WASTE TREATMENT AS A FUNCTION OF
REMOVAL EFFICIENCY-MODEL REFINERY
494
-------
400,000
300,000
H-
(f)
O
O
UJ 200,000
O
100,000
ACTIVATED GRANULAR TERTIARY
SLUDGE MEDIA ACTIVATED
FILTRATION CARBON
Fig.4TOTAL ANNUAL COSTS OF WASTE
TREATMENT-MODEL REFINERY
495
-------
TABLE 1
CASE HISTORIES OF BIOLOGICAL TREATMENT AT PETROLEUM REFINERIES
System
Type
Rotating
Biological
Surface
Rotating
Biological
Surface
Activated
Sludge
Activated
Sludge
Activated
Sludge
Activated
Sludge
Flow
(MGD)
0.9
4.3
4.6
3.9
1.9
2.3
Organic Removal
BOD5 COD
(Ib/day) (Ib/day)
840 1,200
5,300
9,100 26,200
41,500 83,000
8,600 13,000
3,100
Amortized Capital Cost*
($71000 gal) ($/lb BOD) ($/lb COD)
0.23 0.24 0.16
0.19
0.12 0.06 0.02
0.19 0.02 0.01
0.24 0.05 0.03
0.19 0.14
Annual Operation and
Maintenance Costs
($71000 gal) ($/lb BOD) ($/lb COD)
0.06 0.12 0.05
- - -
0.10 0.05 0.02
0.07 0.006 0.003
0.12 0.03 0.02
— — —
* Adjusted to 1977 and 10 percent interest for 15 years.
(Ref. 4)
METRIC CONVERSIONS _
I/sec = (MGD) (43.81)
kg = (Ib)(0.454)
1000 liters = (1000 gal)(3.785)
-------
TABLE 2
Type
Filter
System
Downflow
Deep Bed
Downflow
Shallow Bed
Downflow
Deep Bed
Gravity
Downflow
Deep Bed
Pressure
Tri-media
Gravity
Tri-media
Pressure
* Adjusted
(Ref. 4)
INDUSTRIAL CASE
Solids
Flow Removal
(MGD) (Ib/day)
1.44 240
1.44 240
48.5 24,269
48.5 24,269
48.5 24,269
48.5 24,269
to 1977 and 10 percent
HISTORIES OF
TERTIARY GRANULAR-MEDIA
Amortized Capital Cost*
($/1000 gal)
0.19
0.14
0.05
0,07
0.04
0.04
interest for
($/lb TSS) ($/lb BOD)
1.15 3.83
0.85 2.85
0.11 0.36
0.13 0.44
0.08 0.28
0.09 0.29
15 years.
FILTRATION
Annual Operation and
Maintenance Costs
($71000 gal) ($/lb TSS) ($/lb BOD)
0.07 0.41 1.37
0.06 0.34 1.14
0.04 0.07 0.22
0.03 0.07 0.22
0.03 0.07 0.22
0.03 0.06 0.19
METRIC CONVERSIONS
I/sec = (MGD)(43.81)
kg = (Ib)(0.454)
1000 liters = (1000 gal)(3.785)
-------
TABLE 3
00
T? T S\T,T
E J.OW
System (MGD)
Petroleum Refinery - '
Process wastes without 2.16
biological treatment,
continuous columns
Petroleum Refinery -"'
Storm/process water 0.85
without biological
treatment, fixed beds**
(4)
Mixed Industrial - '
Tertiary treatment , AiMt 48.5
continuous columns
(a)
Petroleum Refinery -
Powdered carbon 1.08
addition to
aeration basin
(21
Municipal - '
Tertiary powdered
activated carbon 10.0
addition to
aeration basin
* Adjusted to 1977 and 10 percent
CASE HISTORIES OF INDUSTRIAL ACTIVATED CARBON TREATMENT
Organic Removal Annual ODcration and
BOD5 COD Amortized Capital Cost* Maintenance Costs
(Ib/day) (Ib/day) ($/1000 gal) ($/lb BOD) ($/lb COD) ($/1000 gal) ($/lb BOD) ($/lb COD)
783 2,090 0.31 0.62 0.23 0.134 0.37 0.14
1,311 - - - 0.56 - 0.25
5,600 67,550 0.17 1.43 0.12 0.27 2.31 0.19
11 347 None, Manual addition of 0.03 2.72 0.09
powdered activated carbon
0.08 - - 0.02
interest for 15 years. METRIC CONVERSIONS
** Intermittent operation.
***Excludes cost of necessary pretreatment by granular-media filtration.
Note: Numbers in parentheses following system description indicate references which
can be found at the end of the text.
I/sec
kg
1000 liters =
(MGD)(43.81)
(Ib)(0.454)
(1000 gal)(3.785)
-------
TABLE 4
CHARACTERISTICS OF MODEL REFINERY
FOR COST-EFFECTIVENESS ANALYSIS
Subcategory B, Cracking
Crude throughput: 42,500 BPSD
Refinery characteristics:
Crude desalting
Atmospheric distillation
Vacuum distillation
Hydrocracking
Fluid catalytic cracking
Coking
Hydrotreating
Catalytic reforming
Asphalt
EPA Configuration: 5.91
METRIC CONVERSION
(0.0159)(BPSD)
3
m /day
42,500
42,500
16,000
9,000
12,000
2,000
8,000
10,000
1,000
BPSD
BPSD
BPSD
BPSD
BPSD
BPSD
BPSD
BPSD
BPSD
(Ref. 3).
TABLE 5
MODEL REFINERY WASTEWATER CHARACTERISTICS
AT VARYING LEVELS OF TREATMENT
Annual Average
Parameter'
Flow, MGD
BOD5, mg/1
COD, mg/1
TSS, mg/1
(Ref. 3)
After
Air
Flotation
1
150
390
30
After After
Biological Tertiary
Treatment Filtration
1
25
125
40
METRIC CONVIISION
499
1
15
97
12
I/sec
After
Tertiary Carbon
Adsorption
1
5
40
6
= (MGD) (43. 81)
-------
TABLE 6
INCREMENTAL MASS
{ ,„„ 1,111,1
Parameter
BOD,., Ib/day
COD, Ib/day
TSS, Ib/day
OF POLLUTANT REMOVED
11 11 i — •• ii —
Biological
Treatment
1,042
2,210
-
BY TREATMENT
^^^^^^^•^^^^^MP^MMto^^^MM^^^MHI
Tertiary
Filtration
83
233
233
PROCESS
1 "- - i • •• i". I, .
Tertiary
Carbon Adsorption
83
475
50
METRIC CONVERSION
(Ib)(0.454)
TABLE 7
INCREMENTAL COSTS OF WASTEWATER TREATMENT FOR MODEL REFINERY
Treatment Type
Annual Annual
Capital Operation and Energy
Cost Maintenance Cost Cost
Activated Sludge
Tertiary Granular-media
Filtration
Tertiary Activated Carbon
$ 461,000
$ 32,000
$ 9,000
$ 220,000
$ 740,000
$ 6,000
$ 92,000
$ 2,000
$ 52,000
Basis: 1977 Gulf Coast Costs.
500
-------
TABLE 8
INCREMENTAL COST-EFFECTIVENESS OF WASTEWATER TREATMENT
FOR MODEL REFINERY
Cost-effectiveness
Parameter
Activated
Sludge
Treatment Method
Tertiary
Filtration
Tertiary
Carbon Adsorption
Columns
Flow, $/1000 gal
BOD5
COD,
TSS,
Capital
Annual O&M
, $/lb removed
Capital
Annual O&M
$/lb removed
Capital
Annual O&M
$/lb removed
Capital
Annual O&M
0.16
0.09
0.16
0.08
0.07
0.04
-
-
0.08
0.02
0.95
0.20
0.34
0.07
0.34
0.07
0.27
0.25
3.20
3.04
0.56
0.53
5.31
5.04
Basis: 1977 Gulf Coast, Capital amortized at 10 percent for 15 years.
METRIC CONVERSIONS
$71000 liters - ($/1000 gal)(0.26)
$/kg = ($/lb)(2.2)
501
-------
TABLE 9
COSTS FOR REMOVAL OF SELECTED WASTE CONSTITUENTS
FROM MODEL REFINERY EFFLUENT
Waste
Constituent
Cadmium*
Cyanides**
Chromium***
Amortized
Capital
Cost
($71000 gal)
0.13
0.22
0.23
Annual
O&M
Costs
($/1000 gal)
0.27
2.02
0.43
Total
Annual
Costs
($/1000 gal) ($/lb removed)
0.40
2.24 2.68
0.66 4.06
* Cost based on treating entire refinery process effluent.
** Assumes only FCC and coker process wastes are treated.
*** Assumes treatment of cooling tower blowdown only; 190,000 gpd.
METRIC CONVERSIONS
1000 liters = (1000 gal)(3.785)
kg = (Ib)(0.454)
502
-------
THE ECONOMICS OF HANDLING REFINERY SLUDGES
Carl E. Adams, Jr., President
and
John H. Koon, Director, Wastewater Management
Associated Water and Air Resources Engineers, Inc.
Nashville, Tennessee
The various types of sludges generated in refinery operations may require handling using
alternative thickening, stabilization and dewatering processes prior to final disposal. Some or all
of these processes may be required depending on the exact nature of the sludge prior to final
disposal. Final disposal practices might include landfilling, lagoohing, land farming, or incineration
of the sludge prior to disposal of ash using one of the above methods. The predominance of each
of these disposal methods as it existed in 1973 in an estimate for 1983 conditions is presented
in Table 1. These data indicate that landfilling and lagooning are the predominant methods in use
at the present time, while it is anticipated that the predominance of lagooning as a disposal
method will decrease significantly in the next several years and will be accompanied by a signifi-
cant increase in the popularity of land farming. Incineration is not expected to be in widespread
use in the industry in the foreseeable future.
Due to the large number of alternative processes for handling and disposal sludges, the
cost associated with the construction and operation of these processes plays a large role in the
selection of optimum sludge handling systems. This paper presents a discussion of the types and
characteristics of sludges originating from refinery operations, alternative techniques for handling
these sludges, and the costs associated with various process sequences.
SOURCES AND CHARACTERISTICS OF REFINERY SLUDGES
Normal refinery operations generate several major types of sludges which can be classi-
fied as follows:
1. Storage tank bottoms.
2. API separator bottoms.
3. Crude desalting sludge.
4. Catalytic solids.
5. Spent clays and coking fines.
6. Solids from utilities operations and biological wastewater treatment systems.
The basic source of many of the solids is the crude oils which contain materials that are
present in the oil as taken from the well and which separate during transportation and storage of
the crude oil. The solids usually associated with the crude oil include iron rusts, iron sulfides, clay,
sand, salt crystals, wax, and paraffin. These solids will generally settle out either in the storage
tank bottoms or API separators. Solids from biological treatment of wastewaters are generated
when soluble and colloidal biodegradable organics are converted into a biological mass which is
separated from the treated wastewater by gravity and concentrated for disposal. The waste
sludge from API separators plus typical secondary treatment facilities will contain approximately
503
-------
1,500 to 2,000 Ib of dry solids per million gallons of effluent treated. A delineation of types and
characteristics of refinery solids are given in Tables 2 and 3.
Those sludges which pose the major handling and disposal problems to a refinery are from
the utilities and wastewater treatment systems. 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
relatively inert and have no heat content, they should be dewatered and landfilled directly. Sludges
gnerated from the wastewater treatment facility 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 organic sludges.
The sludges from a refinery which are readily combustible include the waxy bottoms, oil
sludges, coke fines, and waxy tailings. The non-combustible sludges are sand, rust, silt, tetraethyl
lead sludge, salt, spent catalysts, and lime sludge. The excess biological sludge, although not readily
combustible, can be dewatered to an autogenous state.
SLUDGE HANDLING METHODOLOGY
In order to implement a successful sludge management program in a refinery, both in-plant
recovery and reclamation of applicable solids combined with final sludge handling and ultimate
disposal are required. In-plant management consists of tight operational control, preventive main-
tenance 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 filtration, pressure filtration, pressure
belt filtration, or centrifugation, and final disposal by incinceration, landfill, land farming and
barging to sea. Alternative sludge handling methods for oily and biological sludges are shown in
Figures 1 and 2, respectively.
There are basically three types of sludges which must be handled by refinery wastewater
treatment and sludge handling systems:
1. Recoverable oils.
2. Oily sludges.
3. Biological sludges.
Solids concentrations attainable using various dewatering processes is included in Figure 3.
An in-depth discussion of alternative handling methods has been presented elsewhere (2) and, al-
though many refineries handle the oily sludges in combination with the biological sludges, the
handling will be presented separately for clarification in this paper. Detailed design procedures for
these processes are presented in Reference 4.
Recoverable Oils
Generally, recoverable oils are separated from water mixtures by heating with steam to the
range of 150° to 180° F. Chemicals are added and the emulsion is broken into three phases: oil,
water, and sludge. The oils are usually recovered, the water sent to the wastewater treatment
system, and the sludges discharged into the oily sludge handling system. Separator skimmings,
which are generally referred to as slop oils, require treatment before they can be reused, due to the
high content of solids and water. Solids and water contents in excess of one percent generally inter-
fere with processing. These slop oils are easily treated by heating to 190°F, retained at this tempera-
ture for 4 to 6 hours, then settled for 12 to 24 hours. At the end of the settling period, three layers
exist: a top layer of clean oil, a middle layer of secondary emulsions, and a bottom layer of water
containing solubie components, suspended solids and oils. Frequently, it is advantageous or neces-
to add acid or a specific chemical to destabilize slop oil emulsions. The water layer resulting
504
-------
from this treatment contains higher levels of BOD and COD and must be treated through the
biological system prior to discharge. Slop oils are also successfully treated by precoat filtration(2).
Oily Sludges
Oily sludges are derived from oil-water separators, tank bottoms or cleanings, air flotation
treatment of wastewaters, and cleaning or dredging from lagoons or oxidation ponds. A major
consideration in treating oily sludges is that these materials are generated from various sources
and are discharged at irregular intervals. Thus, the sludge composition is highly variable from day
to day. Often tank bottoms contain heavy emulsions which require special treatment, and the
sludge and skimmings from air flotation systems can present special difficulties in handling also.
In many cases, the skimmings from air flotation are sent to a skimmed oil tank from which they
may be discharged to landfill or pumped to an oil recovery unit. Chemical treatment is generally
necessary for oil recovery.
The most common processes for handling oily sludges are gravity thickening, vacuum fil-
tration or centrifugation dewatering, and disposal by landfill or land farming. A summary of the
sludge handling methods used for oily sludges is given in Table 4.
Thickening. Gravity thickening methods are generally utilized for oily sludges. Solids
loading rates in the range of 5 to 30 Ib/sq ft-day are reported with thickened solids concentra-
tions from 3 to 10 percent. Basically, solids and oils recovery are low from gravity thickeners.
Due to the presence of heavy particles in oily sludges, flotation thickening has not been very suc-
cessful for these sludges. Dissolved air flotation has been successful for thickening froth flota-
tion sludges, but other refinery oily sludges are not deemed practical with this method.
Dewatering. Dewatering alternatives for oily sludges include centrifugation, vacuum fil-
tration, and pressure filtration. Figure 4 presents a schematic of a system being utilized for a num-
ber of oily sludges. This two-stage system utilizes a first-stage horizontal solid bowl centrifuge to
separate oil from the centrate. In this process the sludge is heated to 180° to 200° F prior to centri-
fugation. 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 tank bottoms.
3. A horizontal solid bowl 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
cake will consist of 1 to 5 percent oil and approximately 50 to 60 percent solids.
4. A high-speed nozzle centrifuge separates 95 to 98 percent of the feed oil in the oily
phase and 2 to 5 percent in the nozzle water. Thirty to 50 percent of the feed solids
will be removed with the oily phase with the remainder being in two water phases.
A significant number of refineries use vacuum filtration for dewatering oily sludges; and
if properly implemented, vacuum filtration renders the solids suitable for direct landfill or incinera-
tion. 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. The major conclusions regarding
vacuum filtration of oily sludges are:
1. Increased feed temperature greatly improves vacuum solids performance.
2. Addition of spent clay decreases oil recovery and solids filtration rate.
3. Measured filtration loadings of 0.8 to 3.0 Ib/sq ft-hr are required for oil recovery.
505
-------
A fixed plate, high pressure filter press may be used to dewater many types of oily sludges.
Effective breaking of solids-stabilized emulsions is obtained with cake solids in excess of 50 percent
of oil concentrations in the range of 5 to 20 percent being observed. Increased filtration tempera-
ture has a dramatic effect on cake solids and cake oil concentrations at fixed cycle times. Re-
finery experience indicates that a definite increase in filtration efficiency is obtained with lime
additions. A significant decrease in cake oil content can be obtained by washing the cake with
hot water.
Test results indicate that the recovered 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,
A new method of sludge dewatering in the United States employs gravity draining of sludges
coupled with pressure filtration applied by mechanical means of rollers and belts. These belt
pressure filters are capable of achieving solids content in a range of 12 to 30 percent using poly-
electrolyte additions of approximately 5 to 2-0 Ib of polyelectrolyte per ton of dry solids.
Ultimate 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 is viewed as a
short-term alternative and is eliminated as an option. The disposal of oily sludges on soil is accep-
table if it can be shown that such disposal will not contaminate groundwater or contaminate 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 decompositon of the oil-containing solids. Land farming involved
spreading the sludge in 4 to 6-in. layers, allowing the sludge to dry about one week, adding nutri-
ents, and then discing the sludge into the soil. Decomposition rates have found to average approxi-
mately 0.5 Ib/mo-cu ft wihtout nutrient addition and 1.0 Ib/mo-cu ft with nutrient addition.
Incineration of oily sludges with landfilling of the ash may provide an acceptable means of
final disposal in land limited situations. The three most common types of incinerators include
fluidized bed, rotary kiln, and multiple-hearth furnaces. The fluidized bed incinerator is best suited
for feeds that are partially liquid so that the incinerator can be fed by pumps and screw conveyors.
The multiple-hearth incinerator will be more economical if most of the feed is in the form of cake
or non-pumpable solids.
Biological Sludges
The sludge handling methodology used for biological sludges is similar to that described
for oily sludges. However, due to the biodegradable potential of these sludges, stabilization is
required prior to disposal by sea or land. The basic process sequence for biological sludge handling
consists of stabilization, thickening, dewatering, and final disposal. Often the thickening step will
precede stabilization in order to reduce the stabilization cost which is highly dependent on flow.
A summary of the sludge handling processes and design basis is shown in Table 5.
Biological Sludge Stabilization. The most common method of stabilizing waste biological
sludge in the refining industry is by aerobic digestion. Aerobic digestion is employed to stabilize
the sludge and render it suitable for land disposal, although the overall sludge quantity may be
somewhat reduced and the dewatering characteristics slightly improved. Approximately 10 to 20
days detention time 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 considered suitable for land application.
506
-------
Thickening of Biological Sludges. The three most common processes used to thicken bio-
logical sludges in the refining industry are flotation, gravity thickening, and centrifugation The
belt pressure filter is receiving increasing attention as a thickener and dewatering device for refinerv
sludges.
Results of tests using refinery sludges indicate that these sludges are quite suitable for flo-
tation thickening. Data indicate that the optimum pressure is approximately 50 to 55 psig and no
polymer addition is justified. The optimum air-to-solids ratio appears to be about 0.01 Ib air/lb
solids and that float solids on the order of 3 to 5 percent can be readily achieved.
With gravity thickening (the most common thickening process in refinery practice), loading
rates in the range of 2 to 15 Ib per sq/ft per day are employed with solids concentration being
achieved in the range 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 when
solids loading do not vary substantially. In extremely warm climates, gravity thickeners may
generate obnoxious biological odors if they precede the stabilization process.
Centrifugation has been utilized sparingly for thickening waste biological sludges in the re-
fining industry. Basket centrifuges are capable of thickening waste activated sludge to levels of
5 to 6 percent concentration with an 80 to 95 percent solids capture. One refinery has reported
using basket centrifuges for thickening to a concentration of 8 percent. However, excessive main-
tenance was experienced because of vibrational problems.
Dewatering of Biological Sludges. The most common methods of dewatering biological
sludges are vacuum filtration, centrifugation, and pressure filtration with the belt filtration press
achieving an increasing use in the refining industry.
Vacuum filtration is the most common method used to dewater refinery wastewater sludges.
Chemicals are usually required and 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 combination from an economical and performance standpoint. The optimum ferric
chloride dosage usually ranges from 200 to 400 Ib/ton of dry solids. Tests and full scale experi-
ences indicate that vacuum filtration of refinery biological sludges will usually achieve solids
concentration in the range of 10 to 16 percent at filter loading rates ranging from 1 to 5 Ib/sq ft-hr.
Digested sludge will usually dewater slightly better than raw, undigested sludge.
Basket centrifuges have been found to provide the best centrifugal method of concentrating
waste biological solids, and there are a number of applications utilizing basket centrifuges in the
refining industry. Normally, centrifuges provide a cake solids ranging from 8 to 18 percent con-
centration. There have been some problems in the utilization of the basket centrifuge due to
mechanical vibrations; however, these problems are being solved by constructing the baskets so
that it is driven from the top rather than the bottom.
Pressure filtration will usually achieve a solids concentration up to 50 percent with bio-
logical sludges, thus producing the driest cake for disposal. The pressure levels for the pressure fil-
tration system range from 50 to 225 psig. Normally, it has been found that the pressure filtration
system required higher chemical dosages than the vacuum filtration or centrifugation systems.
The use of various pressure belt filters for dewatering of waste biological solids and some
oily sludges is becoming increasingly widespread in refining and other industries. A summary of
data which has recently been obtained using these devices is shown in Table 6. Generally, pressure
belt filters are capable of dewatering refinery waste activated sludges to keep concentrations be-
tween 15 to 20 percent. More limited data processing a combination of waste activated and oily
sludges indicates cake solids contents of approximately 27 percent are obtainable. The wide range
of polymer dosages used in these tests are indicative of variation both in the types of polymers and
sludges processed. Loading rates to the belt presses range from 1 to 5 Ib/inch belt width for bio-
logical and oily sludges. Figure 5 presents capital costs estimates for pressure belt filters as a func-
tion of the quantity of biological sludge. It is important to note from data in Figure 5 that increas-
ing the solids content applied to the belt filter by a factor of 2 will reduce the capital cost almost
in half. Consequently, it appears that thickening ahead of the pressure belt filters is definitely justi-
fied.
507
-------
Ultimate Disposal. The alternative for ultimate disposal of biological sludges are similar
to those for a waste oily sludge. In many cases, a thick sludge or activated sludge can be placed
in 4 to 8 inch layers and utilized as a limited nutrient and soil conditioner. In this case, additional
waste is removed by natural evaporation and infiltration, into the underlying soil. Following initial
periods of drying, the sludge layer can be disced to encpurage the activity of aerobic soil bacteria.
ECONOMIC CONSIDERATIONS
It is extremely difficult to collect representative and comparable cost data on sludge hand-
ling systems in the refining industry. Different combinations of oily and biological sludges, dif-
ferent labor costs, and construction and capital costs which might be a function of total equip-
ment purchased from a manufacturer (such as a complete wastewater system in which the sludge
handling were only a fraction of the total equipment purchased), make it extremely difficult to
develop cost as a function of size of equipment. Therefore, some data are gathered which are con-
sidered to be reliable and presented in a form for discussion in this section.
In Figure 6, a sludge handling system is shown whereby biological and oil sludges were
combined for dewatering by pressure filtraton or vacuum filtration prior to landfill (5). The bio-
logical solids (1,070 tn/yr) were aerobically digested and gravity thickened. The oily sludges (1,000
tn/yr) were thickened and stored. The combined biological and oily sludges (2,070 tn/yr) were de-
watered by pressure (Alt. I) or vacuum filtration (Alt. II) and transported to landfill. The third
alternative consisted of direct transport by piping of the combined thickened biological and oily
sludges for land farming. x , ,
Costs for Alternative I (dewatering using pressure; filtration followed by landffll dis-
posal of the sludge cake) were based on operation of the pressure filter 140 hr/wk and transport
of the dewatered cake approximately two miles to the landfill site. Costs were included for the
use of 1 Ib conditioner/lb sludge dewatered and it was assumed that a cake having a solids content
somewhat greater than 50 percent would be obtained. The filters had facilities included for pre-
coating and were designed for a two-hour cycle time. The total sludge volume transported to the
landfill site was 18 cu yd/day.
Costs for Alternative II employing vacuum filtration for sludge dewatering included
operation of the vacuum filter 135 hr/wk. The filter was designed using a loading rate of 1.4 Ib/sq
ft-hr and it was assumed that a 40 percent cake would be produced. The vacuum filter was also
designed as a precoat system and it was assumed that the amount of precoat required would equal
10 percent of the weight of sludge to be dewatered. Transport distance to the landfill was also
assumed to be 2 mi and the quantity of cake requiring disposal was calculated to be approximately
40 cu yd/day.
Costs for land farming of the sludge (Alternative III) were based on the application of sludge
in 4 to 6 inch layers at an application rate of 5 Ib/sq ft-yr. It was assumed that following applica-
tion the sludge would be allowed to dry for one week during which nutrients would be added to
the soil, and subsequently followed by discing of the sludge into the soil.
In this situation, dewatering by vacuum filtration was 15 percent cheaper than pressure
filtration; however, the transportation costs were almost twice as much due to the wetter vacuum
filter sludge. In summary, the vacuum filtration scheme was approximately 4 percent cheaper than
the pressure filtration scheme. However, Alternative III, land farming, was about 50 percent more
economical than either pressure or vacuum filtration. The land farming assumes discing with a
tractor and disc rather than the more expensive bulldozer used in some other published methods.
Pressure Filtration Costs
Estimated costs for pressure filtration dewatering of oily sludges is presented in Table 7.
These costs were based on dewatering API separator bottoms, tank bottoms, decant pit sludges,
flotation units solids, spent clays, oil emulsions, and other miscellaneous oily sludges (6). Each
press was designed using a 1.5-inch cake thickness, and 64-inch diameter plates. It was also assumed
that the feed sludges would be heated to a temperature of 150° to 200° F prior to the dewatering
period. It was assumed that lime would be used as a conditioning agent with a dose of 0.1 Ib lime/lb
dry solids dewatered.
508
-------
Dewatering costs presented in Table 7 range between $26 and $40/tn including both capital
and operating costs. Costs for pressure filtration shown in Figure 6 are substantially higher than
those shown in Table 7 where an in-depth cost analysis of pressure filtration was performed. The
higher costs associated with pressure filtration in Figure 6 are partially attributable to the presence
of biological solids which make the sludges more difficult to dewater and require larger quantities
of conditioning chemical. In addition, a substantially larger quantity of sludge was used as a basis
of the estimates presented in Table 7 - 9,200 tn/yr compared to 2,070 tn/yr used in Figure 6.
The analysis presented in Table 7 indicates that dewatering costs are minimized with three
shifts per day operations and become progressively higher with 2 and 1 shift per day operations.
However, the cost estimates for three shifts per day operation contain no flexibility in the design
since only one machine was used for 24 hr/day operation. It was estimated that the use of two
presses, each having the capacity to dewater 60 percent of the total solids throughout, would add
approximately $125,000/yr to the total annual cost. On a unit cost basis, this would increase the
total sludge dewatering costs to approximately $39/tn. In many instances, it might be more eco-
nomical to provide storage facilities for holding sludge when maintenance of the press is required.
In individual cases, the optimum design would require consideration not only of initial construc-
tion and operating costs but also of labor policies which might interfere with three shift per day
operation, the operational flexibility desired, the degree to which alternative designs could accom-
modate plant expansion in the future, etc.
Costs for Centrifugation of Oily Sludges
In Table 8, the cost of a centrifuge system for handling oily sludges is presented. The cost
to dewater 660 tn/yr of flotation unit skimmings using a basket centrifuge was estimated to cost
$160/tn (7). A two-stage system consisting of a horizontal solid bowl and disc-nozzle centrifuge
in series for handling combined oily sludges was estimated to cost only $45/tn for treating 1,750
tn/yr. Costs for the basket centrifuge treating flotation skimmings was estimated based on a 17 gpm
flow rate, 90 percent on-stream factor, and 2 percent influent solids concentration. Costs of the
basket centrifuge and disc-nozzle system in series was estimated based on a 7 percent influent solids
concentration at a flow of 13 gpm and a 90 percent on-stream factor. In using the horizontal
solid bowl and disc-nozzle centrifuge in series, the feed would initially be processed through the
basket centrifuge with the centrate and the emulsion contained in this stream being broken and
handled with the disc-nozzle centrifuge.
Incineration Costs
Costs for the incineration of refinery sludges using alternative dewatering methods is pre-
sented in Table 9. In each case, the sludge consisted of flotation unit solids, refinery oil sludges,
spent clays and emulsions (Whitewater). Each example was based on dewatering 8,400 tn/yr dry
solids. If only the flotation solids are dewatered using centrifugation and the remainder of the
streams are fed to the incinerator without dewatering, the total cost of incineration would amount
to approximately $53/tn. By dewatering both flotation solids and oily sludges using centrifugation,
costs were observed to decrease approximately $45/tn. Pressure filtration of all wastestreams prior
to incineration further reduced the total incineration costs to approximately $41 An.
Land Farming
For Allowable oil loadings ranging from 2.5 to 25 Ib/sq ft-yr, the required land area for
land farming of oily sludges can be determined from Figure 7. Literature costs for land farming
of oily sludges are shown in Table 10. Although the total costs per ton of waste are approximately
the same, $22 and $24/tn, there is a significant discrepancy in each case regarding the loading rate
of oil to the soil, cost of transportation, and cost of cultivation. The cost for land farming taken
509
-------
from Figure 6 totaled $35/tn of solids. Based on a reported 50 percent oil content for this sludge,
the cost would be equivalent to approximately $70 of oil applied to the land. Costs for this latter
example were developed based on an application rate of 5 Ib oil/cu ft-yr applied to a depth of ap-
proximately 6 in. which is equivalent to a surface loading of approximately 2.5 Ib/sq ft-yr. While
the unit costs for land farming estimated in each of these three examples are not extremely diver-
gent, very significant differences exist for cost estimates for individual items. It is felt that this
is partially the result of differences inherent to each individual situation. However, it is also likely
more accurate costs will be available as the practice of land farming continues to be used in the
refining industry.
REFERENCES
1. Jacobs Engineering Company. "Assessment of Hazardous Waste Practices in the Petroleum
Refining Industry." Environmental Protection Agency, Contract Number 69-01-2288,
Pasadena, California, June 1976.
2. Adams, Carl E., Jr., and Stein, Robert M., "Sludge Handling Methodology for Refinery
Sludges," Proceedings of the Open Forum on Management of Petroleum Refinery Waste-
waters. The University of Tulsa, 1976.
3. Cross, F. L. and Lawson, J. R., "A New Petroleum Refinery." American Institute of Chem-
cal Engineering Symposium Series, Vol. 70, No. 136, p. 812.
4. Adams, Carl E., Jr., and Eckenfelder, W. Wesley, Jr., Process Design Techniques for Indus-
trial Waste Treatment. Nashville, Tennessee: Enviro Press, 1974.
5. Ford, Davis L., "A Preliminary Engineering Study for Wastewater Treatment and Pollution
Abatement," September, 1971.
6. Ford, Davis L., "A Preliminary Engineering Study for Solid Waste Disposal and Pollution
Abatement," February, 1972.
7. Kincannon, C. B., "Oily Waste Disposal by Soil Cultivation," Proceedings of the Open
Forum on Management of Petroleum Refinery Wastewaters. The University of Tulsa,
1976.
8. Huddleston, R. L. and Cresswell, L. W., "The Disposal of Oily Wastes by Land Farming,"
Proceedings of the Open Forum on Management of Petroleum Refinery Wastewaters,
The University of Tulsa, 1976.
510
-------
BIOGRAPHIES
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 industries. He holds the
following degrees: B.E., Civil Engineering,
Vanderbilt University; M.S., Environmental
Engineering, University of Texas. He is a pro-
fessional engineer, registered 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 professional societies,
has authored and presented more than 50 technical
papers, and edited and co-authored three books
dealing with industrial wastewater treatment
technology.
John H. Koon is Director of Wastewater
Management for Associated Water and Air Resources
(AWARE), Inc., Nashville, Tennessee. He holds a
B.E. degree in Civil Engineering and a M.S. degree
in Sanitary and Water Resources Engineering from
Vanderbilt University and a Ph.D. degree in
Sanitary Engineering from the University of
California at Berkeley. Dr. Koon is a registered
professional engineer with eight years of experience
in industrial and municipal wastewater management.
He is also the author of numerous technical papers
and reports.
511
-------
TABLE 1. REFINING INDUSTRY SLUDGE DISPOSAL PRACTICES3
DISTRIBUTION (%)
METHOD 1973 1983 (est.l
LANDFILL
LAGOON ING
LAND FARMING
INCINERATION
ONSITE DISPOSAL
OFFSITE DISPOSAL
50
40
9
1
44
56
44
19
34
3
73
27
aJacobs Engineering Company. "Assessment of Hazardous Waste Practices in the Petroleum
Refining Industry." Environmental Protection Agency, Contract Number 68-01-2288,
Pasadena, California, June 1976.
TABLE 2. CHARACTERISTICS OF REFINERY SOLIDS WASTES (Ref. 3)
Typical Composition, Percent
Waste Type
API Separator Sludge
Tank Bottoms
Chemical Treatment Sludge
Air Flotation Froth
Precoat Vacuum Filter
Sludges
Biological Treatment Sludges
Raw
Mechanically Thickened
Centrifuged
Vacuum Filtered
Screw Pressed
Water Treatment Sludge
Oil or
Hydrocarbon Water
15
66
48
5
22
22
40
90
75
29
4 8
5
3
49
0
0
0
0
0
0
98
94
85
75
40
95
Volatile Inert
Solids Solids Characteristics
6 13 Fluid slurry of oil,
water and sand
Oil-water mixture
Slightly viscous fluid
Thick, oily fluid
Temperatures
1.5 0.5 Water Consistency
4 2 Thick, but pumpable
10 5 Viscous-peanut butter
consistency
15 10 Wet crumbly solid
40 Intact, solid cake
5 Pumpable Fluid, some-
times gelatinous
Cross, F. L. and J. R. Lawson. "A New Petroleum Refinery." American Institute of Chemical
Engineering Symposium Series, Vol. 70, No. 136, P, 812.
512
-------
TABLE 3. ESTIMATES OF REFINERY SOLID WASTE QUANTITIES (Ref. 3)
Waste Types Unit Loads
API Separator Sludge 200 mg/l Suspended Solids
Chemical Treatment 50 mg/l Suspended Solids
(API Separator Effluent) Removed Only
Biological Sludges ' 0.7 Ib Dry Solids per Ib
BOD Removed
Water Treatment Sludge
A. Lime Soda Ash 2 parts Dry Sludge per
1 part Hardness Removed
B. Ion Exchange 0.4 Ib Salt per 1,000 Grains
Hardness
Office Wastes 1 -0 cu yd per Employee
per Month
Cafeteria 0.6 Ib per Meal
Cross, F. L. and J. R. Lawson, "A New Petroleum Refinery." American Institute of
Chemical Engineering Symposium Series, Vol. 70, No. 136, p. 812
513
-------
TABLE 4. SUMMARY FOR THE THICKENING AND DEWATERING ALTERNATIVE OF OILY SLUDGES
Type of
Sludg
Oily Sludge
P'rocess
Thickening
Gravity
Flotation
Dewatering
Centrifu-
gation
Vacuum
Filtration
en
Pressure
Average Performance
3-7 percent solids, 50% recovery
84-93% solids recovery, 3.5 - 5.5%
float solids
75-90% solids recovery when charged
with heated oily sludges. Oily phase
will consist of 1-5% solids and cake will
contain 50-60% solids. Nozzle ejector
centrifuge separates 95-98% of oil in
oily phase and 205% in the nozzle and
nozzle water. 30-50% of nozzle ejector
feed solids will be removed with oily phase.
5-20% oil, 40-70% solids cake
Design Parameters
|