United States
                 Environmental Protection
                 Agency
Risk Reduction
Engineering Laboratory
Cincinnati, OH 45268
                 Research and Development
EPA/600/SR-93/191     July 1994
EPA        Project Summary
                 Minimizing  Pollution  in  Cleaning
                 and  Degreasing  Operations
                S. Garry Howell and A.R. Tarrer
                  As  part of an effort to  prevent or
                minimize waste, the U.S. Air F:orce stud-
                ied ways to extend the useful life of
                cleaning solvents, particularly 1,1,1,-
                trichloroethane (TCA). The use of addi-
                tives such as acid acceptors, oxidation
                inhibitors, and metal stabilizers are ex-
                plained as solvent extenders. Because
                knowing the condition of the solvent is
                important  and because visual inspec-
                tion is  unreliable,  several  tests (acid
                acceptance value [AAV], viscosity, wa-
                ter content, visible absorbance, non-
                volatile matter  [NVM],  and electrical
                conductivity) are suggested to analyze
                for stabilizer content and to estimate
                the remaining useful life.
                  When these tests indicate there is
                some useful life left, filtration, replen-
                ishing stabilizers,  distillation, ultrafil-
                tration, and more efficient operation are
                explained  as means to return an al-
                most spent solvent to useful life.
                   This Project Summary was developed
                by EPA's  Risk Reduction  Engineering
                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
                   U.S. Air Force overhaul  and mainte-
                nance operations use large volumes of
                solvents for vapor cleaning and degreasing
                aircraft, including engine parts, gyroscopes,
                guidance systems, and electronic compo-
                nents. A major portion of these solvents
                are chlorinated compounds such as TCA,
                methylene chloride,  and perchloroethylene
(perc).  During normal use, a portion of
these materials are lost through evapora-
tion and are thought to contribute to strato-
spheric ozone depletion.  Further, the
cleaning solvents  are  discarded after
reaching a point where grease and dirt
build up  in  cleaning baths. Reducing
evaporation  losses and the amount of
spent solvent to be disposed would save
the value  of the  solvent, requiring fewer
purchases of fresh solvent, as well as
disposal costs.
  As part of the implementation of its
waste minimization program, the U.S. Air
Force is trying to find the  most efficient,
safe, and economical methods to  extend
the useful life of, or to recycle, cleaning
solvents, particularly TCA.  Other, though
less widely used, solvents considered for
recycling are  methylene chloride, perc,
CFC-113 (trichlorotrifluoroethane) and PD-
680 (a  petroleum fraction  similar  to the
well-known Stoddard solvent). As  part of
this research project, Auburn University,
in  cooperation with major solvent  manu-
facturers, developed a program to develop
fairly simple methods based on key physi-
cal and chemical properties so that users
could get the longest useful life from their
cleaning baths. Then, after testing deter-
mines that the solvent is no longer useable,
two options are offered. Either critical ad-
ditives may be replenished or the solvent
may be recycled by distillation, filtration,
or other means.
  Vapor degreasing is a common indus-
trial process in which the parts  to  be
cleaned are suspended in the vapor space
above a boiling bath of solvent. The most
common solvents for such operations are
                                                                    Printed on Recycled Paper

-------
TCA and perc. Because the vapors are
much heavier than air, a properly designed
and operated bath will confine most of the
vapors in the "head space" and very little
will be lost to the atmosphere. During the
course of operation, however, solvents in
a Vapor degreaser undergo both chemical
and physical changes and solids (such as
metal chips, dirt,  or paint particles) build
up in the boiling sump liquid. The oils and
greases in solution  in the  liquid  phase
may build up to the point where the vis-
cosity is increased and  heat  transfer is
seriously impaired. In electrically heated
baths, heater sheath temperatures  may
increase and cause failure. In addition, as
the concentration of lighter oils increases,
they will appear  in the  vapors  and will
leave a noticeable residue on the "cleaned"
parts. A chemical analysis aimed only at
the composition of the solvent itself  may
show that it is still useable,  but  its  effi-
ciency has been severely impaired by dis-
solved  material. At  least part  of  this
efficiency may be restored by the physical
methods described in this report.
  A number of previous  studies  have
shown that efficient operation of degreaser
baths can also reduce solvent usage and
minimize emissions.  A separate,  but re-
lated, part of this research study  investi-
gated the  operation of  several  vapor
degreasing units at Tinker AFB in Okla-
homa. A survey,  made by a representa-
tive of the Dow Chemical Co.,  resulted in
recommendations for changes which would
result in reduced solvent loss.*

Additives to Maximize Solvent
Life
  As  partially chlorinated  solvents, TCA,
methylene chloride, and perc have a ten-
dency to decompose after prolonged heat-
ing. This is manifested by a chlorine atom
breaking off the hydrocarbon backbone to
form  hydrogen chloride  (HCI). The  HCI
then combines with any water present to
form  hydrochloric acid which  corrodes
metals.  (Fluorine, as in CFC-113, is more
tightly bound, and less  likely to be  re-
leased). The  HCI itself promotes further
dechlorination; this is inhibited by addi-
tives called acid acceptors, which are de-
signed to combine with and neutralize HCI.
  In addition to  dechlorination, relatively
high temperatures can lead to oxidation of
the solvents.  Many halogenated solvents
• Momion of trade names or commercial products does
 not constitute endorsement or recommendation for
 use.
are relatively resistant to attack by oxygen
until a halogen (chlorine, fluorine, etc.) is
abstracted.  Removal of the chlorine  (or
fluorine as in the case of CFC-113) and
an  adjacent hydrogen  leaves a "double
bond," which is susceptible to oxidation,
particularly in the presence of metals such
as copper or vanadium. Perc and trichlo-
roethylene  have these double bonds  al-
ready in place and, consequently, are more
susceptible to  oxidation. Another class of
stabilizers called oxidation inhibitors,  or
antioxidants,  may be added to prevent
this.
  The interaction between acid and metal
can etch or corrode a metal surface. Con-
versely, some metals, such as copper and
vanadium,  promote oxidation  which  re-
quires some way to protect the metal from
the solvent. Metal stabilizers can do this
by forming a protective film that retards
attack by acids, while deactivating the cata-
lytic properties of the surface.
   The concentration of these metal stabi-
lizers in  halogenated solvents is critical,
particularly the ratio of  metal stabilizer to
acid acceptor. Although other stabilizers
decompose chemically or react with acids
and  become depleted,  metal  stabilizers
are intentionally chosen to  be somewhat
volatile so they maintain some  presence
in the  vapor phase. An  almost monomo-
lecular layer is formed  on the  metal sur-
faces so, as the parts are withdrawn, the
adherent stabilizer is consumed.

Methods to Determine a
Solvent's Useful Life
  The full report lists typical acid accep-
tors, oxidation  inhibitors, and metal stabi-
lizers. Visual  inspection  alone cannot
determine the  concentration of these ad-
ditives in the baths nor can it identify other
possible deleterious conditions. Visual  in-
spection  is unreliable as a means of tell-
ing the condition of degreasing solvents:
a perfectly clear bath could be deficient in
inhibitor and a badly discolored one can
still be well stabilized. Although the most
reliable method of determining the  levels
and types  of  stabilizers in  halogenated
solvents is by  gas chromatography/mass
spectrometry (GC/MS), its expense makes
it impractical for most users. There are,
however, several reliable methods for ana-
lyzing for stabilizers and for estimating the
remaining useful life of a degreasing bath.
When  in  doubt, two or more of the tests
listed  below should be used.
Acid Acceptance Value
  The acid  acceptance value (AAV), the
most important parameter, is  an accurate
measure of stabilizer content. It  is nor-
mally measured by a simple titration (see
Figure 1).

Water Content
  Water in halogenated solvents is detri-
mental in two respects: it promotes corro-
sion at metal/water interfaces and it causes
hydrolysis of the halogen-carbon bond re-
leasing  chlorine or fluorine.  The corre-
sponding corrosive acids,  HCI and HF,
are then  formed.

Visible Absorbance
  Color  changes can be quantified with
the use  of a spectrophotometer to mea-
sure light transmission at a set  wavelength.

Non-Volatile Matter
  Evaporating a solvent sample  over  a
steam bath and weighing the residue indi-
cates the content of oil, grease, and dirt.

Electrical Conductivity
  Ionic materials and water decrease the
electrical conductivity of halogenated sol-
vents. Conductivity  measurements are
relatively easy to make with  inexpensive
meters.

Viscosity
  Viscosity is a good indicator  of dissolved
oil and grease content.  Several methods
may be  used for its determination (e.g.,
an  inexpensive, simple, falling ball vis-
cometer).

Returning the Solvent to a
Useable Condition
  As a  degreasing or  cleaning  bath  is
used and dissolved oil  and grease, dirt,
metal particles,  and water accumulate,
some steps  can  be taken to return it to  a
useable  condition. One or more  of the
above tests might indicate that there is
some useful life left and that  filtering the
solvent and monitoring changes in its prop-
erties may be a feasible alternative before
prematurely  discarding the solvent. In an
experimental filtration  setup,  the solvent
was pumped from a stirred tank through  a
series of three filters: the first, a  bag  to
remove very coarse (25 |o.m) particles, then
through 1.0 and 0.5 (xm filters  (see Figure
2). Pumping  continued until the tank's con-
tents were passed through the filters ap-
proximately five times, a single cycle was
about 10 min. Samples  were  taken after

-------
                                                                                                     Solvent condition
                                                                                                     borderline
                                                                                                     - Check daily


                                                                                                     Solvent condition
                                                                                                     unacceptable
                                                                                                     - Discard or distill for reuse
         4

         'S
          i
10.0

9.0

8.0

7.0

60
K n











^

0.2









^

8









^

O.i








^^


'4







^^
""


O.i







^-^"



'0






^^^




0.1





^^





6





---•





0.;




^^'






2



^^,
^






O.t
'/////.
fffff
sssss
j&rft:
fftff.
ffffS:
fffff.
SftfS.

ffffS
fftff
fftfj
ffffj
18
'////,
Jfjfjfjfjf
fWfss
*////,
ffffj
SfSfS,
•fff/S
ffJSf

ft/tf
ffttf
ffftf
•fffff
O.I







,



34

"^"








--
0
                                              Acid acceptance as weight percent sodium hydroxide
 Figure 1.  TCA calibration chart.
 Feed
 tank
                                                               Circulate for 5 cycle times
                                                           Bag
                                                           filter
1' \m
filter
0.5/MI
filter
                                                                           Filtration system
Figure 2.   Filtration scheme for reclaiming solvents.

-------
each cycle to monitor the condition of the
solvent. Filters were cleaned when the
differential pressure across them reached
15 psi. Some test results from these filtra-
tion experiments are given below.

Filtration of 1,1,1-TCA

Acid Acceptance Value
  The  presence  of acid  acceptors (as
measured by AAV) was not affected by
filtration; five passes through the system
left the AAV almost unchanged at  three
times the  minimum value. This is in con-
trast to distillation or adsorption on car-
bon, both  of which decrease AAV.

Water Content
  As stated before, the presence of  water
Is deleterious to the solvent and to  parts
being cleaned. If the starting content was
above  0.5%, repeated filtration cycles ac-
tually reduced it. When the initial concen-
trations were less than 0.5%, water content
remained  almost  unchanged despite the
number of passes through the filtration
system. The reason for this is unclear, but
may possibly be due to the fact that  water
solubility in TCA is  about 500-600 ppm
(0.05-0.06%),  and above  this  level, the
water  exists as a dispersion.  At higher
water contents, filtration would cause  some
coalescence of droplets, causing them to
float to the surface of  the feed tank and
evaporate. Whatever the mechanism, fil-
tration  done in this manner does remove
some of the water.

Visible Absorbance
  New TCA is water white, and its clarity
is decreased by either dissolved materials
or suspended particles. A portable  spec-
trophotometer was  used to measure the
light (580 nm) absorbed or scattered by
the solvent samples; this is reported as
percent transmittance. Although  water
droplets would always be expected to have
some  effect on the light  transmitted (or
absorbed), the presence of water soluble
or wettable impurities  can affect absor-
bance. Dissolved colored waxes, oils, and
grease as well as suspended particles of
dirt, etc. also lower light transmission. Ab-
sorbance measurements should always be
used in tandem with at least one  other
test, as it is possible that  a clear,  clean
looking solvent could be  completely de-
void of acid acceptors or other stabilizers.

Non-Volatile Matter
  Weighed samples of contaminated sol-
vent were evaporated under vacuum (un-
specified) at 105 +/- 5°C.  The  NVM
remained  almost constant during all the
runs. Measuring the solids content  by fil-
tering the samples, then drying the filter
paper at 60°C gave  higher values than
the NVM; this was attributed to volatiliza-
tion of lighter oils in the vacuum oven.
Evaporating the samples on a steam table
(the preferred method) would probably give
higher results than the filter paper method,
particularly if dissolved solids were present.

Electrical Conductivity
  Electrical conductivity tests made after
every pass through the filtration  system
showed very little variation, however com-
parison  to  virgin solvent  indicated con-
tamination by some  conductive material.
The small amount of water dissolved  in
the TCA (the literature values are 500-
600 ppm) could  carry minute amounts  of
HCI or salts,  which would raise the con-
ductivity well  above the 10'10  mhos/cm  of
the new material. Filtration, as stated pre-
viously,  will not remove dissolved materi-
als, so little or no effect on  conductivity
would be expected.

CFC-113 Filtration Tests to
Remove Water and Solids
  Tests were conducted on CFC-113  to
determine if simple filtration is effective  in
removing water  and  solids, thereby ex-
tending  the  useful  life of the  solvent.
Samples of used CFC-113 (trichlorotrifl-
uoroethane) with high water and high sol-
ids contents were reclaimed by filtration  at
Warner Robins Air Logistics Center (ALC).
Water content was reduced by about 90%,
to very nearly the solubility level of water
in CFC-113, which is 0.009%. Solids were
almost completely removed, indicating that
most  solids were particles > 0.05 urn.
Since CFC-113 is primarily used for clean-
ing delicate precision parts, base person-
nel thought that  the filtered solvent would
be usable for preliminary  cleaning, if fol-
lowed by a rinse with virgin solvent.
Replenishing Stabilizers
  If the AAV or GC/MS analysis indicates
that  a solvent is deficient in stabilizer(s)
but still acceptable as a cleaner, two op-
tions  can be used  to  restabilize it. The
first, easiest, and  safest is to buy new
solvent containing  higher-than-usual lev-
els  of stabilizer and add this to the old
bath in sufficient quantity to raise the sta-
bilizer to a safe level. The second option,
which requires  more skill and training,  is
to buy stabilizers (often including acid ac-
ceptor, metal stabilizer,  and oxidation in-
hibitor) and add them as needed. Careful
analysis is  required to determine the
amount to be added, and safety precau-
tions  must be  strictly  observed,  as the
chemicals (see the table below for some
examples) are flammable, toxic, and pos-
sibly carcinogenic. For these reasons, only
those knowledgeable in and adept at han-
dling  chemicals should  attempt to make
up their own stabilizer packages.

Distillation
  If, after filtration, the level of dissolved
oils and grease is too  high, one of the
best options for reclaiming solvents  of all
types is distillation. Properly done, distilla-
tion can  yield the base solvent in almost
virgin condition. Petroleum solvents, such
as Stoddard Solvent, or PD 680,  might
require vacuum to lower their boiling points,
but otherwise they are unaffected by the
process. TCA,  methylene  chloride,  and
perc all boil at  250°F or below and may
easily be handled in either steam or elec-
trically heated  stills at atmospheric pres-
sure. The major disadvantage of distilling
halogenated solvents is that the stabiliz-
ers could be partially retained in the still
bottoms  and the condensate  might  be
understabilized.  Therefore redistilled sol-
vents should be analyzed to confirm sta-
bilizer levels.
 Table 1.  Typical Inhibitor/Stabilizers and Their Functions

   Function                                  Compound
   Acid acceptor

   Acid acceptor

   Metal stabilizer'*

   Oxidation inhibitor
  Butylene oxide
  (1,2-epoxybutane)
  Epichlorohydrin
  (1,2-epoxy-3-chloropropane)
  1,4-Dioxane
  (1,4-dioxacyclohexane)
  /one/™", BHT
  (2,6-di-t-butyl-p-cresol)
 * Flammable and classed as possible carcinogen.
 + Fatty acid amines are sometimes used; these give some rust inhibition after parts are
   exposed to air.
 ++ tono/ and BHT (Shell Chemical Co.) are illustrative of a broad class of phenolic antioxidants;
   they may not be suitable for all types of degreasing baths.

-------
  Many stills have been installed, tried a
few times, and  abandoned presumably
because they are difficult to operate. These
could quite likely be reactivated with more
thorough operator training and retrofitting
with  modern automatic control systems.
An excellent example of this is the opera-
tion at Warner Robins ALC, where distilla-
tion  has been  successfully carried  out
since  1982.  In 1987,  an  estimated
$800,000 was saved on CFC-113 alone.
Personnel at the base found that  redis-
tilled TCA usually retains enough acid ac-
ceptors to  be  safely used,  but  as a
precaution, the personnel keep some 1,2-
butylene oxide on hand for addition  to the
reclaimed solvent.

Ultrafiltration
  Conventional filtration will normally re-
move particles  of  > 0.5  u.m. When par-
ticles smaller than 1 urn are to be removed,
polymeric membranes may be used in-
stead of the  solid media of ordinary filters.
Some newer membranes are resistant to
many solvents and are capable of retain-
ing molecules  larger  than  150 to 200
daltons (molecular weight). This means
that  Stoddard  solvent (ca. 150 to 160
daltons) or TCA (133.4 daltons) would pass
through, but dissolved  oils and greases
would be retained. The higher molecular
weight  inhibitors  such  as  butoxy-
methyloxirane  (130.2 daltons) would also
pass through the membrane. Membrane
filters should always be  preceded by a
conventional filter to minimize fouling. De-
velopment of these solvent-resistant mem-
branes, as well as those resistant to acids
and bases, is  continuing, and membrane
separations are expected to replace distil-
lation in many  instances.

Minimizing  Pollution by More
Efficient Operation
  The period between rejuvenation or re-
cycling solvents can be greatly increased
by more careful operation of degreasing
operations. A technical and economic fea-
sibility investigation was made of several
methods of reducing solvent use and am-
bient vapor  concentrations from a TCA
vapor degreaser at Wright Patterson AFB.
The recommendations resulted in $25,2007
yr savings in solvent costs with a payback
period of 0.6 yr. In addition, almost 4000
gallons of TCA did not escape to the at-
mosphere annually.
  A  similar survey  of  the  various
degreasing operations at Tinker ALC was
made by a representative of Dow Chemi-
cal Co., a major solvent supplier; follow-
ing those recommendations should save
about $68,000/yr. Although the recommen-
dations referred to specific units at Tinker,
they  are applicable to many operations.
The complete survey is too lengthy to be
included  in this summary, but  an edited
version is included in the report summa-
rized here.

Conclusion
  The life of solvents employed to clean
and degrease aircraft engine parts can be
successfully extended by using additives
such  as  acid acceptors, oxidation inhibi-
tors, and metal stabilizers. When the sol-
vent  is  no longer  useful,  it can be
effectively rejuvenated by filtration, distil-
lation, and more efficient operation.
  The full report was submitted  in fulfill-
ment of Contract No. EPA CR816219 by
Auburn University under the partial spon-
sorship of the U.S. Environmental Protec-
tion Agency; the U.S. Air Force supplied
most of the funding.

-------

-------

-------
 S.G. Howellis with the U.S. Environmental Protection Agency and A.R. Tarrer
   Is with Auburn University, Auburn, AL 36849-5127
 Joseph D. Wander was the Air Force Project Officer (AFCESA, TyndallAFB);
 Mary Ann Curran is the EPA Project Officer (see below).
 Ths complete report, entitled "Minimizing Pollution in Cleaning and Degreasing
     Operations" (Order No. AD-A-277094; Cost: $17.50, 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:
         Risk Reduction Engineering Laboratory
         U.S. Environmental Protection Agency
         Cincinnati, OH 45268
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268

Official Business
Penalty for Private Use
$300
      BULK RATE
POSTAGE & FEES PAID
         EPA
   PERMIT No. G-35
EPA/600/SR-93/191

-------