United States
Environmental Protection
Agency
Robert S. Kerr Environmental
Research Laboratory
Ada OK 74820
Research and Development
EPA-600/S2-84-176 Jan. 1985
&ERA Project Summary
Demonstration of a Maximum
Recycle, Sidestream Softening
System at a Petrochemical
Plant and a Petroleum Refinery
Jack V. Matson, Wendy Gardiner Mouche, Eric Rosenblum, and
Larry McGaughey
New full-scale maximum recycle
sidestream softening systems at USS
Chemicals, Houston, Texas, and TOSCO
Refinery, Bakersfield. California, were
evaluated as a technology to achieve
zero wastewater discharge. Softener
process efficiency was optimum at a pH
control range of 10.3 to 10.5 at 40°C
and using a high mixing intensity. A
problem of heat exchanger biofouling
from the high dissolved organics in
recycle water was effectively controlled
by using Bromocide with chlorine. A
total organic carbon balance over the
cooling water system showed raw
makeup water and process water con-
tribute 1 /3 and 2/3 of the organics,
respectively. Major organic sinks were
drift (60%), biodegradation (30%), and
volatilization (10%). Softener sludge as
analyzed for chromium by leachate tests
was classified as nontoxic. Heat ex-
changer equipment averaged two
mils/year internal corrosion. External
corrosion from drift aerosols was cor-
rected by installation of a ferrous sulfate
reactor in the blow down system and
improved drift eliminators in cooling
towers. The TOSCO water problem of
high silica and low magnesium was
corrected by adding caustic and mag-
nesium sulfate to the softener. Both
plants operated satisfactorily at near
zero liquid discharge. Operating costs
and benefits are discussed.
This Project Summary was developed
by EPA's Robert S. Kerr Environmental
Research Laboratory, Ada. OK, to an-
nounce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
The USS Chemicals plant in Houston
produces ethylene from a refinery gas
rich in ethane by a steam cracking
process, and manufactures styrene from
ethyl-benzene by a catalytic process. The
plant is medium sized, with 200 em-
ployees.
Prior to 1979, the plant discharged
roughly 300 to 500 gallons per minute of
effluent to the Houston Ship Channel.
The wastewaters consisted primarily of
oily condensates from the ethylene unit
that were collected and treated in a
coagulation/flocculation unit and added
to the cooling water systems. The blow-
down from these systems with plant
runoff and demineralizer spent regenera-
tion fluid constituted the effluent.
The regulations governing the quality
of the effluent were made more stringent
in a stepwise fashion starting in 1974. At
that time, USS Chemicals initiated a water
management study to determine the need
for wastewater treatment facilities. Var-
ious schemes for upgrading the existing
facilities such as biological treatment,
dirty boiler, and activated carbon were
compared to the sidestream softening
recycle system. The study showed that
the waste from other systems was not
very biodegradable or adsorbable. On the
-------
basis of economics, the recycle system
was chosen. The savings in water
makeup, which was from a surface water
delivery system, coupled with the reality
that there were no feasible alternative
treatment processes, resulted in a deci-
sion to choose the current system.
The strategy for successful implemen-
tation would be quite simple. The chemi-
cal scale from constituents—calcium,
magnesium, and silica—would be re-
moved in the softener. The other ions,
such as sodium, chloride, and sulfate,
would be concentrated and controlled by
drift, the aerosol droplets escaping with
the heated air from the cooling tower. The
potential problems were high total dis-
solved solids (TDS), which increased the
corrosivity of the cooling water; concen-
tration of non-volatile organics, which
increased the biofouling potential; and
external corrosion from the deposition of
the drift aerosols.
The potential advantages were elimi-
nation of blowdown, which was the
largest quantitative source of effluent
requiring treatment prior to discharge;
savings in makeup water; and a way to
continue to use chromate corrosion in-
hibitors. Hexavalent chrome was found to
be toxic at low concentrations so that
regulations called for nearly complete
removal. At that time, there were no
equivalent substitutes for chromate as a
corrosion inhibitor.
Pilot testwork of the proposed system
was conducted in 1974-75. Construction
began in 1977 and coincided with the
conversion from ground water to surface
water supply. The entire facility, including
surface water treatment units, had a
capital cost of roughly $3 million of which
half the cost was for the sidestream
softening system.
This system was unique in a number of
ways. First, carbon dioxide instead of
sulf uric acid was used to control pH in the
cooling water system. The reasoning was
that carbon dioxide preserved the bicar-
bonate in the raw water, and thus greatly
reduced the need for soda ash m the
sidestream softeners. Sulfuric acid would
have destroyed the bicarbonate by con-
verting it to carbon dioxide that was
strippable in the cooling towers. Another
important reason for carbon dioxide usage
was that the pH in the cooling water could
only go as low as approximately four in
the event of operator error or equipment
malfunction, while sulf uric acid overdose
could bring the pH significantly lower and
damage the heat exchangers, which were
of carbon steel with some stainless
material.
All process water was collected, treated
(if necessary), and used as makeup to the
cooling water system. This included oily
condensates which were treated in an
API separator and a coagulation/floc-
culation process, boiler blowdown, and
plant runoff including storm water. The
only discharge was the high TDS de-
mineralizer-regeneration water. Prior to
this project, the effluent discharge was in
the range of 300 to 500 GPM; afterwards,
it was 30 to 50 GPM, a ten-fold reduction.
Study Objective
The objectives of the study were as
follows:
1. evaluate the performance of the
sidestream softening system;
2. determine the impact of the recycle
streams on the cooling water
system;
3. discover the fate of the organic
matter in the cooling water system;
and
4. investigate the potential sludge'
disposal problems.
Findings
/. Performance Evaluation
Prior to startup, there were two impor-
tant questions. First, was there sufficient
magnesium in the raw water to adsorb
the silica in the softening reactions? The
silica-to-magnesium ratio in the raw
water was two to one (in mg/L), which
was marginal. In the reaction, magnesium
is precipitated as a hydroxide onto which
the silica is adsorbed. The question was
answered in the affirmative. As part of
the project, a detailed scientific delinea-
tion of the adsorption phenomena was
made.
The second question involved the
selection of the appropriate scale inhibitor
for the cooling water. An inhibitor was
needed that worked well in the cooling
water but did not significantly interfere
with the softening reactions. Generic
phosphonate inhibitors were found in
laboratory tests to work too well in that
the efficiency of the softening reactions
was decreased. Selected were polymaleic
anhydride polymers which mechanisti-
cally prevented scale formation by fluidiz-
ing small crystals rather than inhibiting
crystal formation. That is what actually
happened. The inhibitor helped maintain
a scale-free heat transfer system.
The most critical element of the side-
stream softening system was the soften-
ers. Operational costs were related to the
consumption of soda ash and lime in the
softening process as was process effi-
ciency. So it was important to develop the
optimum control strategy for the system.
The lime dosage controlled the pH of
the reaction in the softener. Soda ash
was used to make up for any deficiency in
carbonate in the reactor. With the use of
carbon dioxide in the cooling water
system, the bicarbonate in the makeup
water was conserved so that the soda ash
requirement was minimized. The opera-
tors were instructed to control pH within
a narrow band by lime addition, and to
add soda ash based on the difference
5 between the calcium hardness and
carbonate alkalinity run every four hours.
The objective of this part of the per-
formance evaluation was to define the
most cost-effective pH in the softeners
with quality constraints in the cooling
water. The interactions were as follows.
A higher pH required a greater lime
dosage and increased the soda ash usage
as more carbonate was necessary to
react with the calcium ions in the lime.
However, the higher pH improved reaction
efficiency by conversion of a greater
percentage of bicarbonate to carbonate
and precipitation of magnesium. Thus, for
a greater cost, the benefit was a higher
reaction efficiency, which allowed for a
decreased flow requirement through the
softeners. The optimization studies per-
formed in the laboratory then in the plant
indicated that the control pH range was
10.3 to 10.5 at40°C.
The next performance problem analyzed
was the effect of mixing intensity on the
softening reactions. The softeners were
manufactured by Infilco Degrement under
the Densator brand name. Of particular
interest was the mixing in the primary
reaction zone.
Originally, a blade arm rotating on the
same shaft as the bottom scraper provided
slow speed mixing. It appeared to be
inadequate; so two high-speed turbine
mixers replaced the blade in one softener
while the other softener retained the
blade for comparison purposes.
The results of the testing indicated that
mixing intensity was, indeed, an impor-
tant variable. At low mixing intensities,
the reaction/adsorption process involving
magnesium precipitating as the oxide and
the subsequent adsorption of silica was
adversely affected. Also the lime dosage
was much greater because its dissolution
was strongly influenced by mixing. At
very high mixing intensities, the clarifica-
tion process was hindered by the breakup
of the particles. Thus, an optimum mixing
-------
intensity with the turbine mixers was
defined.
Two other observations are worth
noting: with the slow-speed blade it was
difficult to control the addition of lime and
soda ash to the softeners because the
reaction zone was not well mixed, which
created sampling errors, inaccurate test
results, and high feedback control noise.
Also, silica adsorption in the reactor was
much higher than previously predicted in
laboratory studies.
2. Impact of Recycled Water
The recycle water that had a strong
negative impact on the cooling water
system was the process wastewater
which was high in dissolved organics.
After the startup and shakedown of the
sidestream softening system, the major
unresolved problem was biofouling in the
heat exchange system. It was particularly
bad on the critical surface condensers.
The problem was exacerbated by the
operation conditions in the cooling water
system. As previously mentioned, carbon
dioxide was used to maintain pH control.
To minimize the use and expense of the
gas, which was readily stripped out in the
cooling tower, the pH was maintained
between 7.6 and 7.8. Chlorine was used
as the primary biofouling control agent.
At the higher pH, chlorine was relatively
ineffective, as it was largely in the
hypochlorite ion state. Thus, a research
effort was undertaken to resolve the
problem. An organic compound, dichloro-
dibromo dydantoin (trade name Bromo-
cide) was added in conjunction with
chlorine. The combination of these chem-
icals proved to be very effective in
controlling biofilm development at a
reasonable cost. The key was bromine
which, when released in solution as
hypobromous acid, was an effective bio-
cide at the higher pH.
3. The Fate of Organics
A total organic carbon balance was
made over the cooling water system to
determine the sources and sinks of the
organic material. Samples were analyzed
for biodegradability. The results indicated
that the raw makeup water and the
process water contribute one-third and
two-thirds of the organic materials,
respectively. The major sinks were drift
(60%), biodegradation (30%), and volatil-
ization (10%).
4. Sludge Disposal
The most serious potential problem
with the sludge from the softeners was
the classification, i.e., whether it was
considered toxic and had to be disposed of
in a Class 1 site at high cost. The only
constituent of concern was chromium
used in the cooling water system as a
corrosion inhibitor. A fraction of it de-
graded from a hexavalent form to a
trivalent form that precipitated as a
hydroxide in the softeners.
A series of leachate tests were per-
formed using both the EPA Standard and
State of Texas procedures. Under a variety
of conditions at the plant and in the lab,
the sludge met both the EPA and State
criteria. The sludge was consequently
classified as non-toxic.
Other Investigations
After startup, the cooling water system
equilibrated at a total dissolved solids
(TDS) level between 20 and 30,000 mg/L.
This figure was much higher than origi-
nally anticipated. Internal corrosion in the
heat exchange equipment was never a
problem, averaging below two mils per
year during the course of the study.
However, external corrosion caused by
the deposition of drift aerosols by sur-
rounding equipment was a real concern.
As part of the study, an emergency
blowdown system consisting of a reactor
in which ferrous sulfate was added to
precipitate the chromate and a clarifier
was tested in the laboratory. As a result of
the tests, a full-scale treatment system
was placed into operation. Also, the major
cooling tower was rebuilt with improved
drift eliminators to minimize the deposi-
tion.
Summary—USS Chemicals
The sidestream softening concept was
particularly attractive to USS Chemicals
because the cost of their water was
relatively high and their effluent dis-
charge regulations were tough. The
system is continuing to operate in a
steady-state mode, achieving the original
expectation with a minimum discharge.
TOSCO
At TOSCO, different reasons existed for
considering zero discharge, sidestream
softening. In the past the effluent had
been discharged into a percolation field
from which it migrated into an under-
ground aquifer. As regulations stiffened,
the discharges were shifted to an injection
well system at the refinery. The injection
well was costly to operate, and was taxed
by the State as a hazardous waste.
Since the cooling water blowdown was
a major volumetric constituent in the
injection well, it made economic sense to
look for alternative ways to deal with it. In
the mid-1970's several studies favorably
evaluated sidestream softening for the
blowdown. In 1980, the decision to
proceed with that design was made.
The unique problem with the TOSCO
water was its high silica and low mag-
nesium content. Insufficient magnesium
was present in the makeup water to have
much adsorbing power for the silica. An
evaluation of alternative ways to over-
come the problem indicated that the best
way was a combination of caustic as the
source of hydroxide in the softener and
magnesium sulfate as the source of
soluble magnesium for silica reduction.
Laboratory studies had shown that
magnesium in a soluble form had roughly
an order of magnitude greater adsorbing
power for silica than solid forms such as
magnesium oxide and dolomitic lime.
Caustic was attractive as a way to mini-
mize sludge production and soda ash
usage.
The system consisted of dual Densator
softeners, sulfuric acid pH control tank,
and dual media filters. The sludge was
concentrated and pumped to a series of
lined evaporation ponds.
The refinery consisted of five cooling
water systems with a combined recircula-
tion rate of 53,000 gpm, tied together.
Design flow through the sidestream
softener was rated for 250 gpm, of which
150 was from the cooling water system,
and the remainder from the boiler blow-
down and scrubber system.
Operation and Economics
Startup commenced smoothly in March
1982, with the system running at steady
state within one week. A dense sludge
was quickly established in the softeners.
Silica was controlled to below 150 mg/L
in the cooling water system by the
measured addition of dissolved magne-
sium. The total dissolved solids level
fluctuated between 5,000 and 8,000
mg/L.
Reuse of cooling water saved TOSCO
roughly $36,000 in actual water costs.
Another $300,000 was saved in supply
well maintenance, power, and taxes.
Water treatment chemical savings of
$50,000 were also realized. Overall,
savings for the refinery were estimated at
$500,000 per year. The cost of the side-
stream softening system was $2.5 mil-
lion. The capital payback period is five
years.
-------
J. V. Matson, W. G. Mouche, E. Rosenblum, and L McGaughey are with the
University of Houston, Houston, TX 77004.
Donald Kampbell is the EPA Project Officer (see below).
The complete report, entitled "Demonstration of a Maximum Recycle, Sidestream
Softening System at a Petrochemical Plant and a Petroleum Refinery," (Order
No. PB85-121 044; Cost: $19.00, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
P.O. Box 1198
Ada. OK 74820
* U.S GOVERNMENT PRINTING OFFICE; 1985 — 559-016/7885
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
".'•: . '" PDSTA4l«rJpEE&«AiO
' • - ' ' • '.i-tfcut '- ! !, i I
PERMIT No. G-35 *
Official Business
Penalty for Private Use $300
1L 60604
------- |