United States
                    Environmental Protection
                    Agency
Hazardous Waste Engineering
Research Laboratory
Cincinnati OH 45268
                    Research and Development
EPA/600/S2-88/008  Mar. 1988
&EPA         Project Summary
                    Waste  Minimization  in  the
                    Printed  Circuit  Board Industry-
                    Case  Studies

                    Thomas Nunno, Stephen Palmer, Mark Arienti, and Marc Breton
                      The full report presents information
                    on waste minimization practices cur-
                    rently employed in the printed circuit
                    board (PCB) and semiconductor manu-
                    facturing industries. Case  studies
                    conducted at six facilities evaluated the
                    technical, environmental  and cost
                    impacts associated with the implemen-
                    tation of technologies for reducing the
                    volume and toxicity  of PCB  metals-
                    containing sludges and solvent wastes.
                    The analyses of these data are the basis
                    for demonstrating waste minimization
                    technologies  to reduce hazardous
                    waste.
                      This Project  Summary was devel-
                    oped by EPA's Hazardous Waste Engi-
                    neering Research Laboratory. Cincin-
                    nati, OH, to announce key findings of
                    the  research project that is fully doc-
                    umented in a  separate report of the
                    same title (see Project Report ordering
                    information at back).

                    Introduction
                      The purpose  of this project was to
                    evaluate the effectiveness of various
                    waste minimization practices or technol-
                    ogies  in the printed circuit board and
                    semiconductor  manufacturing  indus-
                    tries. The most significant waste streams
                    in these industries are waste halogen-
                    ated solvents from photoresist stripping
                    and developing operations (RCRA Waste
                    Code  F001-F003), and metal-bearing
                    sludges (RCRA Waste  Code F006) from
                    the  treatment  of metal  plating  and
                    etching rinsewaters. This Project Sum-
                    mary  presents the findings of case
                    studies conducted at five printed circuit
                    board  manufacturing facilities and one
                    commercial treatment/recovery facility.
Each facility investigated employs some
practice that  requires offsite disposal.
Two  of the case studies focus on the
recovery of spent halogenated solvents,
and  the remaining four discuss the
recovery or reduction of metal plating and
etching process wastes. Table 1  sum-
marizes characteristics  of facilities
investigated that range from small job
shops to large integrated facilities.

Metal Plating Bath Waste
Minimization Case Studies
  Metal plating wastes generated from
plating bath  dumps,  rinses, etching
machines and  scrubbing operations
generate copper-, nickel-, tin-, and lead-
contaminated wastes.  Four of the six
case studies investigated under this
research project focus on the minimiza-
tion of sludges  generated primarily by
copper plating and etchant baths and
copper and tin/lead rinsewaters.
  The common objectives of each of the
technologies evaluated are: (1) minimi-
zation of metals sludges generated; (2)
compliance with effluent guidelines or
local discharge limitations; and (3)
reduction in operating  costs over other
conventional alternatives. The following
discussion briefly summarizes each case
study, the nature  of the  minimization
technology, the measurements data
collected and the results obtained.

Facility A

Description
  Facility  A is  an offsite Treatment,
Storage, and Disposal (TSDF) facility
which processes concentrated dumps

-------
Table 1.    Summary of Facilities Tested Under Waste Minimization Case Study Program
  Facility Name
Description
                                               Wastes Treated/Reduced
           Technology
    Residuals
Facility A        Treatment storage disposal
                facility handling electro-
                plating baths, waste
                etchants, spills, etc.
                Capacity: I.OOOgph
                (24,000 god).

Facility B        Contract PC board
                manufacturing shop
                Employees: 77
                Production: 500,000 sf/yr
                Sales: $7 Million/yr

Facility C        Computer manufacturer.
                Employees: 10,000  •


Facility D        Electronic equipment mfgr.
                PC board manufacturing using
                the subtractive technique iri-
                the MacDermid process.
                Employees: 260


Facility E        Computer manufacturer.
                PC board manufacturing using
                additive techniques.
                Employees: 600
                Production: 600,000 sf/yr

Facility F        PC Board manufacturer
                2-sided single layer circuit
                boards.
                Production: 480.000 sf/yr
                    -Nickel plating baths
                    -Copper plating baths
                    -Cyanide
                    -Cupric chloride etchant
                    -Electroless plating rinses
                    -Electroplating rinses
                    -Methyl chloroform resist
                     developer
                    -Freon resist developer

                    -1,1,1 -trichloroethane
                     resist developer
                    -1,1,1 -trichloroethane
                     still bottoms
                    —Acid copper plating bath
                    -Acid copper plating
                     rinsewaters
                    - Tin/lead plating
                     rinsewaters
 -Sodium hydroxide precipitation
 -Sodium borohydride reduction
 -Alkaline chlorination
 -Sodium borohydride reduction
 -Memtek ultrafiltration system
—Solvent distillation/
  fractionation recovery of
  resist developers.

—2-stage solvent distillation

—(1) DuPont RISTON SRS-120
     solvent recovery still
—(2) Recyclene Products, Inc.
     RX-25 still

—Activated carbon regeneration
  of spent plating baths.
 -Agmet Equipment Corp.
 -Electrolytic recovery units.
Sludge product
Sludge product
Still bottoms
Still bottoms
                                                                                Spent activated
                                                                                carbon
Metal foil
from the metal plating and printed circuit
board industries, including alkaline
etchants, acid plating baths,  nitric acid
rack strip baths,  and electr.oless plating
cyanide baths. The average total metals
concentration in the incoming waste was
reportedly 12 g/L (12,000  ppm). These
waste streams  are  classified into the
following four  categories:  (1) acidic
metals solutions;  (2)  alkaline  metals
etchant solutions; (3) cyanides;  and (4)
chelated metals solutions.  The  case
study for this facility  focuses on the use
of a sludge minimizing treatment tech-
nology for the  metals  and  cyanides
wastes.
   Initially, the facility was designed to
operate  using lime and ferrous sulfate
precipitation  of  metals  as the primary
means of waste treatment.  When the
high cost of  land  disposal of the lime
sludges  was considered,  alternate
means of treating  and disposing of the
waste were evaluated.
   The unit processes selected to detoxify
the wastes and  recover metals at Plant
A currently include sodium hypochlorite
                 oxidation of cyanides (alkaline chlorina-
                 tion), sodium hydroxide precipitation, pH
                 adjustment, sodium  borohydride reduc-
                 tion (with sodium metabisulfite stabili-
                 zation), sedimentation, plate and frame
                 filter press (for sludge dewatering), rapid
                 sand filtration,  and  ion  exchange
                 columns for effluent  polishing.

                 Results
                   The primary purpose of the Facility A
                 case  study  was  to evaluate sodium
                 borohydride as a viable waste treatment
                 alternative for reducing RCRA Hazardous
                 Waste Code F006 spent electroplating
                 baths. The evaluation criteria were the
                 ability of  sodium  borohydride  (SBH) to
                 effectively meet local compliance stand-
                 ards and  produce a high density, low-
                 volume sludge. The test program evalua-
                 tion  relies mainly on the trace metals
                 results to evaluate system performance.
                   The SBH reactor was sampled for trace
                 metals  on  the  influent, effluent,  and
                 sludge streams. Both filtered and unfil-
                 tered samples were collected  and ana-
                 lyzed for  eight  selected  metals.  The
            unfiltered  sample showed  little or nc
            reduction  as  expected.  However, the
            filtered sample showed individual metal:
            reduction efficiencies which ranged frorr
            16.1 to 99.8 percent. The observed range
            in  efficiency  data was  attributed  tc
            variations in concentration and chemica
            potential   (quantity of  free  energy
            required for an ionic species to obtair
            equilibrium) of each of the metallic ion;
            contained in the solution. Overall, SBr
            was able to reduce 6.91 kg of the  initia
            influent metals loading of 7.25 kg. These
            results represent  a  greater than 95
            percent reduction in total metals for <
            complex waste stream. The remainder o
            the metals influent loading (0.337 kg
            consisted of over 70 percent calcium.
              An additional objective of this progran
            was to evaluate the ability of Facility /
            to  consistently meet local  pretreatmen
            requirements. The resultant data for tw<
            separate batch runs showed discharge:
            in  excess of effluent limits, apparent!'
            due to incomplete  polishing caused  b'
            cation  exchange column  breakthrough
            Since the test program was completed

-------
> Facility A has instituted  the  use of a
 quality control holding tank and further
 waste processing optimization to remedy
 these problems.  Follow-up discussions
 with the  local sewer authority revealed
 that Facility  A's effluent  quality  has
 improved considerably and  is  now con-
 sistently meeting compliance guidelines.
   In addition to  assessing  wastewater
 effluent  characteristics, the testing
 program was designed  to  evaluate
 uncontrolled process air emissions. The
 results were obtained by Draeger tube
 analysis of grab and integrated samples
 of exhaust gases taken from the process
 reactor exchaust ducts. The  emission
 results showed a frequent  presence of
 hydrochloric  acid and  hydrogen  gas
 accompanied by occasional presence of
 ammonia and sulfur dioxide. One of the
 hydrogen emissions grab sample results
 (6.0  percent) is  significant since  this
 value is greater than the lower flamma-
 ble limit for hydrogen (4.0 percent). Grab
 sample concentrations for ammonia and
 sulfur dioxide also exceeded adopted
 short-term exposure limits (STEL) for
 these substances.
   Analysis of  the nickel/cyanide  and
 SBH sludges shows total metals contents
 (dry weight) of approximately  35 and 6
 percent,  respectively.  Neither sludge
 result supported Facility A's claim of 60
 to 70 percent metals content (dry basis).
 While the SBH sludge result was signif-
 icantly below performance expectations
 (70  percent metals), the exact cause of
 these results was not  discernable.
 Possible  explanations include: (1)  a
 possible  process upset;  (2)  sampling
 error; and (3) analytical error.  It  seems
 most probable that a process upset  was
 responsible  for  these results,  since
 blinding of the sludge press occurred on
 the  SBH  press.  Based on other SBH
 reduction case study results conducted
 under this program, it is  reasonable to
 assume  that  these results  are  not
 representative,  since typical sludge
 metals contents should be greater than
 70 percent.
   EP Toxicity analyses were  also con-
 ducted for both the nickel/cyanide and
 SBH reactor  sludges. The results of the
 tests clearly show that for Facility A
 influent metals concentrations, the SBH
 sludge produced  is fairly  stable in  that
 its leachate characteristics are below EP
 Toxicity limits for all metals.  However,
 note that the waste is still classified as
 F006 hazardous waste.
   Another objective of the Facility A case
 study was to evaluate the ability of SBH
to economically reduce  F006  waste
streams. At the time of testing. Facility
A reduction chemistry was very ineffi-
cient at  $19.80/lb of copper reduced.
However, through process optimization,
chemical   costs   have  reportedly
decreased  over 63 percent, bringing
process  economics within acceptable
limits. The case  study follow-up  for
Facility A has indicated that the cost of
copper reduction has been lowered to
$7.27/lb of copper.

Facility B
Description
  Facility B is a captive printed  circuit
board manufacturing facility employing
77 people in Santa Ana, California. Gross
sales  are  approximately $7 million
annually on production of 500,000 ft2 of
board. Production  at Facility  B uses a
special hybrid process, employing ele-
ments of both additive and semi-additive
printed circuit production  techniques.
Process wastes of interest to  this study
include rinsewaters from the electroplat-
ing  and  etchant baths. The  principal
components of the acid copper electro-
plating  baths are copper  sulfate  and
sulfuric acid.  Facility  B uses a slower
acting etchant (sodium chloride, sodium
chlorate, and muriatic acid) which etches
copper from the board, and yields cupric
chloride in the waste stream.
  Facility B uses a rather unique end-
of-pipe treatment system employing SBH
treatment and ultrafiltration  (Memtek)
technology  for solids separation. In this
process,  incoming  plating and etching
wastes are adjusted to pH 7-11 by adding
sodium  hydroxide or sulfuric acid.
Sodium borohydride is added to  obtain
an oxidation reduction potential (ORP) of
approximately -250 mv or  less.  The
reacted  waste  then  feeds  from the
concentration tank to a Memtek ultrafil-
tration unit from which the permeate is
discharged  to municipal treatment, and
the  concentrate  is returned to the
concentration tank. A  small  plate and
frame sludge filter  press dewaters the
sludge which is drawn from the bottom
of the concentration tank.
  Points  of interest in evaluating the
Facility B waste treatment system for this
case study  were: (1) compliance  of the
ultrafiltration permeate  (wastewater
discharge)  with  local  and Federal dis-
charge standards; (2) the volume and EP
toxicity of the sludge filter cake; and (3)
economic evaluation against comparable
technology (lime and ferrous sulfate
treatment).
 Results
  The objective of the sampling program
 was to evaluate the effectiveness of the
 SBH technology in use by Facility B. The
 effectiveness was measured in terms of
 metal reduction efficiency  and minimi-
 zation of hazardous waste streams. Data
 derived from the  metals concentrations
 in the influent and effluent streams were
 used to determine the  effectiveness of
 the SBH reduction system in both meet-
 ing effluent guidelines and minimizing
 releases to the environment.
  Analysis of the influent  and effluent
 streams metals characteristics showed
 that copper was reduced most efficiently
 (99.82 percent), while nickel reduction
 was the least  efficient (45.5  percent).
 Differences in removal efficiencies were
 attributed to variations in concentration
 (higher  removals  for higher concentra-
 tions), but the  chemical potential  may
 also have been a factor. Approximately
 144.7 Ib of combined metals were
 reduced to elemental form  by the  SBH
 reaction  system,  representing  a com-
 bined reaction efficiency of 99.8 percent.
 Despite deviations from design operating
 conditions, the SBH/ultrafiltration sys-
 tem  performed very well.  EP Toxicity
 leachate test results for Facility  B filter
 press sludge clearly show that the SBH
 sludge produced is fairly stable. Leachate
 characteristics  are below  EP Toxicity
 limits for all metals. However,  note that
 the waste is still classified  as F006
 hazardous waste.
  In  addition, an  economic comparison
 of the use of SBH versus  lime-ferrous
 sulfate chemistries was conducted. The
 results demonstrate that in this applica-
 tion, SBH would  be superior  to lime-
 ferrous sulfate for the following reasons:
 (1) sludge disposal costs and  volumes
 would be reduced by 93.5  percent; (2)
 overall operating expenses would be 48
 percent  lower; and (3) sludge generated
 by the SBH reduction process was 78
 percent copper and suitable for reclama-
tion (due to the high copper content).
  The use of the SBH and ultrafiltration
treatment at Facility B is favored by the
 use of the chloride etch process  in lieu
 of the more commonly preferred ammo-
 nium  peroxide  etch. The  ammonium-
 based etchants  create borohydride
 sludge stability problems which require
tighter treatment process control and use
of stabilizers such as  sodium metabisul-
fite.  Additional  factors  that favor the
economics of SBH treatment at Facility
B include: (1) the  use of cupric chloride
etchant; (2) nigh copper concentrations

-------
and low organic loadings seen at this
facility;  and (3) low effluent limitations
required by the sanitation district.
  Based on the above results, it appears
that SBH  reduction is an  effective
technology which can  be  utilized to
reduce  complex  metal electroplating
sludges and  render them reclaimable,
and possibly  less hazardous. Note that
the economics of SBH  technology  is
highly dependent on site-specific factors
and warrants a detailed study prior to
implementation.

Facility E

Description
  Facility E began operations in January
1982 as a manufacturer of customized,
fine-line  multilayer printed  circuit
boards.  Facility E initiated an ambitious
waste  minimization program  in mid-
1984. Since  that time,  production has
roughly doubled, but liquid discharge to
the wastewater treatment plant has
remained  constant and wastewater
sludge generation  has dropped roughly
30 percent. Waste minimization  efforts
continue to  center around in-process
modifications  to use nonhazardous or
reclaimable solutions, to reduce water
consumption  and bath dump frequency,
and to  optimize wastewater treatment
operations.
  At Facility E, boards are pattern plated
with eight acid copper and one aqueous
tin/lead plating  baths  in  a  48-tank
plating line. The line begins with a nitric
acid (HNO3)  rack  strip  tank. After the
racks are stripped, boards are loaded and
then  undergo rinsing, cleaning  with
phosphate solutions (H3PO<, Electro-
clean PC2000), and more rinsing before
being plated. Acid copper baths contain
CuSO4, organic brighteners, and chlo-
rides with  copper  concentrations of 24
oz/gal.  The  general processing  proce-
dure is to activate the board surface (HCI),
plate, clean/rinse and replate.
  In plating  operations, addition  agent
and photoresist breakdown products will
incrementally accumulate and contam-
inate  an  electrolytic (charge  carrying)
plating  bath.  In the  absence of  a bath
regeneration  system, the  manufacturer
would  typically  be forced to  either
discharge the spent plating bath to the
wastewater treatment  plant or send it
offsite for disposal. In either case, large
quantities  of metals containing  sludge
(RCRA  Waste Code F006) would be
generated  and subsequently land dis-
posed. At Facility E, these spent  plating
baths are regenerated through activated
carbon filtration (used to remove  built-
up organic bath contaminants) and then
returned to the process. Copper and
solder plating baths are  treated with
activated carbon once  every three
months and every month, respectively.
The frequency of cleaning is determined
by organic  contaminant build-up.  Elec-
troplating baths never have to be dumped
with  this  arrangement under  normal
processing conditions.
  Activated  carbon treatment  is  per-
formed in a batch mode for acid copper,
solder and  nickel microplating  baths in
three separate systems. The bath recla-
mation system consists of a holding tank,
mixing tank, and MEFIAG paper-assisted
filter. For acid copper treatment, 2,400
gallons of contaminated solution  is
pumped into a 3,000-gallon mixing tank.
Hydrogen  peroxide is  added  and the
temperature of the bath is maintained at
120  to  130°F for  1  hour.  Powdered
activated carbon (80 Ib)  is added and the
contents are  mixed for 3 to 4 hours to
oxidize  volatile organic species. The
solution is recirculated through a paper-
lined MEFIAG filter several  times  to
remove  the activated carbon. The filter
solids and paper are removed as needed
when a  predetermined pressure drop
across the  filter is reached. When the
bulk of  the activated carbon has been
removed (generally after three passes of
the solution through the filter), the filter
is precoated with 5 gallons of diatoma-
ceous earth. The  solution  is again
recirculated through the  filter  until a
particulatetest indicates sufficient solids
removal (no residue detected on visual
examination of laboratory filter paper).
Total spent solids from  plating bath
purification is 1-V4 drums every 3 months
which is landfilled.

Results
  The purpose of this case study was to
evaluate the extension  of electroplating
bath  lifetimes (and subsequent waste
reduction) by activated carbon  removal
of organic brightener breakdown  prod-
ucts. The acid copper baths were selected
for study since recovery of this sojution
results in the most significant.amount
of waste minimization.
  Sampling and analysis were conducted
on three process streams associated with
activated carbon bath reclamation. Based
on resultant analytical data the following
conclusions were drawn:

• Forty-seven percent of the organic by-
   products  and  brighteners  were
   removed  from the contaminate)
   solution:

 • Low molecular weight organics sue)
   as carboxylic acid  derivatives are no
   preferentially adsorbed;

 • Reduced  sulfur (a  brightening an<
   leveling agent) is  oxidized and vola
   tilized during treatment; and

 • Inorganic contaminants  such as tir
   and lead  are  also removed  (37.£
   percent and 24.5 percent,  respec
   lively) as a beneficial by-product of th(
   treatment process.

  In recovering spent electrolytic platini
baths. Facility E was able to save ove
$50,000  in  hazardous waste disposa
and raw  material purchase  costs. Thesi
savings represent  a  payback period o
only 3  months for purchasing the  acti
vated carbon recovery  system. Thi
relatively short payback period, combinei
with the volume  of plating  solutioi
regenerated,  make  activated carboi
treatment a cost-effective and environ
mentally safe technology for reducing thi
quantity of hazardous waste that wouli
otherwise be land disposed.

Facility F

Description

  Facility F is an independent manufac
turer of  printed  circuit boards.  Thi
normal production volume of the facilit
is 40,000 ftVmonth. The major wasti
streams of interest to this case study an
rinsewaters that  follow  electroplatini
and etching processes. Prior to imple
mentation of  the electrolytic  recover
technology being  studied, these rinse
waters contained copper and  lead  a
concentrations of  up to 3,000  mg/L
Because  of  this,  the concentration  o
these  metals in  the final  effluen
exceeded pretreatment  standards  (4.!
mg/L for copper and  2.2 mg/L for leaC
for discharge to the  city sewer system
To decrease the concentration of metal
in the  effluent,  the  facility convertei
several rinse tanks into  static dragou
tanks in  order to recover metals fron
rinse baths following  copper electroplat
ing, tin/lead electroplating, electroles
copper plating, and a copper microetc!
process. The quantity of metal recovere
from the electroless copper rinse and th
copper microetch  was  small. Thus, th
reactors were removed from these bath
and installed at the copper and tin/lea

-------
rinse baths  where there was  more
potential for metal recovery.
  The electrolytic reactors used at this
facility  are Agmet Equipment  Corp.,
Model 5200* reactors. They consist  of
a wastewater sump, a pump, and the
anode and cathode, contained within a
rectangular box with dimensions  of
approximately 22 in. x 10 in. x 22 in. The
anode was cylindrical and was encircled
by  a  stainless  steel cathode  with  a
diameter of 8 in. and a height of 6 in.
The anode material  used for copper
plating solutions is titanium. For tin/lead
plating solutions, however, a columbium
anode was required because the fluoro-
boric acid in the tin/lead plating solution
was extremely corrosive to titanium. The
columbium anode increases the cost  of
these electrolytic units to $4,500, as
opposed to $3,500 for the titanium anode
units.

Results
  At the time of testing, four electrolytic
reactors were being used for recovery of
copper,  and three were being used for
recovery of tin/lead. To  evaluate the
performance of these units, samples  of
the plating bath,  dragout, and rinse bath
were analyzed. Conclusions that  were
drawn  based on  the resultant  data
include:

•  Recovery of  copper from the acid
   copper solution  is very  effective—
   rates of recovery were 4 to 5 grams/
   hour/unit,  representing a current
   efficiency of nearly 90 percent.

•  Recovery of  tin  and  lead was not
   effective at  the  itme of testing-
   concentrations of these two metals in
   the dragout were not significantly less
   than  in the plating bath.  However,
   evaluation of the data was difficult
   because the analytical results for
   some of these samples were  incon-
   clusive due to matrix interference.

•  Use  of  in-line electrolytic recovery
   was not able to reduce metal concen-
   trations sufficiently as to enable this
   facility  to   meet  pretreatment
   standards.

•  Electrolytic  recovery  would  signifi-
   cantly reduce the amount of sludge
   generated  if a  lime  precipitation
   system were utilized to remove metals

"Mention  of trade names or commercial products
 does not constitute endorsement or recommenda-
 tion for use.
   from the final plant effluent. For this
   facility,  a reduction of 32 tons/year
   would be realized.

 • At a sludge disposal cost of $200/ton,
   the annual cost of electrolytic recovery
   would exceed the savings. However,
   if sludge disposal costs increased to
   $300/ton, the  savings (at least for
   copper  recovery)  would  exceed the
   processing costs.

  Electrolytic recovery methods remove
 metals from an aqueous solution in  a
 metallic form which allows  for the use
 of the recovered material as scrap metal.
 Conversely, hydroxide precipitation
 removes the metal  from  solution and
 generates  a sludge  with  a  low  metal
 concentration.  In most cases, the only
 method of  handling this sludge is land-
 filling at a  high cost. Therefore, electro-
 lytic recovery is useful in minimizing the
 quantities  of metal-bearing  sludge that
 must be landfilled. The cost effectiveness
 of this type of  technology will increase
 as sludge disposal costs increase in the
 future.

 Resist Developing Solvent
 Recovery Case Studies

  Two case studies evaluated under this
 program focused upon the minimization
 of developer solvent wastes and sludges
 which might require either land disposal
 or incineration.  In general, the recovery
 of resist stripping and developer solvents
 is not unique within the  PC board
 manufacturing  industry. However, the
 recovery systems evaluated at the two
 facilities discussed  below  represent
 state-of-the-art technology applications.
 In the case of Facility C, the technology
 involves the separation of a two-solvent
 system with subsequent recovery and
 reuse of each  solvent. In the  case of
 Facility D,  the technology  evaluated
 further  recovers  the solvent bottoms
 product of the initial recovery unit.

Facility C

 Description
  Facility  C manufactures  computing
 equipment  including  logic, memory and
 semiconductor devices,   multilayer
 ceramics, circuit packaging, intermediate
processors and printers. One of the major
 hazardous waste streams that is gener-
 ated is spent halogenated organic sol-
vents (RCRA Code  F002).  Methylene
chloride is  used in  resist stripping of
 electronic panels. Methyl  chloroform
 (1,1,1-trichloroethane) used  in  resist
 developing of electric panels  and sub-
 strate chips. Freon  used  in surface
 cleaning  and developing of  substrate
 chips. Perchloroethylene used in surface
 cleaning of electronic panels.
   The spent solvents from  photoresist
 stripping  and developing  are  contami-
 nated with photoresist solids at up to 1
 percent,  and the solvents  used for
 surface cleaning  are contaminated  by
 dust, dirt  or grease. Waste solvents are
 recovered at Plant C by  distillation or
 evaporation and returned to the process
 in which they were used.  Several types
 of  equipment are used including  box
 distillation units to recovery methylene
 chloride and  perchloroethylene,  flash
 evaporators to recover methyl chloro-
 form, and a distillation column to recover
 Freon.
   There are two identical  flash evapor-
 ators at the facility, each with a capacity
 to recover 600 gallons  of methyl chlo-
 roform (MCF)  per  hour.  The  flash
 chamber operates  at a vacuum of 20 in.
 Hg, allowing the MCF to vaporize at 100
 to 110°F.  The units are operated one to
 two shifts/day depending on the quantity
 of waste solvent being generated.
  A packed  distillation column is used
 to  recover pure Freon from  a waste
 solvent stream containing approximately
 90 percent Freon and 10 percent methyl
 chloroform. Waste  is continuously fed to
 a reboiler  where it  is vaporized  and rises
 up the packed column. Vaporized Freon
 passes through the column, is condensed
 and recovered at a rate of 33 gal/hour.
 MCF condenses on the packing and falls
 back  in the reboiler. The  distillation
 bottoms are removed when the concen-
 tration of methyl chloroform reaches 80
 percent (approximately 1 to 2 weeks).
  There are also two identical  box stills
 at the facility, each with  a capacity to
 recover 475 gph of methylene chloride.
 These are very simple units consisting
 of ah 800-gal still pot  with hot water
 heating coils. The  contaminated meth-
 ylene chloride is  heated to  between
 103°F and 108°F,  and clean solvent is
 condensed overhead.

 Results
  Sampling and analysis were conducted
 on process streams associated  with two
 of the solvent recovery processes. One
 of these processes was the flash evap-
 orator used for  recovery of methyl
 chloroform (1,1,1-trichloroethane), and
the other  was  the distillation  column

-------
used to recover Freon TF from a Freon/
methyl chloroform mixture. The conclu-
sions drawn from the sampling  and
testing program were:

• At least 95 percent of the solids are
   removed from the  solvent waste
   influent;

• The recovered  product is at least as
   clean as the virgin material; and

• The still  bottoms from recovery of
   contaminated solvent still contain a
   high fraction (90 percent) of solvent.


  The recovery of spent solvents at the
facility is motivated primarily by eco-
nomic benefit.  In recovering spent
solvent, the company saves  over $10
million  annually, compared to offsite
recovery. The  savings  per  pound of
methyl chloroform, methylene chloride,
and Freon recovered is $0.18, $0.18, and
$0.61, respectively.
  The high cost savings are primarily due
to the fact that the solvents recovered
are reused  onsite,  thus reducing the
quantity (by greater than 95 percent) of
new or  virgin solvent that must  be
purchased.  Offsite  recovery  could  be
conducted,  but  at  much  higher  cost.
Since the  rate of generation of spent
solvent is so high, the initial  expense of
purchasing  recovery  equipment  is
quickly returned.
  To landfill or dispose  of such a  large
quantity of spent solvent  by any other
method   would   be  economically
unacceptable.  Incentives other  than
economic reasons for onsite recovery
include:


• Reduction in  the risk of a  spill of
   solvent in transporting the waste to
   aTSDF; and

• Reduced liability related to an accident
   at the TSDF resulting in the  release
   of spent solvent.


  Facility C is trying to further reduce the
quantity  of waste solvent that must be
sent offsite for recovery. They intend to
do this  by  recovering the still bottoms
generated by distillation of Freon/methyl
chloroform waste.  In  addition,  they
eventually plan to phase out the use of
methyl chloroform and methylene chlo-
ride  and replace  these  materials with
aqueous-based photoresist developers
and strippers.
Facility D

Description
  Facility D manufactures mobile com-
munications equipment components  in
Florence, SC. The operation consists  of
a small metal-forming shop, prepaint and
painting lines, electroplating,  printed
circuit board manufacture, and a 30,000
gpd onsite wastewater treatment plant.
  Printed  circuit boards are  produced
using the  subtractive  technique  and
solvent-based  photoresists. Methylene
chloride resist stripper and 1,1,1-
trichloroethane (TCE)  developer are
continuously  recycled  in closed-loop
stills. The TCE developer wastes (Waste
Code F002) are recovered in a  DuPont
Riston SRS-120 solvent recovery still
(referrred  to as  the primary  still) and
returned to the developer line.  Until
recently,  all  still   bottoms from the
primary still were drummed and shipped
offsite  for reclamation at.a solvent
recycling facility. Facility D purchased a
Recyclene Industries  RX-35 solvent
recovery system (referred to  as the
secondary  still)  in  October  1985, to
further  remove TCE from still bottoms
onsite.
  The Recyclene  Industries  RX-35
solvent recovery system  is a batch
distillation system  with a  30-gallon
capacity, silicone oil immersion heated
stainless steel boiler,  a  non-contact,
water-cooled condenser, and a 10-gallon
temporary  storage  tank. The boiler  is
equipped with a vinyl liner inside a Teflon
bag. The Teflon bag provides temperature
resistance and the vinyl bag collects solid
residue, eliminating  boiler clean-out and
minimizing sludge generation  after
distillation. Two thermostats control the
temperature of the boiler and the vapor,
automatically shutting down the boiler
when all the solvent  has evaporated. The
maximum  operating temperature of the
still is 370°F, so recovery of solvents with
higher  boiling points  would  not  be
practical. Recovery of a 20- to 25-gallon
batch of still bottoms  requires approxi-
mately 90 minutes at Facility D, and four
to six batches are completed each day.


Results

  Evaluation of the  system consisted  of
the analysis of the  contaminated feed,
overhead product,  and  distillation  bot-
toms. Based on a  mass balance and
analytical data, the following conclusions
were made:
 • Purity of recovered solvent was 99.9$
   percent;

 • Total  solvent  recovery was 99.7(
   percent;

 • Still  bottoms contained 7.5 weigh
   percent 1,1,1-trichloroethane; and

 • Reduction in waste generation was
   97.5 percent.

  An  additional objective of the  studi
Was to evalute the  economics  of  the
batch solvent recovery unit. Annual cos
savings ($43,000) and  waste reductior
(110,602 gal) were calculated for Plan
D,  based on the first year of  RX-3E
operations. In addition, the investmen
payback period  for the  RX-35  was
calculated  considering  credit  fo
reclaimed  solvent and  reductions ir
waste transportation  and  disposal costs
The estimated  payback period was  7.2
months, given the  current  level of solven
reclamation. Thus, the low capital cos
of the unit and  the relatively high cost!
of virgin solvent  ($4.50/gal) favor  the
second-stage recovery of  TCE developei
still bottoms.
  There are several potential drawback!
to the implementation of RX-35 batcr
still that should be discussed. The firs
is that since the  bottoms product con
tains  7.5 weight  percent 1,1,1-TCE,  ii
remains classified as RCRA Waste Code
F002  (halogenated organic solvents) anc
is among those  solvent  wastes beinc.
considered under  the land disposal ban
Thus, while this technology significant^
reduces the volume  and  toxicity of the
solvent still  bottoms,  it  continues tc
generate a hazardous waste. A seconc
potential concern  is the accumulation 01
contaminants  and/or  breakdown  pro-
ducts. For example, 6.7 to 11.0 percent
concentrations  of carbon tetrachloride
were  found in process  feed  and  exi
streams, indicating  a  build-up of this
contaminant. Another  significant con
taminant found was 2-Butanone, which
represents 3.6  percent of the solven
waste feed  stream. It  could not be
determined  whether a  build-up  of 2
Butanone was occurring or if it is harmfu
to the system. However, its presence anc
effect on the solvent properties of 1,1,1
TCE should be considered.
  A final consideration in the implemen-
tation of any solvent recovery still is the
issue  of safety.  The unit at Plant D was
housed in a separate structure  and
provided with  adequate  ventilation  tc
minimize the risk of  exposure or explo-

-------
sion. The RX-35, according to the manu-
facturer, is safe for flammable materials,
and is rated for NFPA Class I, Division
I,  Group D environment. These safety
considerations should help to minimize
the  risk  of chronic exposure or danger
from explosion to personnel. Neverthe-
less, explosion risks from solvent recov-
ery operations should be carefully eval-
uated in  planning  the layout  and
installation of the unit.
Conclusions

  The findings of the waste minimization
case studies evaluated under this pro-
gram are presented in Table 2,  which
includes data collected by the facilities
and verified by sampling and laboratory
results. These results indicate that a good
variety of technologies exist to minimize
metals-containing and solvent  wastes
produced by the printed circuit board and
semiconductor industries. The technol-
ogies  discussed range  from  simple
changes in  treatment  system reagents
with nominal capital costs to large onsite
                        solvent reclamation facilities with signif-
                        icantly higher capital costs.
                          Four of the case studies investigated
                        under this program focused on technol-
                        ogies to reduce metal-plating rinsewater
                        sludges. The use of SBH as a substitute
                        for lime/ferrous sulfate was found to be
                        viable in one case and appeared to be
                        marginally  acceptable  in  another.  The
                        case study on carbon adsorption recovery
                        of plating bath wastes found that  this
                        technology  significantly  reduced  both
                        disposal  costs and waste volume.  The
                        case study of electrolytic recovery indi-
                        cated that this technology is highly waste
                        stream specific. An acid copper electro-
                        plating rinse is an ideal waste stream for
                        electrolytic  recovery.  However,  other
                        metal-bearing rinses, such as those from
                        solder (tin/lead)  plating or etching, are
                        not appropriate  for  use  of electrolytic
                        recovery. Electrolytic recovery units are,
                        however, generally inexpensive to pur-
                        chase and  can be used in many cases
                        to supplement an end-of-pipe treatment
                        process.
                          Two of the case studies presented in
                        this  Summary involve the recovery of
                                  spent halogenated solvents using batch
                                  distillation  units.  Both  case studies
                                  indicate that onsite solvent recovery is
                                  successful  from a technical and  an
                                  economic standpoint. In both cases, over
                                  95 percent of the  waste solvent was
                                  recovered  and reused  onsite. Solvent
                                  recovery appears to be a technology that
                                  could be applied to  a number of printed
                                  circuit board manufacturing facilities.
                                    The results of this project indicate that
                                  waste reduction can be achieved through
                                  the use of appropriate  technology, and
                                  it  can  be achieved with  significant
                                  reductions in cost. The case studies also
                                  indicate that  the  success of waste
                                  reduction  is in many cases waste stream
                                  specific. The technologies will not neces-
                                  sarily be successful in all cases. A slight
                                  variation between one waste stream and
                                  another may  make waste reduction
                                  either  technically or economically
                                  impractical. Therefore, successful waste
                                  reduction  is dependent on  a  thorough
                                  knowledge of waste  quantities and
                                  characteristics.
 Table 2.    Summary of Findings of Waste Reduction Case Studies
Facility Name
Facility A
Facility B
Facility C
Technology
Sodium borohydride reduction
Sodium borohydride reduction
Solvent batch distillation
Waste Reduction
Metals sludge
Metals sludge
Methylene chloride
Methyl chloroform
Freon
Annual Waste
Reduction
Achieved
a
962 tons
6. 152, 000 gal
Capital Costs
{$)
Nominal
Nominal
709.400
Projected
Annual Cost
Savings
(9)
b
/ 15.870
16.000,000
Facility D



Facility E


Facility F
2-Stage solvent distillation
Carbon adsorption
plating bath reclamation

Agmet electrolytic
recovery unit
1.1,1 • Trichloroethane
Resist developer
still bottoms

Plating bath wastes
(metals sludge)

Metals sludge
10.625 gal



10.600 gal


32 tons
26.150



 9,200


30.350
 43.105



 57.267


(10.685)°
'Not quantifiable, but a significant waste reduction was realized.
"Not demonstrated during testing.
c( ) indicates negative value.

-------
      T. Nunno. S. Palmer, M. Arienti, and M. Breton are with Alliance Technologies
        Corporation, Bedford, MA 01730.
      Harry M. Freeman is the EPA Project Officer (see below).
      The complete report, entitled "Waste Minimization in the Printed Circuit Board
        Industry—Case Study,"(Order No. PB88-161 575/AS; Cost $19.95, subject
        to change) will be available only from:
              National Technical Information Service
              5285 Port Royal Road
              Springfield, VA22161
              Telephone: 703-487-4650
      The EPA Officer can be contacted at:
              Hazardous Waste Engineering Research Laboratory
              U.S. Environmental Protection Agency
              Cincinnati,  OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
     BULK RATE
POSTAGE & FEES PA
        EPA
  PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/S2-88/008
                                                  QC01961    HWER  GN

                                                  LIBRARY  REGION  V

                                                  230  S  DEARBORN  ST
                                                  CMCAGO
                                                                                 60604
                                                                              
-------