PROJECT SUMMARY
          HASTE MINIMIZATION  AUDIT REPORT
     Case Studies of Corrosive and Heavy Metal Waste
     Minimization at a Specialty Steel Manufacturing
     Complex
                   Submitted  by:

                    Versar  Inc.
                6850 Versar Center
              Springfield.  Va.  22151

                In Response to:
          EPA Contract No.  6F-0107053
             Work.  Assignment  No. 46
                Project Officer

                 Harry Freeman
           Thermal Destruction Branch
       Alternative Technologies Division
HAZARDOUS WASTE ENGINEERING RESEARCH LABORATORY
       OFFICE OF RESEARCH AND DEVELOPMENT
      U.S. ENVIRONMENTAL PROTECTION AGENCY
            CINCINNATI, OHIO  45268

-------
PROJECT SUMMARY

WASTE MINIMIZATION AUDITS AT GENERATORS OF CORROSIVE AND HEAVY METAL
WASTES

M. Drabkin and E. Rlssmann

    The USEPA 1s encouraging hazardous waste generators to develop
programs to reduce the generation of hazardous waste.   To encourage such
programs the Agency's Hazardous Waste Engineering Research Laboratory Is
supporting the development and evaluation of a model hazardous waste
minimization audit procedure.  The procedure was tested 1n several
facilities In the summer of 1986.

    Waste minimization audits (WMAs) have been carried out in an electric
arc furnace (EAF) specialty steelmaking complex.  These audits were
Intended to develop waste minimization options for two hazardous waste
streams at this facility:  corrosive waste and heavy metals waste.  Waste
minimization options considered were 1n one of three categories:  source
reduction, recycling or treatment On the same order of preference).

    Application of WMA methodology to a corrosive waste stream (K062)
generated at one plant in this complex, resulted in the development of a
promfsing recycling option for the recovery of calcium fluoride
(fluorspar) which is directly usable as a metallurgical flux (replacing
presently purchased material) In the EAF steelmaking process at this
facility.  Savings obtained by using this option (Including $68,000
savings from a thirty percent reduction 1n offsite nonhazardous waste
disposal, and $100,000 savings 1n purchased chemical costs) were
estimated at $168,000 annually, and the proposed process could largely
use existing process equipment.

    Application of the WMA methodology to a heavy metals-bearing waste
(EAF dust-listed waste K061) generated at another plant in the
steelmaking complex did not result in any viable source reduction or
recycling options for this waste.  However, a detoxification treatment
step proposed for this material is economically attractive based on
preliminary estimates, and bench-scale development of this option appears
warranted.

    This Project Summary was developed by EPA's Hazardous Waste
Engineering Research Laboratory, Cincinnati, OH, to announce 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 U.S. Environmental Protection Agency (EPA) 1s expanding its
efforts to promote waste minimization activity in the private sector by
providing technical assistance to generators of hazardous waste.  As part
of this effort, the EPA Office of Research and Development/Hazardous
Waste Environmental Research Laboratory (ORD/HWERL), Cincinnati, Ohio, 1s
promoting the development and testing of a generalized or model waste
minimization audit (WMA) procedure and test this procedure in actual
production facilities agreeing to cooperate with the audit teams selected
for this task.  Initially, four hazardous wastes were selected by
EPA/ORD/HWERL to be studied 1n this effort.  These wastes Include the
following:
    1.  Corrosives;
    2.  Heavy metals;
    3.  Waste solvents; and
    4.  Cyanides
    In  this report, results are presented of WMAs conducted at generators
of corrosive and heavy metals wastes.  A specialty steel manufacturing
complex agreed to provide host facilities for the WMA effort reported
herein.  This complex Includes among others, the following plants which
were audited in the present effort:
      An annealing and pickling facility for finishing stainless strip
       only.  This facility is designated as Plant No. 1.

-------
      Cold rolling, annealing and pickling facilities for finishing both
       stainless and electrical steel product strip.  These facilities
       are designated as Plant No. 2.
      The melt shop employing electric arc furnaces (EAFs) for the
       manufacture of stainless and electrical steels, as well as hot
       rolling furnaces for fabricating these steels into product strip,
       and EAF emission collection and cleanup equipment.  This entire
       facility is designated at Plant No. 3.
Description of the MMA Protocol
    The function of the protocol is to force the use of an orderly
step-by-step procedure for conducting a WMA at a host site.  The initial
WMA protocol was developed in earlier work, and was further refined
during the course of the present EPA-sponsored audit effort.  The
protocol  is applicable to all three categories of waste minimization
(source reduction, recycling and treatment).
    The teams employed in carrying out the audits described 1n this
report were composed .entirely of employees of outside
consulting/engineering firms.  There were 8 sequential steps executed by
the audit team, following selection of the host facility:
    1.  Preparation for the audit.
    2.  Host site pre-audit visit.
    3.  Waste stream selection.
    4.  Host site waste minimization audit visit.
    5.  Generation of waste minimization options.
    6.  Preliminary evaluation (including preparation of preliminary cost
        estimates) and ranking of options in three categories
        (effectiveness, extent of current use, application potential).

-------
    7.  Presentation, discussion and joint review of options with plant
        personnel.
    8.  Final report preparation and presentation to host site management.
    This protocol  was followed 1n carrying out the WMAs as summarized
below.
Results' of the MMA Conducted at a Generator of Corrosive Haste
Audit at Plant No.  1
    Following the  selection of a stainless steel strip pickling facility
In the EAF steelmaklng complex (designated as Plant No. 1 1n this study) 
as a host site for a WMA for corrosive waste (listed waste K062), a
pre-audlt visit was made to the facility 1n order to become acquainted
with process and waste treatment operations.  The Plant No. 1 facility
consists of a stainless steel strip annealing and pickling line for
processing of 300  and 400 series stainless steels and a pickling waste
water neutralization plant.
    Following annealing of the stainless steel strip, this material Is
fed through the continuous pickling process.  The strip 1s first treated
In a Kolene bath (the latter a mixture of molten sodium and potassium
hydroxides) at 800F for Initial dissolution of surface scale on the
strip, followed by water quench and rinse tanks to cool and flush the
treated strip, and then a nitric-hydrofluoric acid pickle using a mixture
of 8-10 percent nitric acid and 2-4 percent hydrofluoric add.  Following
a water rinse, the stainless steel strip is coiled and shipped.  Air

-------
emissions from the Kolene treatment and acid pickling tanks are

controlled through the use of fume scrubbers.   The following waste waters

are generated in the pickling process:

      A Kolene rinse water which is highly alkaline and contains sodium
       and potassium hydroxides, sodium and potassium carbonates, and
       chromates resulting from some oxidation of the chromium on the
       steel  surface during the Kolene descaling treatment.  Chromate
       levels are below 200 ppm, and the spent rinse water has a pH of
       about  12.  Approximately 45 gallons per minute (gpm) of Kolene
       rinse  water Is discharged from the process.

      Spent  HF/HNOs pickle liquor waste.  This stream is periodically
       dumped and replaced with a mixture of fresh HF/HN03 and recycled
       spent  pickle liquor.

      Rlnsewater from the HF/HN03 pickling operation.  This stream 1s
       generated continuously from the rinsing operation.

      The combined spent pickle liquor and rinse water waste stream from
       the HF/HN03 pickling operations has a pH of about 2 and cental n.5__
       dissolved metals, nitrate and fluoride.  The combined wastewater
       stream flow from the pickling and rinse tanks averages 150 gpm.
       Average composition of the spent pickle liquor/rinse water stream
       over a one-week operation (which includes a once-per-week dump of
       spent  pickle liquor into the combined wastewater stream) is given
       as:

       Parameter                            Average Concentration, mq/1

         Cr (trivalent)                                 -164
         N1                                              47
         Cd                                              <0.02
         Fe                                           1,110
         F                                            1,100
         pH                                            -2.0

    The Kolene waste rinse water and the combined spent acid/rinse water

waste stream  are treated as follows:

      Raw Kolene rinse water is treated in a mix tank with excess
       ferrous sulfate heptahydrate and sulfuric acid with about 80
       minutes retention time.  The ferrous ion reacts with chromate to
       reduce hexavalent chromium to trivalent chromium.  The pH of the
       treated waste is approximately 4.

-------
      The combined spent add/rinse water waste and the treated Kolene
       rinse water are pumped to a mix tank where slaked lime Is added.
       The final pH after 11 me addition is about 8.   The addition of lime
       causes the heavy metals present to precipitate as hydroxides and
       the fluoride to precipitate as calcium fluoride.   The resulting
       mixture of treated wastewater and precipitated solids is treated
       in a second mix tank where coagulant is added and the stream 1s
       then fed to two 30-foot diameter clarifiers operated In parallel.
       The clear overflow from the clarifiers 1s discharged to the
       outfall (a local creek) meeting the conditions of an NPDES permit
       and the underflow 1s fed to two vacuum filters operated in
       parallel.  Nonhazardous solids recovered from the filters are
       disposed of offsite and the filtrate Is recycled to the treatment
       process.
    During the formal  audit phase of this study at the Plant No. 1 acid
pickling facility, process and waste treatment operations were
Intensively studied by the audit team.  The use of various potential
source reduction and recycling options was reviewed with plant
personnel.  Plant No.  1 already recycles part of the spent add mixture
to the pickling line thus reducing fresh add use.  Based on the audit
team's evaluation and  discussions with plant personnel,  there did not
appear to be any other significant source reduction options available.
    With respect to recycling, the present neutralization treatment of
the combined Plant No. 1 pickling line wastewater stream (K062) generates
a mixed sludge for which there 1s essentially no potential for reuse.
The audit team determined, however, that the raw waste (the spent
HF/HN03 pickle Hquor/rinsewater discharged from the pickling
operation) does contain a constituent (fluoride ion) that could be
converted into a useful product-calcium fluoride (fluorspar).  The

-------
electric arc furnace (EAF) facility at this steelmaklng complex
(designated as Plant No. 3 in this report) presently purchases about
1,000 tons per year of fluorspar for use as a furnace flux material In
the steelmaklng process.  Current cost for metallurgical grade fluorspar
(approximately 75-80 percent calcium fluoride) for flux use Is $100 per
ton at the plant.  The audit team proposed a waste minimization option
for recovery of calcium fluoride wherein the combined Plant No. 1
wastewater stream at pH ~2 (excluding the treated Kolene waste) 1s
treated with slaked lime at a controlled rate so that pH "2.5 1s not
exceeded.  Calcium fluoride will precipitate selectively, and at this pH,
fluoride solubility data Indicate that a level of 65 ppm dissolved
fluoride will be achieved.  With about 1,100 ppm dissolved fluoride in
the raw wastewater, approximately 95 percent of the fluoride will
precipitate.  This 1s equivalent to about 1,300 tons per year of calcium
fluoride potentially recoverable (based on 330 days per year operation),
which more than equals the annual consumption of calcium fluoride
(fluorspar flux) in the EAF operation and Plant No. 3.  Hydroxides of
Iron, nickel, and chromium are all highly soluble at pH values below 3.0
and thus would not be expected to co-precipitate with the calcium
fluoride.
    The combined spent HF/HNO, pickle liquor and rinse water discharge
would be treated in the same waste acid neutralization system now  used to
generate the neutralized nonhazardous solids discharged offsite and NPDES
                                     7 .

-------
effluent to the outfall.  However, the neutralization would be done 1n
series In two stages, thereby effecting the recovery of a reasonably pure
calcium fluoride 1n the first stage.  After the first stage of
neutralization, the presently treated Kolene waste would be combined with
the partially neutralized waste pickle Hquor/rlnse water stream.  The
combined stream would then be neutralized and discharged to the outfall.
    If the above described option were to be put Into operation at Plant
No. 1, not only would the generation rate of sludge from K062 treatment
be reduced (resulting 1n a saving 1n offslte sludge disposal costs), but
a substantial potential savings 1n chemical purchases could be made.
Plant personnel agreed with the audit team that this was a worthwhile
option.
    With a recycling option established for the Plant No. 1 corrosive
waste, the preliminary engineering design and cost estimate for this
option was developed.  A preliminary estimate (using 1986 capital and
operating cost data) of the economics of this recycling option Indicates
the following:
      Total  capital cost (Including new drying and
       brlquetting equipment and retrofitting of one
       existing clarlfler vacuum filter system to
       make it corrosion-resistant at pH 2.5)              $300,000
      Annual operating cost                               $ 46,000/yr
      Savings due to replacement of purchased
       fluorspar                                           $100,000/yr
      Savings due to lower cost of offsite waste
       disposal                                            $ 68.000/yr

-------
      Total potential  savings                             $168,000/yr
      Estimated payback period                             2.5 years
      Estimated Internal  rate of return (based on
       an economic life of 5 years)                         28 percent
Audit at Plant No. 2
    Plant No. 2 was also selected as a host site for a WMA for corrosive
waste.  This facility consists of cold-rolling equipment to convert raw
stainless steel and electrical steel slabs Into strip, annealing ovens,
and a series of six countercurrent flow pickling lines which add pickle
stainless and electrical steel strip produced 1n the cold-rolling
equipment as well  as stainless and electrical hot-rolled steel strip
produced 1n the EAF raw steel manufacturing facility.   The six pickling
lines 1n Plant No. 2 use HF, HN03, H2S04,  and mixtures thereof and
have a total nameplate capacity of 1,833 tons per day of processed steel
strip.
    Emissions from all  of the pickle lines are controlled by use of fume
scrubbers.  Following pickling, the treated strip Is wound Into colls and
shipped.
    The six pickle lines generate the following spent pickling acids:
dilute spent sulfurlc add, dilute spent mixed sulfuric-hydrofluoric add
mixtures and dilute spent mixed nitric-hydrofluoric acid mixtures (also
containing dissolved iron salts and traces of dissolved chromium,
cadmium, and nickel), as well as rinse waters containing lower
concentrations of these components.  About 313,000 gallons of spent

-------
sulfuric acid, 144,000 gallons of spent suIfuric/hydrofluoric acid,
12,000 gallons of spent hydrofluoric/nitric acid pickle liquors and 9.7
million gallons of rinse water are generated in this facility per week.
The average composition of the combined waste stream (In terms of
critical metal and non-metal parameters) 1s as follows:
    Parameter                               Average Concentration, mq/1
       Cr              .                                     49
       N1                                                   28.5
       Cd                                                   <0.02
       Fe                                                  500
       Fluoride                                             39
       pH                                                   -2
This stream also contains emulsified and free fatty oil and grease from
the cold rolling operations at this facility.  The combined wastewater
discharge from Plant No.  2 Is sent to a central treatment facility onslte
where 1t Is treated with Urne to effect precipitation of metals as the
corresponding hydroxides and removal of fluoride as calcium fluoride.
The treated wastewater slurry 1s then pumped to a series of onslte large
lagoons where the suspended solids are removed by sedimentation and the
treated wastewater is then discharged to the outfall.  The precipitated
solids meet EP-tox1c1ty test levels for hazardous metals and the treated
wastewater is discharged to a local creek under an NPDES permit.
    During the formal audit phase of this study at the Plant No. 2 acid
pickling facility, the use of source reduction and/or resource recovery
options was studied by the audit team.  No source reduction options were
identified by the audit team.  Furthermore, the current neutralization
                                    10

-------
treatment of the combined pick-Ung and cold-rolling aqueous waste stream
(listed waste K062) generates a mixed metal hydroxide-gypsum-calcium
fluoride sludge for which there is no reuse potential.   The key to waste
minimization at this plant 1s suitable segregation of selected spent acid
wastes before these wastes are combined at the process  building outlet
flume (the latter discharging to a centralized wastewater neutralization
process)  an option that plant personnel Indicated would be highly
disruptive to plant operations and costly as well.  However, assuming
that pickling waste stream segregation is feasible, the following waste
minimization option is proposed by the audit team:
      As a recycling option segregate an appropriate amount of waste
       sulfurlc acid pickle liquor (13;5 million gallons^per year are
       potentially available from three pickling lines).   Two uses for
       this material were Identified by the audit team:
       -  Reduction of hexavalent chromium in a bleed stream of venturi
          scrubber waste resulting from scrubbing of EAF dust 1n Plant
          No. 3.
       -  Reduction of hexavalent chromium in the Kolene waste in Plant
          No. 1.
       Both of these uses presently employ purchased solid ferrous
       sulfate heptahydrate.  Recycled sulfurlc acid pickle liquor
       (containing 5 to 10 percent dissolved ferrous sulfate) is usable
       for this purpose.
*Based on 300 days per year (43 weeks/yr) operation of the Plant No. 2
acid pickling lines.
                                    11

-------
    The technical and economic feasibility of this recycling option was
evaluated by the audit team.  A preliminary estimate of the economics of
this recycling option Indicates the following:
      Total capital cost (Including new tankage,
       piping, and pumps, a tank truck to haul  the
       spent add between-the various facilities
       and suitable permitting)                            $255,000
      Annual operating cost                               $ 20,000/yr
      Savings due to replacing purchased ferrous
       sulfate heptahydrate with waste ferrous
       sulfate/sulfuric add pickle liquor                 $ 44,000/yr
      Savings due to lower 11 me usage for
       neutralization at Plant No. 2                       $  8.000/vr
      Total potential savings                             $ 52,000/yr
      Estimated payback period                              8.0 years
    The payback period of eight years 1s almost three times the usually
acceptable period of three years, making the proposed project distinctly
unattractive from an economic standpoint.  In addition to the obvious
economic disadvantage of this option, plant personnel are concerned with
the-cost and disruption of decoupling the internal discharge points of
this waste (in the pickling process building) needed to segregate the
appropriate amount of this waste for recycle.  No information was
available from the plant to permit estimation of the cost of decoupling
this waste quantity from the remaining waste streams discharged from the
facility.  However, since the use of this option does result in some
waste minimization as well  as a small but measurable extension in the
                                    12

-------
life of the present waste lagoon disposal system (approximately 10
percent less calcium sulfate and metals-containing solids would be
deposited 1n the lagoons), the audit team believes that the option should
continue to be reviewed at Plant No. 2.
Results of the HMA Conducted at a Generator of Heavy Metals Haste
    During the pre-aud1t visit to this facility (designated as Plant
No. 3) the audit team became acquainted with process operations at the
EAF melt shop and the wastes generated from these operations.  The melt
shop contains three 165 ton capacity EAFs used to manufacture 300 and 400
series stainless steels and silicon steels, as well as one 175 ton
argon-oxygen decarburlzer (AOD) used to further refine the raw stainless
steels.  Steels leaving the furnaces are processed through continuous
casting and hot-rolling operations to produce colls of strip which are
sent to the acid pickling facility (Plant No. 2) for final processing.
Production of stainless and electrical steels is approximately 270,000
tons per year.
    The pre-audit visit yielded the following waste stream
characterization and treatment Information at:
      The three EAFs and the AOD at the melt shop generate about 8,000
       tons per year (TPY) of particulate emissions (listed waste K061).
      Of these emissions, approximately 7,000 TPY are removed from the
       EAF vent gases using venturl scrubbing; the remaining 1,000 TPY
       Include EAF fugitive emissions as well as emissions removed from
       AOD vent gas, and are recovered in a baghouse.
                                    13

-------
      The venturi scrubber slurry 1s clarified and filtered with the
       filter cake, containing about 30 percent water,  discharged at the
       rate of 10,000 TRY.

      The combined EAF dust and sludge (listed waste K061) leaving the
       plant totals about 11,000 TPY and Is sent to offsite hazardous
       waste landfills at an annual  cost of approximately $1.0 million.

    During the detailed audit phase  of this study at Plant No. 3, the use

of source reduction and/or recycling options for EAF dust emissions was

reviewed with plant personnel.  The  following Information was developed

which provided a disincentive to further exploration of these waste

minimization options for the K061 waste generated at the plant:

      With the primary steel products of this facility being 300 and 400
       stainless grades, there will  always be significant levels of the
       following hazardous metals 1n the EAF dust:  chromium and nickel
       (because of the alloying requirements of stainless steels),
       cadmium ajid lead.  Contributions to these metals In the EAF dust
       come from the scrap feed as well as from the alloying additives.
       These EAF dust constituents are expected to always generate levels
       of one or more of these hazardous metals In excess of the
       presently allowable RCRA levels 1n leachate from the TCLP
       procedure (using acetate buffer).  Source reduction 1s therefore
       not an available option to the plant.

      Plant No. 3 has made a number of attempts (without success) to
       recycle EAF dust to the steelmaking furnaces.  Plant personnel
       also Indicated that because of the volume of dust generated and
       the sensitivity of the steel  product quality to tramp elements 1n
       the recycled dust, 1t 1s unlikely that a large percentage of this
       waste could ever be recycled  to the process.

      Hith respect to use of the K061 waste by a metals reclaimer, the
       principal ingredient of value (zinc) is only present to the extent
       of 8 to 10 percent in the Plant No. 3 EAF dust.   In order for EAF
       dust to be economically attractive to reclaimers, it should have
       at least 20 percent zinc and  preferably nearer 50 percent.
       Additionally, internal recycling of the EAF dust, which would tend
       to enrich the zinc content of the residual material, is presently
       not available to Plant No. 3  due to factors discussed above.
                                    14

-------
    No source reduction and/or recycling options appear to be available
to reduce the quantity of K061 waste generated by this plant.  The last
resort 1s an alternative treatment step proposed by the audit team which
appears to be the only option available for this waste stream and 1s
summarized as follows:
    Plant No. 3 would (1) convert the material Into a nonhazardous waste
using a chemical stabilization technique (using 11me kiln or cement kiln
dust and water blended with the EAF dust to generate solidification
reactions and create an essentially Insoluble matrix), (2) apply to have
the waste dellsted by EPA, and (3) dispose of the dellsted material In an
onslte dedicated landfill.
    A preliminary estimate (using 1986 capital and operating cost data)
of the economics of this option (for 10-year onslte disposal) results In
the following:
      Total capital cost                                 $1.75 million
      Annual operating cost                              $0.42 m1l11on/yr
      Estimated payback period                               3 years
      Estimated Internal rate of return
       (based on an economic life of 5 years)                20 percent
      Present annual K061 waste generation rate         11,000 TPY
      Proposed annual treated K061 waste
       generation rate                                   22,000 TPY
      Cost of treatment and disposal of waste              $35/ton
      Annual savings (over amortized life of
       treatment plant and onsite landfill),
       1f this option was implemented                    $227,000/yr
    The disposal cost of K061 waste using this treatment option is
estimated at $35/ton of treated waste compared to $100/ton for the
current cost of disposal of the raw waste even though the treated waste
                                    15

-------
generation rate has doubled.  It 1s therefore recommended that this

treatment option be pursued further with an Initial  bench-scale effort to

establish the appropriate waste stabilization technique.

    It should be noted-that plant personnel are In agreement with this

assessment of available waste minimization options.

Conclusions

    The following conclusions can be drawn based on the study summarized

herein:

      The WMA methodology can be successfully applied to the
       minimization of hazardous waste, 1n this case corrosive waste
       (K062), 1n at least one Industry - the specialty steels segment of
       the EAF steelmaking Industry.  Application of the WMA protocol to
       a plant 1n this Industry (a stainless steel pickling facility In
       an EAF steelmaklng complex), resulted 1n the Identification of a
       technically and economically feasible recycling option for the
       recovery of fluorspar (calcium fluoride) from a corrosive waste
       stream.  The fluorspar would be used Internally 1n place of
       purchased material.  Use of this waste minimization option results
       in savings of $168,000 annually and a 30 percent reduction in
       final  waste disposal volume.

      Application of the WMA methodology to another corrosive waste
       generated at another steel pickling facility in the same
       steelmaking complex, resulted in the development of a recycling
       option requiring segregation of a portion of this waste (waste
       ferrous sulfate/sulfuric acid pickle liquor) for internal recycle
       replacing purchased ferrous sulfate heptahydrate.   However, this
       option is not economically feasible and this disadvantage
       outweighs the principal advantage of a small  prolongation of the
       life of the onslte facility for disposal of neutralization sludge
       from present waste treatment.

      An attempt to apply the WMA methodology to another hazardous waste
       stream:  heavy metals - containing EAF dust (K061) generated in
       the EAF steel plant at the same steelmaking complex, could not be
       considered as successful inasmuch as both of the more preferred
       waste minimization approaches (source reduction and recycling),
                                    16

-------
were not technically feasible in this case.  However, the
identified treatment approach:  detoxification of the waste by
onsite chemical stabilization/solidification treatment, followed
by onsite disposal in a dedicated landfill, appears to be worth
investigating further based on the results of a preliminary
technical and economic evaluation of this option.

It is believed that the on-site audit (employing the WMA
methodology developed in this study is a distinctly useful tool
for waste minimization due to the one-to-one contact with the
Industrial waste generators by qualified engineering professionals
on the audit team.
                             17

-------