-------
TABLE 3-7.  FOUNDRY LOCATION WITH RESPECT TO ATTAINMENT STATUS
.
Attainment


State
Iowa
Ohio
Oklahoma
Total
No. of
ft
foundries
34
103
6
Pennsylvania 63
Vermont
Washington
Total
9
16
231
Not meet
primary
No.
7
50
5
30
0
12
104
%
20.6
48.5
83.3
47.6
0
75.0
45.0
status*3
Not meet
secondary
No.
-7
42
0
0
6
_2
57
%
20.6
40.7
0
0
66.7
12.5
24.7
Unknown
No. %
4 11
0 0
0 0
11 17
0
o. _9.
15 6


Better
than
national
standard
No.
.8 15
11
1
.5 22
3
	 _2
.5 54
%
44.1
10.7
16.7
34.9
33.3
12.5
23.4

a Source:
b Source:
NEDS Printout,
June 1979.
Federal Register,
Vol. 43,
No. 43
, March
3, 1978,
Part II.
pp.
      8962-46019.  Federal Register, Vol. 43, No. 194, October 5,
      1978.  pp. 45993-46019.
                              30

-------
     Davis et al. developed estimates of industry-wide production from vari-
ous foundry size classifications.2  They used foundry employment as a basis
for the following size categories:

     Large     Over 250 employees
     Medium    50 to 249 employees
     Small     49 or less employees

The Penton  capacity  data and the assumption that large foundries melt for
16 h/day, medium foundries 8 h/day, and small foundries 4 h/day were used
to calculate industry-wide daily capacities for 12 foundry size/furnace-type
categories.1  The results are shown in Table 3-8.

     While these data  are  not particularly useful in developing specific
compliance s-trategies, two  general observations can be made.  The data in
Table 3-8 indicate that over 65% of foundry production capacity is centered
in large foundries, which gives further support to foundries being signifi-
cant sources of  air  pollution.   The data also indicate that almost 75% of
foundry melt capacity is from cupolas, further indicating that these furnaces
will continue to be a compliance problem.

     Obviously, the data most useful in developing specific compliance strat-
egies are actual production  rates  of  individual  foundries.  However, since
these data may  sometimes be difficult to obtain and since employment data
are generally available from industrial publications, production rates were
examined as a  function of employment.  Using the size classifications and
work rates  from  Davis  et al., Penton capacity data  (see Appendix D) were
used to calculate daily production capacity for each U.S.  foundry.   The aver-
age daily production and the range of production for the various size found-
ries (as defined by employment on p. 38) were found to be:
Size (based on
 employment)

    Large

    Medium

    Small
Average production
	(tons/day)	

       820

       100

        20
Range of production
    (tons/day)	

    10.2-9,670

     1.5-1,600

     0.2-138
As can be seen from the extreme range of capacity shown in the table above,
foundry employment  can not be used as an indicator of production capacity.
                                   31

-------
    H
CO
en


I
    CO

    0)
•O   CU  P-i x-s
 cu   p. cd  ^
4J      13  cd
 cd   ^     *o
 g   4J  oo*-.
•H  iH  C  CO
4J   O  iH  F3
 co   cd  .M  o
pq   p, }j  4J
     cd  o  ^
                     cd
                                     CO
                                     a
                                 P-  o
                                 CO   4J
                                o  >-*
                                o
                 CO
                 o  CD
                 •H  0)
                 W  0)
                 TJ  !*
                 c  o
                 3  rH
                 o  a
                         cu  o.  cd
        T3  cd

     $  00^
 •S  >H  C  co
 4J  CJ -H  (3
 en  ca A!  o
 W  P.  M  4J
                  0)
                  N  CTl
                  a)
                                     CO
                                     C
                                 a,  o
                                 CO
                                 O
 CU
 o   co
3%

 si
T3   i
*O   Q)
 §0
                  0)
                  00  a]
•o   H   (3   CO
                         w   a  t-i   c
                              a.  i-i
                                  o   co
         cd  4J
     IH  oo  cu
      o  s   a
         •H   cd
      cu 4-i   (3
      D. rH   VI
     >,  cu   3
                                              m
                                              oo
                             
-------
                                REFERENCES
1.   Penton Computer Printout of Gray Iron Foundries in the United
     States.  The Penton Publishing  Company,  Cleveland, Ohio.  March 1974.

2.   Davis, J. A.,  E.  E.  Fletcher, R. L.  Wenk,  and A. R. Elsea.  Final
     Report on Screening Study  on Cupolas and Electrical Furnaces in Gray
     Iron Foundries.  U.S.  Environmental Protection Agency Contract No.  68-
     01-0611,  Task 8.  August 15, 1975.

3.   Metal Casting Industry Census Guide:  November 1978.  Foundry
     Management and Technology.   April 1979.

4.   U.S. Department of Commerce Industry and Trade Administration,
     1979.  U.S.  Industrial Outlooks, January 1979.  pp. 178,  179.

5.   A.  T.  Kearney  Company.   Systems Analysis of Emissions and Emissions
     Control in the  Iron Foundry Industry.  U.S. Environmental Protection
     Agency PB 198 348, February 1971.

6.   Foundry  Equipment Inventory  and Buying Plans.  Foundry Management
     and Technology, April  1979.  pp. 34-60.

7.   City County Data Book.

8.   Federal Register, Vol. 43,  No. 43,  March 3,  1978, Part II.  pp.
     8962-46019.

9.   Federal  Register, Vol.  43, No. 194, October  5,  1978.  pp.  45993-
     46019.
                                   33
-------
               4.0  FERROUS FOUNDRY PROCESSES AND EMISSIONS


     A ferrous foundry is composed of numerous unit operations, many of which
have the potential for the emission of- gaseous and/or particulate pollutants
to the atmosphere.  A basic understanding of these foundry operations and
their associated  emissions  problems  is  a prerequisite to any  analysis of
the compliance problems associated with ferrous foundries.

     This section briefly describes the ferrous foundry process and summarizes
available information on foundry emissions.  The discussion is divided into
three sections:   (a) description of ferrous foundry processes; (b) identifi-
cation of emissions sources;  and (c) inventory of particulate emissions.
More detailed information can be found in Appendix A, "Description of Ferrous
Foundry Processes," and Appendix B, "Quantification of Particulate Emissions
for Major Foundry Emissions Sources."

4.1  DESCRIPTION OF FOUNDRY PROCESSES

     A ferrous  foundry processes various  grades of iron and steel scrap to
make cast iron and steel products.  The  four basic  operations present in
all foundries are raw materials storage and handling, metal melting, pouring
of the molten metal into some type of mold, and removal of solid castings
from the mold.  Other operations which occur in many foundries are prepara-
tion and assembly of  sand molds  and  cores,  mold cooling, shakeout,  casting
cleaning and finishing, sand handling and preparation, and hot metal inocu-
lation.

     Six basic operating areas can be found in the typical fetrous foundry.
These are:

     •    Core and mold preparation

          Furnace charge preparation

          Pattern making

          Melting and casting

     •    Cleaning and finishing

          Sand-handling system

Since pattern-making  operations  are not significant sources of emissions,
this area was not included in the study.  A flow diagram for the other five
operations  in a "typical"  foundry is shown in Figure 4-1.   The paragraphs
below describe  the operations that are found in each of these areas.  De-
tails of the individual operations are provided in Appendix A.

                                     34
-------










,
'












1


1

























f


























— »































h
J

4
a






i
i

t

Purchased



















j[.







i
!
»







p
!
i

i

8-
£










i
,<

t
S
J


i
-








-»























p
5-









j
S
:!
1





























SL









1
H
= S
s u
1













*
<
1
z




8-5
ji


•§
a
S




f
j 1

> 1.


F
•s
J *
6 1
i>
s ^
l|
""*
i h


1
g>
I I
§ 2


£

01
-• r--, -
U
S « Si
s 3 |
I







T
5





























?








•o
o
o
-5
I'
3
u



t
_i
u







-rs
c

1
















••-





1
g
3
£
i




3











"*~















	








a
1
g
i


1

•p


!


1
,
U
I
.

s

s
•*

,
1





u

-






c
^

_
o
in
E
to








1

g
i
fe
0)
VI
(

O)

"g
U

f
'

a



»
*>

— -






~ S .5 J
•2 "" !i 2 s ;|
r. JL ^u i
t 1
u

0 1
1 I s.
fc s s •• j g*
1 1— u g 3 g
o ,5 cd
U S S

• - *H
}i a


T ji
n
N J
31 2
E 	 	 » 01
i ill"
01 § ™ °* *
x U "t
"? A 5-1
i 	 T a
in 60

Jl
52
.


5 c »

ee i/i
_ 2
,,„„ 	 	 ^ T _g
... 	 ^ ^ -g
-— 1 M
t i




35
-------
     The basic raw materials which enter the foundry process are:   (a) sand
and binders for core and mold preparation;  (b) metallic materials including
iron and steel scrap, borings and turnings, limited quantitites of pig iron,
and foundry returns;  (c)  coke  for cupola  fuel;  (d)  fluxing agents;  and  (e)
limited quantities of inoculants and alloying agents.  The  sand is generally
stored in closed silos, and the other materials are stored  under cover when-
ever possible to prevent material degradation.

     Upon leaving the storage area, the sand and binders go to the core and
mold preparation  area.   The mold provides  the basic  exterior  form  of the
casting, and  cores  are used to form  the  interior of the castings (e.g.,
cylinders in an engine block).  The most common type of mold used in ferrous
foundries is  the  green sand mold.  The sand  is prepared by mixing  silica
sand, water, bentonite clay, and binder materials such as cereal binder (de-
rived principally from corn flour) and sea  coal (finely ground coal commonly
mixed with foundry sands) in a muller.  The damp sand mixture is packed around
patterns in  one of  a variety of molding machines  to form the two halves of
the mold.   A typical green sand mold is shown in Figure A-2 (Appendix A).
The cores are then placed in the mold, and  the molten metal is poured while
the mold is  still moist.  The  cores  are  generally prepared by a chemical
binding process and contain no moisture when the metal is poured.

     The production of  other  types of molds, such as dry  sand molds, pit
molds, and permanent molds,  is described in Appendix A.   Another type of
mold, the chemically bonded mold, is produced in the same manner as the chem-
ically bonded cores described below.

     As with molds, cores are produced by a variety of processes.  The tra-
ditional method uses oil and cereal binders to maintain the core shape.   In
this method  sand,  core  oil,  and cereal are mixed; the core is shaped; and
finally, the  core  is  baked in an oven to solidify it.  The oil/core  oven
method is being replaced by chemically bonded sand processes.   These pro-
cesses work  through  thermal  setting (hot box and shell  core  systems) or
through catalytic reactions (cold box or no-bake processes).  These methods
are described in detail in Appendix A.

     The metallics, coke,  and  fluxing materials are removed from storage
and prepared  for  charging to the  furnace.   Coke and fluxing agents undergo
minimal processing prior to charging.   The amount of metal processed is  de-
pendent upon the type of scrap received and the type of melting furnace  used.
If the scrap received by the foundry is too large to be charged to  the fur-
nace, the size is reduced by breaking, abrasive cutting, or torch cutting.
Unless an electric induction furnace is used,  sizing is the only preparation
needed.  If an electric  induction furnace  is used as the primary melting
unit, however, the scrap must be  clean and  dry, or explosions will result.
Acceptable scrap is obtained by purchasing high quality scrap and  storing
it in an enclosed area  or by preheating the scrap before it is charged  to
the induction furnace.

     The melting area is  the  most visible  area in the foundry.  All melt
shops will include a melting  furnace and some type  of process  for  pouring
the molten iron into a mold and subsequent cooling of the casting.   In addi-
tion,  some foundries have-duplexing furnaces (primarily associated  with  mal-
leable iron)  and  inoculation  stations  (associated  with ductile  iron).

                                     36
-------
     Four types of  furnaces  are used as primary melting units  in  ferrous
foundries.   Currently 75% of all molten iron "is produced in cupola furnaces,
17% in electric arc furnaces,  7%  in  electric induction  furnaces, and  1%  in
other types of  furnaces  such as gas- or oil-fired reverberatory furnaces.
These four types of furnaces are described in Appendix A.   Diagrams of these
furnaces can be found in Figures A-6 through A-10.

     Each of the four types  of furnaces receives scrap  metal and heats the
metal until desired physical and chemical properties are achieved.  After
the melt is completed, the metal is tapped from the furnace into a hot metal
transfer ladle.  The metal may then  be transferred to a holding furnace, a
duplexing furnace, an inoculation station, or directly  to the pouring sta-
tion.  In some foundries the metal is transferred to a mixing ladle or fore-
hearth before going to the pouring ladle.

     A holding  furnace is an electric arc or an electric induction furnace
which is used to maintain the metal in the proper condition until the foundry
is ready to pour.   A duplexing furnace is an electric furnace which is used
in malleable iron production to increase the  temperature of the metal in
the absence of  slag.   Duplexing is necessary when a cupola is used as the
primary melting unit.   Iron  inoculation is  the addition of magnesium (or
other inoculants) to  gray iron to produce ductile iron.  It is generally
accomplished in the ladle by one of several methods described in Appendix A
(see Figure A-ll).

     After the  above  steps  have been completed, the molten metal is ready
to be poured into the mold.  The,pouring method used is dependent  upon the
type of mold,  the  size of the casting,  and the degree of mechanization in
the foundry.  Various types  include  permanent mold pouring; floor  pouring,
in which the ladles are moved to stationary molds; and pouring stations,  in
which the ladle is held at one place and the molds are moved to the station
on conveyors.   After pouring is completed, the mold and casting are cooled
until the casting is ready for removal from the mold.

     The final processing area is cleaning and finishing the casting in prep-
aration for shipping.   Cleaning and finishing are accomplished in several
steps by a variety of methods.  The first step is to remove the casting from
the mold.  If a sand mold is used, this process is termed shakeout.  Shake-
out is accomplished in a variety of ways, but the most typical is the use
of a vibrating  or  rotating screen to remove the  sand from the castings.
The castings are then sorted, and the sprues, gates, and risers are removed.
Depending upon the type and size of casting, this may be accomplished through
impaction, abrasive cutting, band cutting, or torch cutoff  (including air-
carbon-arc cutting).  After  the appendages have been removed, the surface
is cleaned by processes such as shot and sand blasting, tumbling, and vari-
ous types  of grinding.  Finally,  especially  with malleable  iron and steel,
the casting is heat treated and final forming is completed before shipping.

     The final  area of operation which can be found in all foundries that
use some type  of  sand molding is the sand-handling system.  Sand handling
comprises a number  of  transfer and conditioning operations  which vary sig-
nificantly among foundries.  The most important processes from "the standpoint
                                     37
-------
of  compliance are those involving the return sand between the shakeout and
the  sand mixer or muller.  In  a  mechanized foundry transfer is generally
accomplished  by conveyor.  In smaller, less  mechanized  foundries much of
the  sand transfer may be accomplished manually  and by front-end  loader.

     In  reviewing the above description  of the ferrous foundry relative to
compliance with air pollution regulations,  it  is  important to note a "typical"
ferrous  foundry does  not exist.  Any given foundry consists of a  sequence
of  unit  operations which can be  accomplished  in  a  variety of ways.  Thus,
an agency charged with monitoring the compliance  of ferrous foundries  should
be familiar with  the  available operations,  the relative impact of  these opera-
tions  on air quality, and  the limitations  in  the application of different
methods  of operation.

     It  is not within the scope of this  study to provide  detailed descrip-
tions  of every available iron foundry process.  But some  additional  detail
is provided  in Appendix A.   For the reader who desires further information
about  foundries in general  or more  details  about  specific processes, excel-
lent information  can  be found in  References 1  to  9.

4.2  IDENTIFICATION AND CHARACTERIZATION OF EMISSIONS  SOURCES

     The characteristics  and quantity of emissions  from the various processes
are apparent  factors  which  affect the compliance  of ferrous foundries.  This
section  describes  those characteristics of  emissions sources which can affect
a foundry's compliance with,various regulations.  The  major characteristics
examined include  the  types  of pollutants emitted  by the source, the type of
emissions source,  e.g., ducted or stack, process  fugitive, or open fugitive,
and emissions stream  properties,  such as temperature and moisture that have
an  effect on the controllability of  the source.  Emissions quantities are
discussed in  Section  4.3.

     Each of  the  processes described in Section  4.1 is a potential source
of gaseous or particulate emissions.  Gaseous pollutants emitted from found-
ries include  carbon monoxide (CO), gaseous  hydrocarbons or volatile organic
compounds (VOC), and  limited quantities of  sulfur dioxide (S02).  Major con-
stitutents of the particulate emissions are fine metallic fume from the molten
metal, silica dust from the core  and  mold sand, metallic oxide, organic partic-
ulates,  and general dust.  Table  4-1  presents  a detailed listing of all the
foundry  operations which  are potential emissions points.

     Obviously not all the sources listed in Table  4-1 will present.a compli-
ance problem  at any particular  foundry; many  are  generally minor emissions
sources.  However, depending  on the  type of operation, degree of control,
attainment status of  the area surrounding the foundry, and regulations appli-
cable  to the  foundry, each of these  sources  may  affect the compliance of
some foundries.   The  number and variety of  sources  also highlight  the  com-
plex problem of developing compliance strategies for foundries.

     Another  characteristic that  affects  foundry compliance is the nature
of the source.  Foundry emissions sources can be classified as one of three
types:  ducted or stack sources, process  fugitive sources, and open fugitive
                                     38
-------
           TABLE 4-1.  POTENTIAL SOURCES OF EMISSIONS IN FERROUS FOUNDRIES
                                       Gaseous
                         Particulate
                                                   Metal  Silica
                                   Coarse  Other
                                     CO  VOC  SOo  fume
                     dust   Smoke  metals  dust
Raw material receiving and storage
  Scrap storage
    Transfer to/from pile
    Wind erosion
  Coke/limestone storage
    Transfer to storage
  Sand receiving
    Manual or mechanical transfer
                                             X
                                             X

                                             X
Mold and core preparation
  Mixing (mulling)
    Charge to muller
    Dry mixing
  Sand molding
  Shell or hot box core or mold
    Charge shell machine
    Heating
    Cooling pallet
  Cold box
    Introduce catalyst
    Air sweep
  No-bake
  Oven bake
  Core washing
    Apply wash
    Burn off

Charge preparation
  Metal screening
  Coke screening
  Sizing
    Abrasive cutting
    Torch cutting
  Preheating

Melting and casting
  Cupola furnace
    Charging
    Melting
    Tapping
X
X
X

X
X
X
X

X
X
X
X

X
X
X   X
X   X
X        X
X
                       X
                       X
                       X
                                         X
                                         X
                                         X

                                         X
                              X
                                             X
                                             X

                                             X
                                      X
                   X
           X
                          X
X
X
X
X
X
X
                                             39
-------
TABLE 4-1 (continued)


Electric arc furnace
Charging
Melting
Oxygen lancing
Tapping
Electric induction furnace
Ductile iron inoculation
Pouring and cooling
Cleaning and finishing
Vibrating conveyor to shakeout
Shakeout
Appendage removal
Impact
Abrasive cutting
Torch removal
Carb on-air-arc
Surface cleaning
Abrasive blast
Grinding
Heat treating
Painting or coating
Sand-handling system
All conveyor transfer
Magnetic separator
Aerator
Screening
Reclaimer
Waste disposal
Sand handling
Baghouse catch removal
Landfill
Transfer
Erosion
Gaseous Particulate
Metal Silica Coarse
CO VOC S02 fume dust Smoke metals

X
X . X
X X
x x . ...
X
X
XX X X '
X XX
X XX-
X
X X
x x x
X
X X
X X
XX X
X
X
X
X
X
X
X
X X
X , X
X X

Other
dust

X



X
X
X
X
X


X
X
X
X

            40
-------
sources.  A ducted emissions source is one in which the emissions are con-
fined within the  processing  equipment and are released to the atmosphere
only through a well-defined duct or stack.  Cupola melting is an example of
a ducted emissions source.  Process fugitive emissions are emitted directly
from a particular process to the foundry environment.   These emissions reach
the atmosphere through  windows,  doors,  wall vents, and roof monitors.  To
control these emissions some type of capture mechanism must be used to con-
fine the emissions stream to a duct.  Iron ^curing and electric arc furnaces
are examples of process fugitive emissions sources.  An open fugitive emis-
sions source is one which is not associated with a particular piece of pro-
cessing equipment, e.g.,  storage piles and road dust.  Dust  generated from
movement of equipment on dusty foundry floors can also be considered an open
source.

     Emissions stream properties, particularly those  for  fugitive sources,
also impact on the  compliance status of the foundry.   Some of the proper-
ties which have the greatest impact are the  temperature,  moisture content,
and flow variations in the gas stream.

     For a fugitive source the chance of  emissions reaching  the atmosphere
is directly related to the temperature.   Those particles emitted at ambient
temperature have a greater chance of settling in the .foundry than those re-
leased in a bouyant plume.  In addition, highly bouyant plumes are difficult
to capture.

     Moisture content and the  variation in gas stream flow also impact on
the controllability of  an emissions stream.  Higher  moisture systems are
more difficult to control and can cause corrosion problems with control equip-
ment.  Ease of control is also dependent on the variation in gas stream flow.
A continuous emissions  stream can be controlled more easily and much more
economically than an  intermittent stream.  In addition, the determination
of compliance status is much easier for a continuous stream than for an in-
termittent stream.

     Table 4-2 characterizes  each of the emissions sources  identified  in
Table  4-1  with  respect to type of  source,  temperature, moisture  content,
and flow variability (both frequency and location).  The  information in Table
4-2 represents a  "best estimate" based on literature review and a limited
number  of  foundry visits; as such, some  characteristics  may differ  for a
similar operation at a particular foundry.

4.3  QUANTIFICATION OF FOUNDRY EMISSIONS

     The primary  goal  of this study was  to provide an overview of ferrous
foundry compliance  problems.   Given the broad scope  of the  study,  it was
not  feasible  to  perform a detailed quantification for each  of the sources
and pollutants listed in  Table 4-1.  Since initial contacts  with state agen-
cies  indicated  that their primary concern was particulate emissions, and
since  earlier studies by  MRI indicated that  gaseous emissions data are scarce,
this study focused only on particulate emissions.
                                      41
-------
TABLE 4-2.   POTENTIAL SOURCES OF EMISSIONS IN FERROUS FOUNDRIES

Raw material receiving and storage
Scrap storage
Transfer to/from pile
Wind erosion
Coke/limestone storage
Transfer to storage
Sand receiving
Manual or mechanical transfer
Pneumatic transfer
Mold and core preparation
Mixing (mulling)
Charge to muller
Dry mixing
Sand molding
Shell or hot box core or mold
Charge shell machine
Heating
Cooling pallet
Cold box
Introduce catalyst
Air sweep
No-bake
Oven bake
Core washing
Apply wash
Dry wash
Burn off
Charge preparation
Metal screening
Coke screening
Sizing
Abrasive cutting
Torch cutting
Carbon-air-arcing
Preheating
Melting and casting
Cupola furnace
Charging
Melting
Tapping
Type of
source


OFa
OF

OF

S
S


PFC
PF
PF

PF
PF
PF

Se
S
PF
S
PF
PF
PP
PP

PF
PF

PF
PF
PP
PF or S


PF
S
PF
Temperature


Ambient
Ambient

Ambient

Ambient
Ambient


Ambient
Ambient
Ambient

Ambient
. 350-450°F
Near ambient

Ambient
Ambient
Ambient
400 °F
Ambient
Ambient
Ambient
Above ambient

Ambient
Ambient

Ambient
High
. High
O
200-1200°F


High
.1400-1500°F
2600-2650°F
Moisture
content


Low
Low

Low

Low
Low


High
Low
73%

Low
Low
Low

Low
Low
Low
Low
Low
Low
Low
Low

Low
Lbw

Low
Low
Low
Low


Low
Low
Low
Flow
Variability


Ib
I

I

I
I


I
I
. I

I

cd

I
I
I
c
sc
sc
I
I

I
I

I,Mf
I.M
y ^*
I
I


I
• I or C
I or C .
                                  42
-------
                                 TABLE 4-2  (continued)-


Electric arc furnace
Charging
Melting
Oxygen lancing
Tapping
Electric induction furnace
Ductile iron inoculation
Pouring and cooling

Cleaning and finishing
Vibrating conveyor to shakeout
Shakeout
Appendage removal
Impact
Abrasive cutting
Torch removal
Surface cleaning
Abrasive blast .
Grinding
Heat treating
Painting or coating
Sand-handling system
All conveyor transfer
Magnetic separator
Aerator
Screening
Reclaimer
Waste disposal
Sand handling
Baghouse catch removal
Landfill
Transfer
Erosion
Type of
source

PF
PF
PF
PF
PF
PF
PF


PF
PF

PF
PF
PF

PF
PF
S
PF '

PF
PF
PF
PF
PF

OF .
OF

OF
OF
Temperature

High
High
High
High.
High
High
Moderately
High

Above ambient
Above ambient

Ambient
Ambient
High

Ambient
Ambient
High
Ambient

Near ambient
Near ambient
Ambient
Ambient
Ambient

Ambient
Ambient

Ambient
Ambient
Moisture Flow
content variability

Low
Low
Low
Low
Low
Low
High


High
High

Low
Low
Low

Low
Low
Low
Low

Moderate
Moderate
Moderate
Moderate
Moderate

Low
Low

Low
. Low

I
I or C
I '
I
I
I
I,M


I or C
I or C

I
I,M
I,M

C
I or C
C
I

C
C
c
c
c

I
I

I,M
I

a  OF - Open fugitive
b   I - Intermittent
c  PF - Process fugitive
^   C -' Continuous
e   S - Stack
f   M - Moveable
                                             43
-------
     In order to further limit the scope with respect to emissions quantifi-
cation, control technology analysis, and regulatory analysis, initial effort
was directed toward identifying these particulate sources which had the great-
est emissions potential.  Data which had previously been compiled were used
to determine emission factors for most particulate emissions sources.1'16 21
These emission  factors  were  then  applied to  annual production rates  (based
on data in Wallace and Cowherd16) to estimate the total annual emissions.
The results  are shown in Table 4-3.  The data indicate that the following
sources have the greatest potential impact on the environment:  (a) cupola;
(b) electric arc furnace; (c) pouring and cooling; and (d) shakeout.  Other
sources that might possibly have a significant impact are the cleaning room
(grinding, blasting, and cutting) and the sand system.  Each of these sources
was considered  in  -detail,  and the results are summarized below.  Further
analysis, of  emissions  data  for these sources is  presented  in Appendix B.

4.3.1  Cupola Emissions

     The cupola is the  one source of  foundry emissions for  which extensive
data are available.   These  data indicate that emissions from cupolas vary
widely (3.8  to  75.5 Ib/ton)  and suggest that these variations are due, at
least in part,  to  different design and operating parameters of cupolas. .,  ,
Some of the parameters which have been shown to affect cupola emissions are
specific blast  rate,  blast  temperature, melt rate, and in some cases, the
coke-to-melt ratio.   In some cases, the effects  of these parameters  have
been quantified  (see Appendix B).

     Two observations from the data in Appendix B are of particular signifi-
cance when evaluating and enforcing compliance of cupola furnaces.  First,
the variations  in  cupola emission factors has a dual impact on enforcement.
The wide range  of  emission factors makes the use of an average emission fac-
tor to enforce a process weight regulation questionable.   On the other hand,
the measured relationship between foundry emissions and operating character-
istics can be an enforcement tool.  If measured emissions data and associated
operating  characteristics are  available for a particular cupola, control
agency personnel can estimate the effect of changes in its operating charac-
teristics on its emissions and in that way make an initial determination of
any change in compliance status.

     Another observation which may be particularly useful for small foundries
is the result  of testing at Foundry A shown in Table B-7 in the appendix.
At this foundry screening of the scrap and careful handling to prevent charg-
ing of loose sand,  rust, and coke fines resulted in a 50% reduction in emis-
sions .  This practice may be an economically feasible way of reducing emis-
sions in smaller foundries where fabric filter system costs make the system
economically infeasible.

     The development and installation over the past 10 years of the divided
blast cupola is a  technological step which  has the potential to decrease
cupola energy use  and emissions.  This system, described in Appendix A, has
been shown to  significantly  reduce coke consumption.  It is quite  likely
that the decreased coke consumption will result in lower emissions per ton
of iron produced.  However,  no data are available which quantify this reduc-
tion.

                                     44
-------













w
w
M
C*

S
p^
g

a
§
M
1
1
cn
to
M
S
^H
J
CJ
M

re AJ
!— 1 1— (
U
>1
*•*«.
en
AJ C
•

o c
U 0
b u
S
«


«
1
•a
a
u
i





c a e ca 1*2 u u


O  AJ
3 t Bt
gi* 	 g *
S^ ** m
i-l CM <• O 1O
O O •* O O O ^ *>tf *^ O ^ "T *^ *!f
cj in CM tn *3- eg NN— i cgoje«i
. o
a> ' o
SfS CO ' 00
e c e
•rt 3 OS \w O O fc< 00 O
U U4C O**4 Q) C -H
tn bXOC *M^H -u

RlC)WbW>O "* C 3 *•< C «4 b V . C b
COQQOStt U b CO *Q C >, <9 C 60
-HQ.OO«'0> 0 3 -HJi 4J B «« « » C
A43 -*C« B O fin O -H J5> g •*
»^u. www M a* "OT w S -oo *•• " 5J u
53 •«* S U O « Q
ca
M •
a a £















.

§
«IH
kJ
0)
c

a
• u
a
3
1
i  o b
41 N b *E
N -H O -H
«rt a n
to i a
rH i-4 U
~ 3 C E
3 C 
-------
4.3.2  Electric Arc Furnace Emissions

     Emissions data for the electric arc furnace  (EAF) are much, more limited
than  the data for cupolas.   It appears that little effort has been made to
relate emissions  data to EAF operating characteristics.  Data  compiled  dur-
ing an  earlier Environmental Protection Agency  (EPA)  study indicate that
EAF emissions range from 4 to  40  Ib/ton with an average of 13.8  Ib/ton.1
All other  data identified during this  study fall  within  this  range.

     Available data indicate that EAF  emissions are related to scrap quality
and cleanliness.   Data  presented in Appendix B show increases in  emissions
of 30 to 100% when dirty or low quality scrap  is  used.   Observations of in-
dustry personnel  and from past  plant visits  indicate that visible emissions
from  charging increase  appreciably when dirty, particularly oily,  scrap is
charged  to the furnace.

     Limited  data (see Appendix B)  also indicate  that the practice of oxygen
lancing  in steel  foundries also has an impact on EAF emissions.  The data
indicate that gas temperature, gas flow rate, pa'rticulate loading, and CO
emissions  increase during  lancing.   These  factors are  important considera-
tions in the  sizing of steel foundry EAF control  equipment.

4.3.3  Pouring and Cooling Emissions

     The emissions generated from pouring and  cooling castings have generally
not been considered to be a problem either by  foundry personnel or air  pollu-
tion  control  agencies.   However, limited test data  indicate  that  if sand
molds are  used, pouring and  cooling operations may be  a  significant  source
of particulate emissions.  Test data show that pouring emissions range  from
0.6 to 24  Ib/ton  of iron poured with an average  of about 6  to 10  Ib/ton of
iron poured.   This level, is particularly important because most operations
are not  controlled and  may be  difficult to  control as  described in Section
5.0.

     Foundry personnel report that  the  quantity of pouring and cooling  emis-
sions is probably related  to such  factors as  mold size, mold  composition,
sand-to-metal ratio, pouring temperature, and  pouring  rate.   Test data  are
not sufficient, however, to quantify the effects  of these parameters on emis-
sions.

     It  should be  noted that the estimates of  emissions are based on limited
test data, and some of  the data were obtained  from pilot scale operations.
It is suggested that  these data are not sufficiently  reliable to use for
enforcement purposes.

4.3.4  Shakeout Emissions             ,

     The removal  of castings  from a sand mold releases  moisture that has
been trapped  in the mold,  dust from the sand .and binders which have dried
during pouring, and products of thermal decomposition of the chemical binders
as they  are exposed to air.- Available emissions  test data range from 0.17
                                     46
-------
to 18 Ib/ton of iron castings with an average of about 3 Ib/ton of iron cast-
ings.  Limited data indicate the wide variation in the emissions may result
from variations in such parameters as sand to metal ratio, length of cooling
time prior to shakeout, size of casting, and number of cores in the casting.

4.3.5  Cleaning Room Emissions

     As reported in Section 4.1, cleaning room emissions are generated by a
number of operations.  Available data are not sufficient to quantify emissions
from most of these operations, and available data are certainly not sufficient
to determine the effect of operating and design parameters on emissions quanti-
ties .

     The only available cleaning room test data are for sand and shot-blasting
operations. • These data indicate  that uncontrolled  emissions range  from 27
to 500  Ib/ton  of castings  cleaned, and  several  tests  show emissions in the
range of 250 to 400 Ib/ton of castings cleaned.  It should be noted, however,
that most  of  these  emissions  were controlled at the 98  to 99+% level.

     Limited engineering estimates are available for emissions from grinding
wheels.  These data  indicate that emissions in the range 1.6 to 15 Ib/ton
of castings cleaned are generated from grinding wheels.

4.3.6  Sand-Handling Emissions

     Sand handling,  like  cleaning, has  a number of unit  operations which
generate particulate emissions, but test data are not sufficient to quantify
emissions from these unit operations.  Limited test data indicate that emis-
sions from  the sand-handling system  (starting  at the  point the  sand leaves
the shakeout and  eanding  when it  enters  the muller)  range from 0.6 to 50
Ib/ton  of  sand handled.   Given this  wide  range  and  the  limited  quantity of
data, it is not possible to estimate an average.
                                     47
-------
                                REFERENCES
 1.   A.  T.  Kerney Co.   Systems Analysis of Emissions  and Emissions  Control
     in the Iron Foundry Industry (Volume I,  Text,  and Volume II, Exhibits).
     U.S. Environmental Protection Agency. PB 198  348 and PB 198 349.   .
     February 1971.'

 2.   Sylvia, J.  G.  Cast Metals Technology.  Addison-Wesley Publishing  Co.
     1972.

 3.   Cupola Handbook.   The American Foundrymen's Society.  1975.

 4.   Cleaning Castings.  American Foundrymen's Society, Inc.   1977.

 5.   Malleable Iron Castings.   Malleable Founders Society.  1960.

 6.   Chemical Binders  in Foundries.  British  Cast Iron Research Association.
     Birmingham England.  1976.

 7.   Beeley, P.  R.  Foundry Technology.  Butterworth  and Co., Ltd.,  London.
     1972.

 8.   Quality Ductile Iron Today and Tomorrow, Proceedings of Joint  AFS-DIS
     Conference.  October 1975.  Rosemont, IL.  American Foundrymen's Society,
     Inc.   1975.

 9.   Heine, R. W., C.  R. Loper, Jr., and P. C. Rosenthal.  Principles of
     Metal  Casting. McGraw-Hill, Inc.   1967.

10.   Wallace, D. and C. Cowherd, Jr. Fugitive Emissions from Iron  Foundries.
     U.S. Environmental Protection Agency. EPA-60017-79-195,  August 1979.

11.   Gutow, B. S.  An Inventory of Iron Foundry Emissions.  Modern  Casting.
     January 1972.  pp. 46-48.

12.   Steffora, T. J.  Induction Furnaces, Preheaters, and Air Pollution.
     Foundry.  August 1978.  pp. 82-86.

13.   Recent Tests on the Cokeless Cupola.  Foundry  Trade Journal.   February  19,
     1976.   pp.  234-235.

14.   Warda, R. D., and R. K. Buhr.  A Method  for Sampling Cupola Emissions.
     AFS Transactions.  Volume 81.  1973.  pp. 24-31.
                                  48
-------
15.


16.


17.



18.



19.


20.



21.
Warda, R. D., and R. K. Buhr.  A Detailed Study of Cupola Emissions.
AFS Transactions.  Volume 81.  1973.  pp. 32-37.

Davis, J. W., and A. B. Draper.  Effect of Operating Parameters in Cupola
Furnaces on Particulate Emissions.  AFS Transactions.  1973.  pp. 287-296.

Patterson, W., E. Weber, and G. Engles.  Dus.t Content of Cupolas for
Cupolas of Different Designs and Modes of Operation.  The.British Found-
ryman.  March 1972.  pp. 106-117.

Crabaugh, H. R., A. H. Rose, and R. L. Chass.  Dust Fumes from Gray
Iron Cupolas - How They Are Controlled in Los Angeles County.  Air Repair,
4(3).  November 1954.  pp. 125-130.
Compilation of Air Pollutant Emission Factors.
tection Agency.  AP-42.
U.S. Environmental Pro-
Miller-, W. C.  Reduction of Emissions from the Gray Iron Foundry Industry.
Paper 71-134 presented at the 64th Annual Meeting of the Air Pollution
Control Association, June 27-July 2, 1971.

Bates, C. E., and W. D. Scott.  Better Foundry Hygiene Through Permanent
Mold Casting.  National Institute of Occupational Safety and Health.
Contract No. 1 R01 OH 000456-01.  January 1976.
                                  49
-------
                     5.0  EMISSIONS CONTROL TECHNOLOGY
     The development  and application of emissions control  technology for
control of  foundry emissions  is  complicated by the wide  variance  in opera-
tions among foundries,  the number of emission points within any particular
foundry, and the differences  in pollutants and gas stream characteristics
of these sources.  This study was motivated in part by reports that the above
factors have resulted in significant technical problems associated with in-
stallation  of air pollution control equipment at foundries.

     The examination of these problems was directed toward two specific ob-
jectives.   The  first  was  to determine the  type of technical problems  asso-
ciated with design or operation of  control equipment.   The second  was to
investigate exemplary control systems which had successfully alleviated these
problems.

     The above  objectives  were accomplished through five basic  activities.
First, selected state and local agencies  having a number  of foundries in
their jurisdiction  were contacted by Midwest Research Institute  (MRI) to
obtain information  on the  types  of problems associated with control equip-
ment in ferrous foundries and to identify foundries which had a record of
continual compliance.   Concurrently,  a thorough literature  search was ini-
tiated, and control device manufacturers were contacted to obtain data on
foundry controls.  The foundries that were identified by the state and local
agencies were contacted by telephone to obtain design and  operating  data
for their control  systems.  Finally,  a limited number of foundries  appear-
ing to have the best systems for continual emissions reduction were visited.

     It became apparent early in the study that the scope was not sufficient
for a detailed  analysis of all  sources  and pollutants.  Based  on initial
contacts with state and local agencies, it was decided to limit the analysis
to particulate  controls only.   The study was  further limited to  the  five
operations  (see Section 4.0) having the greatest impact:   (a) cupola melting;
(b) electric arc furnace melting; (c) pouring and cooling;  (d)  shakeout and
sand handling; and (e) the cleaning room.

     The review of foundry control technology resulted in the following con-
clusions:

     1.   Adequate control technology is available for most foundry emissions
sources.   The major exceptions are pouring and cooling emissions,  emissions
from carbon-air-arcing'operations and some grinding operations.

     2.   Concerns of the Occupational Safety and Health Administration (OSHA)
and the National Institute for Occupational Safety and Health (NIOSH) about
                                  50
-------
the internal foundry sources which result in worker exposure to contaminants
as well as  fugitive emissions  from the  foundry have 'led  to  the development
and implementation of improved capture systems for these sources.

     3.  Malfunction of  control  equipment is cited as a major problem by
both control agency and foundry personnel.

     4.  Reduction"of malfunctions  is possible through improved operation
and maintenance procedures and proper equipment design.

     The remainder of  this  section presents the  data  that  form  the basis
for the above  conclusions.   The discussion  is divided into three parts.
The first is  a general overview of foundry emissions control.  The second
summarizes  the  status of emissions control  systems with  respect  to availa-
bility and,  -when possible, extent of application.   (More detailed descrip-
tions  of the  available  control systems are presented in Appendix C.)  The
final  section  identifies some of the major malfunction problems that were
identified  by  control  agency and foundry personnel and  describes some of
the design  features  and maintenance procedures employed by some foundries
to reduce malfunctions.

5.1  FOUNDRY EMISSIONS CONTROL

     The principal components  of an effective emissions control strategy
are availability  of  an effective control system, installation of the con-
trol system and operation and maintenance of the system in a manner which
ensures continued compliance.  The paragraphs below summarize the status of
foundry controls with respect to each of these components.

     Before discussing  the  availability of  foundry control systems, it is
helpful to  describe  briefly the components  of a  "typical" foundry control
system.  The components  of the control  system depend upon- the nature of the
emissions source.  For a  stack or ducted emissions source, the control sys-
tem consists  of a particulate removal  device (generally referred to as an
air pollution  control  device) and possibly  some  type  of gas  conditioning
equipment positioned ahead  of the  removal device. The primary removal de-
vices  used  in  foundries  are wet scrubbers, and fabric  filters.

     The control  system for a fugitive emissions  source consists of  a  cap-
ture device which contains  the particulate and exhausts  it  to a  duct where
it  is  then  collected by  a removal device.   Some  type  of gas  stream condi-
tioning may also be used with  fugitive  emissions  sources.  The primary cap-
ture devices  are hoods,  either close  capture  or canopy, and enclosures.
The removal devices are  the  same as those for stack sources.

     One alternative  to the fugitive  source control  system is preventive
measure  to  reduce or eliminate the  generation of emissions.   One example
might  be  water sprays at sand conveyor transfer points or  the use  of  the
Schumacher  system  (see Section 5.2.2.2) to  inhibit dust  generation.  Another
method for  either stack or fugitive sources is to replace  one process  with
another that  is less polluting.  Other examples are the  use of in-mold rather
than ladle  inoculation of ductile  iron and  the replacement  of a  cupola with
an  induction  furnace.

                                   51
-------
     Data from the literature review, industry personnel, and air pollution
control agencies  indicate  that for many foundry particulate sources, both
stack and fugitive control systems are available.  There are some operations
for which,  depending  on the  type  of  casting  (e.g.,  shakeout  of a pit mold)
or size of the foundry  (e.g., manual sand handling in a small foundry), emis-
sion control systems may not be technically feasible and are very expensive.
For a limited number of sources, control systems are not available.

     The extent of installation of available  emissions  control-systems de-
pends on several factors.  The first and most obvious is the size and degree
of mechanization  of the foundry.   Larger foundries  and  those that  are more
mechanized tend to have more operations that are controlled and also tend
to have a higher level of control.  A second factor, especially among small
to medium sized foundries, is management attitude.  During telephone contacts
and plant visits, MRI found that some foundry managers considered the instal-
lation of an acceptable level  of  emissions control  to be part of corporate
responsibility.  Other  managers,  however, considered air pollution control
systems to have an unwanted adverse impact on production.  The former shops
generally had better overall control.  Another factor in the degree of appli-
cation of emissions  control  systems is the regulatory stance of local air
pollution control officials.   MRI visited one foundry which, as a result of
aggressive  enforcement  by  the  local agency, had installed a well-designed
fabric filter  system  on a cupola.  "A  foundry of similar size in the same
state but under the jurisdiction  of  another  agency  less than 50 miles away
was still operating with an uncontrolled cupola stack.

     Both air  pollution control agencies and industry personnel indicate
that the greatest foundry compliance problem is the malfunction of air pol-
lution control equipment.  This is sometimes  the result of poorly  designed
equipment, but more often  it is the  result of improper  operation and main-
tenance either through lack of knowledge or limited resources.

     The following section provides  in some  detail the available  data on
control system availability and application.  Section 5.3 discusses malfunc-
tions operation and maintenance.

5.2  AVAILABILITY AND EXTENT OF INSTALLATION OF EMISSIONS CONTROL EQUIPMENT

     Information from control agency and industry personnel and the litera-
ture indicate  that  adequate  control technology is generally available for
both stack and fugitive emissions sources in foundries.  The discussion below
briefly summarizes available control systems for the major foundry emissions
sources identified in Section 4.3.  More detailed descriptions of these con-
trol systems can  be  found in Appendix  C.  The discussion is divided into
two sections.  The first describes melting furnace  control,  and the second
control of all other fugitive emissions sources.

5.2.1  Melting Furnace Controls

          MRI examined control technology for two melting furnaces, the cupola
and the electric arc furnace (EAF).  Cupola control systems generally consist
of a wetcap or a system comprising an afterburner, a gas cooler,  and a venturi
scrubber or fabric filter.  The EAF control system consists of a  capture de-
vice and a particulate collection device.

                                  52
-------
5.2.1.1  Cupola Controls—•
     The particulate collection devices most 'frequently used on cupolas are
wetcaps, venturi scrubbers,, and fabric filters.  Wetcaps which may be either
single or  double  are low efficiency scrubbers which can still be found on
some smaller  cupolas  in  nonmetropolitan  areas. . The wetcap  is  installed  in
the cupola stack above the charging door.  Flow through the wetcap is main-
tained by  the draft of the cupola without the use of auxiliary fans.  The
wetcap is effective in controlling the large particles generated from loose
dust and sand in the charge but is ineffective in controlling fine particles.
In general, single wetcaps control about 50% of the total particulate while
double wetcaps control about 85%.  A diagram of a wet cap is shown in Figure
B-l (Appendix B).

     If a fabric filter or venturi scrubber is used for particulate collec-
tion, the typical control system includes an afterburner, a gas cooler, the
collection device, and a fan.  The afterburner, usually located in the cupola
stack above  the  charging door  or just before the  gas  takeoff duct,  is  used
to oxidize the carbon monoxide and to burn tars and oils.  The oxidation of
these compounds  prevents explosions  in the control system and reduces the
potential  for plugging that  can occur in some control systems.  The after-
burner  raises the temperature of the  gas  stream  to about 1200 to  1400°F,
which is  too high for operation of  either a scrubber or a fabric  filter.
Cooling of the gas  stream can be accomplished in  one  of three  ways:  evap-
orative cooling,  dilution cooling,  or radiant cooling.  Evaporative cool-
ing, or quench cooling with water sprays used to  cool the gas stream is pre-
ferred  by most foundries.  However,  some  foundries that have cooler gas
streams from other  processes that can be mixed with  the cupola gas stream ,
use dilution cooling.  A few foundries may use a  long duct  system to obtain
radiant cooling.  However^ a more common form of  radiant cooling is the  in-
direct  heat exchanger where the heated  air is used  for cupola blast air.
Some systems  may  have  a  combination  of two  of these cooling methods.  A  fan,
which may be located either in front of or behind the collector,  is needed
to overcome  the pressure drop  across  the collector.     ;

     As indicated above  the  two  particulate  collection  devices most  commonly
used with ferrous foundry cupolas are fabric filters  and venturi scrubbers.
Davis et.al.1 indicate that  fabric filters  are used more  frequently in small
and medium sized cupolas while scrubbers are commonly found in larger  cupolas.
Detailed  descriptions  of both  devices are  included in Appendix C.   The para-
graphs  below describe some of  the design features identified during the  study
as  factors that  might affect on compliance  status, such as  ease of  monitoring
and  equipment performance and  condition.           •     , .

     The  major design features that  may  affect fabric filter  compliance  are
 filter  material, cleaning method,  fan location,- and air-to-cloth  ratio.
 Some type of glass or Teflon is usually chosen as the bag material because
 of the high  temperature of the cupola gas stream.  These fabrics can with-
 stand  temperatures  continuously in  the 450 to 500°F range and  maximum  temper-
 atures  of about  550°F.   The literature  does indicate that if  fluorspar  is
 used in the  charge,  the resultant  fluorides in the  emissions stream will
 destroy glass fibers.2  Under these conditions Nomex  or Teflon bags or Teflon-
 coated bags should  be used.
                                   53
-------
      Information compiled during the study indicates that both reverse air
 (including pulse)  and mechanical  shaker mechanisms  are being  used  for  clean-
 ing  cupola filters.  No particular  advantages were  identified for  any  partic-
 ular cleaning mechanism.   However,  at least one small foundry, which melts
 only 3 h/day,  did  indicate  that they have  improved  bag life by going to daily
 inspections  coupled with manual  shaking rather than automatic mechanical
 shaking.

      Typical  air-to-cloth ratios  for metallurgical  furnaces are in the range
 of 1.5 to  2.5:1, and the  air  to cloth ratio  for a cupola  filter should fall
 in this  range._ The American Foundrymen's Society  (AFS)  suggests  that the
 ratio should be 2:1 with the gas  volume based on the maximum design volume
 that occurs during burndown (the  end of the  heat).

      Both  positive and negative pressure fabric filters were  used  on foundry
 cupola systems.  Both appear  to operate with equal  effectiveness,  and  neither
 type system is inherently better.   Positive  pressure baghouses  have the ad-
 vantage  of lower capital cost, greater ease of inspection and  maintenance
 of the bags,  and reduced  fan noise.  Negative pressure baghouses generally
 have less  fan wear and maintenance  as  well as low'er operating costs.

      Two other design features were identified  during the study which  impact
 on the wear of the bags and the baghouse.  First, it is essential  that fabric
 filters  be well insulated,  especially those located in northern climates.
 If not,  condensation in the baghouse  is likely,, and the corrbsive  nature of
 the  condensate will result  in early deterioration of both the bags and the
 housing. Second, it  is important  to have a mechanism whereby  the gas stream
 bypasses the  baghouse  if the inlet temperature is  above  550°F. The best
 designs  identified during the study have a double fail-safe system.  A tem-
 perature sensor  immediately downstream from  the primary cooler controlls a
 dilution air  damper.  If  the  gases  leaving the  cooler are too hot, the damper
 opens providing ambient dilution  air  for cooling.  Another temperature sensor
 is located between the dilution air damper and  the  baghouse inlet. If the
 temperature at this  sensor  is above 525°F, the  damper to the baghouse inlet
 closes and a  bypass  damper  to the atmosphere opens.  In addition,  an alarm
 sounds to  alert the  operator  to the problem.  One foundry visited  installed
 this  type  system 2 years ago and has had no problems with bag burnout and
 has  had  minimal  time in  the  bypass mode.  Since this same system  is well
 insulated, they  have also had no  corrosion problems and have  had extremely
 good bag life.

     Two types of venturi scrubbers are used frequently for cupola control,
 the  typical venturi  tube  with cross current  water introduction  and a fixed
 throat and the variable  throat flooded disc scrubber.  For both types of
 scrubbers, the primary design parameters which  affect compliance are pres-
 sure drop across the scrubber and the material of construction.

     It is well established that the fractional efficiency of any wet scrub-
ber  is strongly  related  to  the pressure drop across the  scrubber.  Data in
                                  54
-------
Appendix B show that the cupola emits significant quantities of fine particu-
late.  Thus, a  relatively high pressure drop  is needed  to  effectively  con-
trol cupola emissions.  AFS indicates that scrubber pressure drops of 60 to
90 in H20  are  necessary to attain grain loadings of 0.02 to 0.03 gr/scf.2
The higher pressure drops are needed for dirty or oily scrap.

     Because the emissions  from the  cupola tend to  result  in highly  corro-
sive substances in  the wet scrubber, choice  of construction materials  is
important.  Both the literature and the foundries contacted during the study
say that  it  is  imperative that the venturi throat, spray nozzles, and the
separator be constructed of stainless steel if maintenance problems  are to
be avoided.  In addition it is suggested that the fan housing be epoxy coated
and the fan blade be constructed of stainless steel.  It has also been sug-
gested that  less efficient  self-cleaning paddle wheel fans are often used
because they are easily maintained.  Foundries contacted that used the mate-
rials  suggested above  had  experienced minimal maintenance  difficulty.

     Information gathered from contacts with control agencies and foundries
strongly suggest that the systems with the designs described above can achieve
initial compliance with the existing regulations.   To achieve continued com-
pliance, further operation and maintenance procedures are available  (Section
5.3)

5.2.1.2  Electric Arc Furnace Controls—
     Emissions from the electric arc furnace (EAF) are particularly difficult
to control since at least some of the emissions are fugitive, i.e., they do
not necessarily enter the atmosphere from a well defined duct or stack.  As
such the  emissions  control  system  comprises a capture mechanism to contain
the emissions stream  and  a particulate collection device.   The paragraphs
below describe  the  various  capture mechanisms that may be used by EAFs.
These descriptions  are  followed  by a discussion of particulate collection
devices.

     Three distinct emissions  streams  are generated by the various phases
of the EAF melt cycle:  melting and refining, charging, and tapping.  The
capture device is generally designed to contain one or more of the streams.
Charging and tapping emissions are often captured by one system and the melt-
ing and refining by a different system.  Melting and refining capture systems
are described below, followed by a discussion of charging and tapping systems.

     EAF melting and  refining emissions are generally controlled  by  one of
three systems:  roof  hoods;  side draft hoods; or  direct furnace (or  shell)
evacuation.  Each of the systems controls emissions during melting but does
not operate  when the  roof is removed for charging  or during tapping. Two
other systems,  the  furnace  enclosure and the  close  capture hood,  have been
used on  a  limited basis.  These systems capture charging and tapping emis-
sions as  well  as melting and refining emissions.   These five  systems are
described briefly below and in more  detail in Appendix  C.  Diagrams  of  the
systems are shown in Figures C-10 through C-16.   Typical flow rates and cap-
ture efficiencies for each are shown in Table 5-1.
                                  55
-------
        TABLE 5-1.  TYPICAL EXHAUST FLOW RATES AND EMISSION CAPTURE EFFICIENCY
                      OF MELTING CONTROL SYSTEMS
Typical exhaust flow rate for
model furnaces , in. f ts/min

Roof hood
Side draft hood
Direct evacuation
Furnace enclosure
Close capture hood

4 ton/h
16,300
27,500
7,000
17,000
27,500
Furnace size
10 ton/h
25,200
60,000
18,000
25,000
42,000

25 ton/h
63,600
150,000
45,000
-60,000
106,000
Emission capture
efficiency (percent)
Range
95-100
90-100
90-100
80-100
90-100
Typ i c a l~ma ximura
99
99
99
99
- 99
Source:  Electric Arc Furnaces in Ferrous Foundries - Background Information for
           Proposed Standards (Rough Draft).  U.S. Environmental Protection
           Agency, Research Triangle Park, North Carolina.  April 1980, pp. 4-32.
            A typical  roof  hood is mounted directly on the furnace and pulls air
       through the  annular  openings  around the  furnace electrodes.  The  roof hood
       is the most effective device for capturing EAF emissions.  It has the addi-
       tional advantage of  muffling  noise.   Its major disadvantages are  stress on
       the furnace  roof and supports due to  the  substantial weight of the hoods
       and difficulty encountered in maintaining furnace roofs.

            The side draft hood is also mounted on the furnace roof but is open on
       the top and  on one side to allow  free movement of  the  electrodes.  Since
       the side draft hood collects emissions after they have escaped from the fur-
       nace through the annular openings around the electrodes, it requires greater
       air volume and  is  sometimes slightly  less efficient  than the  roof hood.
       However, the side draft hood is easier to retrofit than the roof hood, places
       less stress  on  the roof  and supports,  and  allows easier maintenance of the
       furnace roof.

            Direct furnace  (or shell) evacuation is often called fourth hole evacu-
       ation because it collects  gases from the furnace through a fourth hole in
       the roof. A  heat  resistant elbow  is  mounted  above  this ventilation hole
       through which the gases from the furnace are pulled into a duct.   This evacu-
       ation system provides  good emission control and minimizes both the space
       required on  the furnace  roof and  the  gas  volume which  must be withdrawn.
       Disadvantages are that the ingress of air  to the furnace, although slight,
       cools the slag, makes control of the temperature difficult, and reduces the
       carbon level in the melt through formation of carbon monoxide.3  It has also
       been suggested  that  direct evacuation is not applicable to small furnaces
       because of lack of space for a  fourth hole, pressure fluctuations in the
       furnace which are too rapid for the automatic control dampers and deteriora-
       tion of the  shell refractory  because  of  excess weight on  the furnace roof.

                                         56
-------
     A furnace enclosure is a metal shell which completely encloses the fur-
nace and tapping  area.   It captures emissions from charging, tapping, and
melting.  The major advantage of this system is that emissions from all phases
of the melt  cycle are captured.  Although  there  is  some  speculation  that
the system may  inhibit  operations of the charging or  tapping cranes,  one
steel plant using the system has encountered no problems.  Only two installa-
tions in the United States currently use total enclosure systems.  More ex-
perience at both facilities is needed before the -system can be fully evalu-
ated.

     The close capture hood is a collection of hoods connected to an exhaust
plenum.  Dampers allow the system to regulate the exhaust volume to the ap-
propriate hoods during different phases of the melt cycle.  Melting and re-
fining emissions are captured by a rectangular hood which surrounds the elec-
trodes, acting  much like  a side draft hood.   Capture of charging emissions
is accomplished by an annular ring hood which has slots on  the inside that
collect the fumes during charging.  Tapping emissions are evacuated through
an inverted U-shaped hood  that covers the tapping spout.

     The advantage  of the close capture design is that control of, charging
and tapping is provided at an exhaust flow rate much less than the flow rate
for canopy hoods or furnace enclosures.  This significantly reduces the quan-
tity of exhaust gas delivered to the particulate  control device, thus  cutting
costs of gas cleaning.  Also, the close capture hoods are simpler and  consider-
ably less  expensive to  install than a  furnace enclosure or  canopy hood.
The  disadvantage  is that  complete control of charging  and tapping may not
always be  provided because the charge/tap hoods  do not completely enclose
emission sources.3

     In addition  to the furnace enclosure  and close capture hooding just
described, four techniques are available for  capturing  charging  and tapping
emissions. They are:   (a) canopy hoods;  (b)  building  evacuation;  (c) bay
evacuation;  and (d) ladle pit  enclosure.  The exhaust  volumes and capture
efficiencies of all six systems are shown in  Table 5-2.

     Canopy hoods  are the  capture mechanism most  frequently employed  to col-
lect charging emissions.   The canopy hood is  suspended  at a sufficient height
above  the  furnace to allow clearance  for  the crane, or it is attached to
the  foundry  roof.  If the furnace has  a  melting emission capture system,
the  hood is  operated  only during  charging and tapping.

     Effective  capture  of emissions is not always attained by the use of a
canopy hood.   As the furnace  is  charged, emissions  are sometimes  diverted
away from  the canopy because  of impingement on overhead cranes and the charge
bucket.  Another problem  is caused by cross drafts  in the shop which lower
canopy hood  collection efficiency.   Upward flow of the emission plume from
the  furnace  is  easily disrupted by drafts from openings along foundry walls
and  doors, passage of shop vehicles, temperature  gradients  within the shop,
and  even suction hoods which may ventilate other nearby foundry processes.
A canopy hood  is  not generally as effective for small  furnaces because there
is less thermal uplift  generated.  Meteorological conditions  may also influ-
ence the plume conditions.  High pressure systems and  low humidity tend to
                                   57
-------




C3
p*j
*M

a
P^
H"t
CD
p£
«gj
*^
C^

fg*
o

^H
CJ
y^
ft]
1—4
o
M CO
pC| pi]
£&4 KH
PL} p£
pn
PS-] f^*
p^ £3
b^? C5
fcH £KI
£U
*C J™H"
u w
r. S
P^ ^J
O S
(H
CO H
CO <
M
33 CO
£x3 £x*t
CJ
P ^
£c j>
«ffi f-o
p
CO
Cs^ H^
EH C)
^3 P*
CM fc^
jzj
3: O
S"
P&4 CD
. 525
c~* &™H
CO OH
3 P-i
«Cg ^J
{*H E"^
PS
pC|

S3
O
HH
Q4
pH
H

•
CM
1
in

S
£Q
^j
H








-
(U
J"^
£j
p
(i
CO
CJ

d
o
•H
to
ca
•H
m












^4
o
4H d
•H
CJ g

CQ W
U 4->
MH*
^
o d
rH -H
tfH
ca
-P 0)
co cj
3 co
CO CS
sf^ Jj
*M* *•<
!"• *J
01 4H

rH rH
CO Ol
CJ TJ
•H O
PH ^
fa»\
H





































d
0)
u
J*4
CJ
CX
**—S

^
U
C
Qj
•rt
U
•H
M-l
^•t
0)



i
*!•
oc
£

r~
03
C
•r
a
-l^1
^





0)
6C
d
R
Q£













0)
N
•H
M

O
CJ
S
M
•3
tr4

rH
01
-a
0
23







*c
"^1^
d
o
p

m
CM





j3
-»,
g
4J

O
rH





tE*
****„
d
o
^J

^*f





































XI
O ON O O O O\
ob ON co ON oo ON










o o
- o o o m . o o
O\ rH OO C^ CT\ rH *-
1 1 II 1 1
in m o o o o •
rH cj\ in 00 \O CO






o o o o o
o o o o o
VO O O 0 O
•k *k ^ •* •*.
rH - cj ' o cj 01 - cj
W CO rH d SH d
•o > o 01 3 » vH CJ d
ft "O co SH 01 0)
O iH d 3 W rH
d H SH 4H O 13
CO 3 3 - rH CO
cj pq to cj IH"
0
CO P
TJ
SH co
co d
T3 O
d XS -rl
CO P P
•P SH -H
CO O t3

13 (0
4) ~
ca M >>
0 S^ 0
ft CO rH
O PH rH
SH • co

iH O 60
SH 60 ja d
o d co -H
4H 03 CO
•H d CO
d SH -H oi
OH SH
•H 4J CJ
p xl d d
CO O Ol -rH

SH co 0) xl
O U SH 4J
tH CO ft -H
d oi S
H f**| 0) "
SH 13
"O •> CO 01
d >> CJ
3 u • ca 3
O d 4-> 13
j-i cj m .
CO d , rH
pa o' co xi
•H CO CO
IP O SH
CJ SH 01
CO Ol O TJ
01 P -H
SH o 4H co
•H SH -H d
'O PH O
§•0 o
rH 0)
o co cj co
fn P • 3 d
rt r^^ ^3 o
CO 01 1 01 -H
O § CO
SH o • >. SH
SH SH ft rH 0)
Ol -H ft rH ft •
to > eg o co
d -r-t oi
d w • P 60 cj
•HO d d «
• 00 CO ' -H d
co co cn P ft SH
01 • rH CO ft 3
CJ E3 Xl CO 4H
05 rH 3 p
d -H W rH
SH • SH SH Ol
3 >N ft >> 00)
to P < cj 4-1 P
4H d W
CJ (0 0) >»
SH SH -H UP
< Q 03 U C 03
d -H oi
U X! -H 4-1 -H -
•H M rH 4H U •
SH 3 O 01 -H 01
p O SH • 4H •
U PS CO d 4H -H
O) ^ CJ O 0)
rH -H ••>
M -P Ol Ol
U SH rH
01 3 13
rH P 03
Ol rH ftrH
U ' O 03
SH CJ CJ
3
0
CO CO XI
58
-------
allow efficient upward  flow of the plume  to  the  canopy.  However, during
periods of low pressure, high humidity, and/or strong winds, thermal uplift
may not be sufficient to carry fumes directly into the canopy.3

     Some iron foundries  capture charging and tapping emissions by evacu-
ating the complete  melt shop.   Since building evacuation systems require
greater air  flow  than do  canopy  hoods,  they are generally used  only if the
building is  not structurally suited for canopy hoods  or  if  there is also a
need to collect fugitive emissions from other sources.

     A recent modification to the building evacuation system is the bay evac-
uation system, which is in  limited use  in  foreign steel  mills.  In the bay
evacuation system, each separate shop bay, or furnace area,  is isolated from
the others by walls with closed doors.  A canopy hood is then placed at the
top of the bay.  While no foundries in the United States have employed this
system, it is anticipated that it may eliminate the problems associated with
cross drafts  that occur when canopy hoods  or  building evacuation are used.

     One steel mill in  the United States  uses a  tapping pit enclosure to
capture emissions from  EAF  tapping.   In this  system the  ladle is placed in
the pit with a standard overhead crane.  The crane is retracted and a move-
able cover seals  the pit.   Hot metal  flows through a  closed launder to the
ladle, and air is exhausted from the enclosure.  This system,  which visibly
appears to control emissions well, could easily be designed into a new melt
shop.  However, structural limitations, primarily space around the pit, may
limit its retrofit applicability.

     The systems described above are considered to be effective in capturing
EAF melting emissions; in many cases charging and tapping emissions can also
be captured.  However,  more experience with  capture  systems for charging
and tapping is needed before these systems can be said to assure compliance.

     The particulate collection device used with virtually all domestic EAFs
is the fabric filter.  Industry data indicate that both positive and negative
pressure units are  used.  Newer  installations tend to be positive pressure
units because of  lower  capital costs  and simpler  inspection procedures for
detecting damaged bags.4  Information collected during the  study indicates
that EAF fabric filters perform satisfactorily and that no major design prob-
lems have been identified.

5.2.2  Other Fugitive Emissions Sources

     Control  systems for  the two melting furnaces described above are well
developed and have  been applied to foundries of  all  sizes  and  locations.
Such a high degree of availability and application is not found for the other
fugitive emissions  sources  in  the foundry, however.   This lack of control
results in part from two  factors.  First,  unlike  melting operations, other
foundry processes which produce fugitive emissions vary considerably between
foundries. These  variations are  dependent  on  a number of factors including
foundry size, degree of mechanization, type of product, and size of casting
produced. As a result of these process differences, the development of uni-
form or standard  control measures is not possible.  A second reason identi-
fied during  the  study is the lack of concern about fugitive emissions on

                                  59
-------
the part  of both foundries and many state agencies.  Since fugitive emis-
sions from foundries are not perceived to be a major problem, little research
or regulatory effort has been directed toward their control.

     While  fugitive  emission  controls  are  not  as  well  developed as  melting
controls, some fugitive control can be found at almost all foundries.  Cur-
rent control is in part a result of OSHA and state health department pressure
to maintain a clean internal environment and in part results from a commit-
ment on  the part of many foundries  and  the American Foundrymen's Society
over the  last 40  years  in developing adequate  foundry  ventilation systems.

     For this study three major areas which contribute to fugitive emissions
were examined:  (a) pouring and cooling; (b) shakeout and sandhandling; and
(c) the cleaning  room.  The sections below briefly describe the control sys-
tems that are available for these three sources.  More detailed descriptions
can be found in Appendix C.  The sections describe alternate processes which
reduce the  emission  of particulate as well as capture/particulate removal
combinations which can be used on traditional installations.

5.2.2.1  Pouring  and Cooling Controls—
     Control systems for pouring and cooling operations,are the least devel-
oped of  all controls in ferrous foundries because of the large volumes of
air and  low pollutant concentrations.   During the course of this study no
well controlled pouring and cooling installations were identified even though
a significant number of foundry personnel, control  agencies, and equipment
manufacturers were contacted.  However, some systems of limited effectiveness
were identified.  The  paragraphs  below describe some of the problems that
have inhibited the development of controls for pouring and cooling emissions.
In addition the limited controls that are available are described.
                                    I
     If sand molds are used, most pouring operations are one of three types.
For large castings, pit molds are often used.   These molds are not moveable
and the  ladles  must  be moved to the mold.   Pouring occurs at a relatively
stationary point  but in a large area.   In small  jobbing foundries, floor
pouring is generally used.  In these foundries molds are placed on the floor
in a large room.  The ladle is then moved to the molds by overhead conveyor,
and the casting is poured and cooled on the floor.  In more mechanized found-
ries producing  small to medium  sized castings, pouring occurs at a defined
station.  In this case the ladle is placed in a stationary position (or along
a line)  and the molds are moved to the  ladle by  conveyor or rail.  After
the pouring  is  complete,  the molds move  along the conveyor or rail through
a cooling line or cooling tunnel.   The control problems and control  availa-
bility are dependent upon the type of pouring operation.

     For pit molds, no known capture system exists.   Because of the  size of
the operations, localized hoods cannot be used.  On the other hand, canopy
hoods or building evacuation  are  considered not economically feasible and
in all likelihood would not provide a high degree of control.  Because the
thermal rise  of pouring emissions  is significantly less than the thermal
rise of melting emissions, the volumes necessary  for control would be cost
prohibitive and might well result in such low inlet grain loadings  that par-
ticulate collection devices would have little  effectiveness.
                                  60
-------
     Capture mechanisms for floor pouring operations are limited.  Some small
nonferrous foundries have installed moveable -localized hoods that are attached
to the pouring  ladle  to capture hazardous metal fumes.  These ladle hoods
are described in Appendix C (see Figure C-19).  Based on observations during
earlier plant visits,  the hooding system visibly appears to control emissions
during pouring.   In addition, the concentration is sufficiently high to allow
good control in a fabric filter.  However, the system is limited in that it
only controls emissions during pouring.  Even small ferrous castings continue
to smoke  for  up to half an hour; and  as  the  data  in Appendix B  indicate,
about 50% of the emissions occur during this cooling period.  While systems
have been used in nonferrous foundries, no ferrous foundries were identified
which use a pouring hood such as the one described above.  Because of the
differences in  emissions  from pouring of ferrous and nonferrous castings,
this system probably has limited usefulness in ferrous foundries.

     The greatest degree of control of pouring and cooling emissions is pos-
sible when mechanized  pouring lines are used.  Several commercial pouring
hoods such as those described in Appendix C  (Figure C-17)  are available.
The most  effective  of  these systems use  a push/pull air  flow which blows
air over  the top of the molds  to contain  the  emissions and  draws the emis-
sions stream to the back of the hood.  These pouring hoods are usually coupled
with an enclosed  cooling  tunnel which also exhausts the emissions stream.
Data from the literature and from foundry personnel indicate that these sys-
tems have been applied to several foundries and that they effectively capture
pouring emissions.  However,  the emphasis of  these controls was  on the in-
plant environment, and no foundries were identified which use a particulate
collection device with the pouring  hood.  Some  foundry personnel indicated
that, particulate collection may be quite expensive with these capture systems.

     One alternative that is available to some foundries is the use of per-
manent molds  rather than  green sand molds.  Permanent molds, made of metal
or graphite emit almost no particulate during pouring and cooling.  However,
current technology  for permanent molds is limited  with respect to the size
of castings that  can  be produced and  is only economical if at least 2,000
copies of the same casting are needed.5

5.2.2.2  Shakeout and Sand Handling Controls—
     Because of the internal environmental problems caused by the silica
dust emissions  from shakeout  and sand  handling,  the capture mechanisms for
these sources are well developed and have been installed in most foundries.
The few exceptions are the nonmechanized jobbing foundries which often per-
form shakeout and  sand handling manually.  Since  the emissions  from sand
handling  can  be quite  extensive, most  of  the  installed capture systems are
connected to  particulate  collection devices.  The  paragraphs below briefly
describe the capture systems used on shakeout and  sand handling and the as-
sociated  particulate  collection devices.   The last paragraph describes an
alternative sand handling system which can be used to reduce emissions from
the process.

     Three types of hoods are  used  to  capture emissions from shakeout:  total
enclosure, side draft, and double  side draft (see Figures C-19, C-20, and
C-21).  If  the  size of casting permits,  the  preferred  capture method  is to
                                  61
-------
enclose the  shakeout operation with openings  for  the mold  to  enter  and  the
casting to exit.   Details  on flow rates  and design parameters are given in
Appendix C.   Based on visual observations during plant visits, a properly
operating enclosure is nearly 100% effective in capturing the shakeout emis-
sions .

     If the  size  of a casting or the foundry operating characteristics do
not permit the use of an enclosure, side draft or double side draft hoods
can be used  to capture shakeout  emissions.  Because the emissions from  the
plume are somewhat buoyant,  the  side draft hood is most effective if  it is
placed at an angle above and to one side of the operation.  The double side
draft hoods  should be placed as  closely  as practical on either side of  the
shakeout grate.6   The two  side draft hoods observed during the study were
not effective  in capturing the emissions stream.  In both cases it appeared
that the hoods were too small and did not extend far enough above the shake-
out.  However, Kane suggests that a properly operating side draft hood cap-
tures 90% of the emissions.7

     Both low  energy wet scrubbers  (8- to 10-in.  pressure  drop) and fabric
filters are used to collect the particulate from 'shakeout.  Because the emis-
sions stream often has a high moisture content, the scrubber is more frequently
used.  No particular design problems were identified for either control sys-
tem.  However,  if  a fabric filter is used, it  is  suggested that it  be well
insulated to avoid condensation and bag blinding.

     Once the  sand leaves  the shakeout hopper, a portion of the sand that
was near the mold/metal interface is dry, and there is a high potential for
dust emissions.  It is essential for reduction of emissions in both the in-
ternal and external environment  that the handling and transfer operations
be hooded as  well  as possible.   Hooding  systems for the various operations
are well defined and widely implemented and are not described here.   As with
shakeout emissions, particulate emissions from sand handling may be collected
in either a fabric filter or a low energy scrubber.

     An alternative  concept  (U.S. Patent No. 3,461,941) has been developed
which has the potential to control fugitive dust emissions from most sand
handling operations  other  than shakeout by reducing rather than capturing
emissions.  The process is called the Schumacher  Sand Process  System.   The
normal sand-to-metal  ratio in a  green sand foundry is between 5 and 7:1.
The Schumacher process uses  a  sand processed to metal ratio of 20:1.  This
is the quantity of sand put through the muller.  However,  the extra sand is
not used to  produce molds, but is diverted to an inundator.   Here the hot
dry sand taken off the shakeout is mixed with the moist sand from the muller
to produce a  moist cool sand.  This sand is then taken through the normal
sand handling processes.  However, the now moist sand presents no emissions
problems.                                        '

     No problems with either the design or operation of control equipment
for shakeout  and sand handling were identified during the study by either
foundry or control agency personnel.  Thus,  there  is  no evidence to  suggest
that these operations should have an uncorrectable adverse effect on foundry
compliance.
                                  62
-------
 5.2.2.3   Cleaning Room Controls—
      As  with shakeout and sand handling, no control agency or foundry per-
 sonnel  identified the cleaning room as a compliance problem.  In addition,
 the  limited  data  suggest that emissions  from the cleaning room are  less  sig-
 nificant than those  from the  other sources.   As  a result of these consider-
 ations ,  no effort was made to obtain any information on the cleaning room
 controls that was not readily available  in the literature.   The limited  in-
 formation that was obtained is summarized in Appendix C and is not  repeated
 here.

      The one factor  regarding cleaning room controls that may impact on  fu-
 ture foundry compliance  is the increased concern about the industrial health
 hazards  of the cleaning  room.  These concerns are likely to result  in pressure
 on foundries to improve  their cleaning room capture and ventilation systems.
 However, since many  of the particles generated in the cleaning room are  rela-
 tively  coarse, control should not  present a problem.

 5.3   MALFUNCTION  OF  CONTROL EQUIPMENT

      The major control problem identified  by state agencies and confirmed
 by foundry personnel is the malfunction of control equipment, particularly
 cupola  control systems.  As a result of these concerns, effort was made to
 identify the problems which led to the malfunctions and to examine  possible
 ways to  reduce the incidence  of malfunction.   MRI efforts were directed  pri-
 marily  toward cupola controls, although  some of  the findings are applicable
 to other control  systems.

      Based on a  number  of telephone contacts with foundry personnel and a
 limited number of plant visits, MRI concludes that extensive malfunctions
 of cupola controls are avoidable.   They  are a result of improper design  and,
 more frequently,  of  improper  operation and maintenance of the control equip-
 ment.   Section 5.2.1.1 described some of the design considerations  that  have
.an impact on malfunctions; this section  describes operation and maintenance
 practices that can help  reduce the incidence of  malfunction.

      The discussion is divided into three sections.  The first two cover
 proper  operation  and maintenance of wet  scrubbers and fabric filters.  Mate-
 rial is  summarized that was compiled in  a previous Environmental Protection
 Agency (EPA) study of operation and maintenance  of control devices  for iron
 and  steel processes.9  More detailed information from this EPA study is  in-
 cluded  in Appendix D.  While  these procedures were not. developed specifically
 for  the  cupola,  they were  developed for processes with similar emissions
 stream characteristics and are applicable.   The third section summarizes
 the  information  obtained from those foundries experiencing a low incidence
 of malfunctions.

 5.3.1  Operation  and Maintenance of Venturi Scrubbers

      The typical  scrubber system  associated with ferrous foundry cupolas
 consists of a gas prequencher to  reduce  the temperature of the cupola exhaust,
 a flooded disc or fixed throat venturi scrubber, a mist eliminator  with sump,
 recirculation pumps, and an  induced draft fan.   Each of these components
                                   63
-------
can. be a source of malfunctions; however, the main problems identified dur-
ing the study were fan bearing and wiring failure, feedwater nozzle plugging,
and corrosion  and erosion of the venturi  throat and mist eliminator.   It
appears that proper operation and maintenance  of the scrubber can reduce
the occurrence of these problems.  The sections below describe typical oper-
ating procedures that can be used during startup, normal  operation, and shut
down and  some  routine maintenance procedures that  can be used to  improve
equipment performance.

5.3.1.1  Operating Procedures—
     Before  initial startup, all major equipment  including fan, pumps, con-
trol and  safety  systems,  connecting  pipes,  and  utility feed systems should
be inspected and cleaned.  All fluid flow systems should  be checked for leaks
and instabilities, and an initial water test should be run on newly installed
systems to  ensure  that all items, particularly monitoring instruments and
the control  safety system, are operating properly.  After the preoperational
checks are  completed,  the system can be started using procedures outlined
in the designer's operating manual.  Typical startup procedures and preopera-
tional checks are included in Appendix D.

     During  normal operation, the operator should monitor the system to ensure
that control variables such as pressure drop, recycle pump rate, makeup water
rate, slurry density  and purge rate, and  sump  level are  operating within
prescribed ranges.  An  alarm system should be used to notify the operator
of abnormal  conditions.  An interlock system that can be  operated both auto-
matically and manually to open a bypass and shut down the scrubber in cases
of major failure should be available.

     After a melt is complete, the system is shut down again using the pro-
cedures described in the  operator's manual.  In particular, care should be
taken to flush and drain water lines and slurry lines to  reduce the possi-
bilities of  corrosion and plugging.  More complete shutdown procedures are
covered in Appendix D.

5.3.1.2  Inspection and Maintenance During Normal Opeifation--
     Many items checked before operation should be inspected during routine
maintenance; this generally includes unplugging lines, nozzles, pumps,  etc.;
replacement  of worn equipment  parts,  erosion/corrosion prevention liners,
and instruments (level indicators,  density indicators, etc.);  and repairing
damaged components (when practical from the standpoint of labor and materials).
In addition, the crossover duct between the cupola stack  and cupola should
be checked for wear  and corrosion and fan mufflers should be  checked  on a
weekly basis.

     Table 5-3 indicates  the  manpower  requirements for maintenance due to
scaling and plugging for both the wet approach and liquid injection venturi
scrubbers.
                                  64
-------
       TABLE 5-3.  MAINTENANCE FOR PLUGGING AND SCALING VENTURI SCRUBBER
                     (From interview with P. Wechselblatt, Chemico)
Type of
venturi
scrubber
Wet
approach
Liquid
injection

Type of
Plugging
Mechanical
cleaners
1 man/shift/
mo
1 man/shift/
mo
Cylinder
cleaners
1 man/ shift/
mo
1 man/ shift/
mo
problem

Scaling
Chemical
•cleaning
3 men/ shift/
wk
3 men/shift/
wk
Hand
cleaning
1 man/shift/
wk
1 man/shift/
wk
Source:  Szabo, M.,  and R. W. Gerstle.   Operation and Maintenance of Partic-
           ulate Control Devices on Selected Steel and Ferroalloy Processes.
           EPA-60012-78-037.  U. S. Environmental Protection Agency, Research
           Triangle Park, North Carolina.   March 1978.
      Table 5-4 lists maintenance  requirements  for two ranges of pressures
 and various lining materials and gas characteristics.   This table should be
 useful in the selection of scrubber liners or venturi units for the various
 iron and steel applications, including iron foundry cupolas and sand system
 scrubbers.

      The incidence of  malfunction can be reduced by periodically checking
 the system and performing,the necessary preventive maintenance.  Major items
 which should be checked are  scaling,  corrosion and  erosion of  all  internal
 surfaces, nozzle plugging or erosion, improper operation of the mist elimi-
 nators, fan balance and power requirements, and  instrumentation.   A proper
 inventory of spare parts should also be maintained to allow quick correction
 of problems identified during the check.  A more detailed inspection check-
 list and spare parts inventory are included in Appendix D.

 5.3.2  Operation and Maintenance of Fabric Filters

      The typical  fabric filter  control  system applied to  a ferrous foundry
 cupola consists of a cooling mechanism, usually  a prequencher, to  cool the
 gas stream to  about  450°F,  the fabric  filter (or baghouse)  including its
 cleaning mechanism, a  fan  which may be either upstream or downstream from
 the baghouse,  and a  dust removal system  to handle  the  captured dust.  As
 with the  scrubber system, each  of the components of the fabric filter sys-
 tem is subject to breakdowns which  can  lead to a malfunction of the entire
 system and excessive emissions from the cupola.   "Proper operation and main-
 tenance of the system  will  reduce the frequency  of  the  malfunctions to low
 levels (in some cases  1 to 2% of the operating schedule).

      This section describes  operation and maintenance  procedures  for the
 fan, fabric  filter, and dust removal systems.  The  section is  divided into
 three parts:   (a) preoperational  checks and startup;  (b)  shutdown; and  (c)
 maintenance during normal operation.

                                   65
-------






























^fc
o
a

1


C.3 \j


!? "*"*

H «
53 *H

*-j *_j
Si r~i
JC w
CO

pa o
FQ 
C? !^



H
B
0
Fm



























i





(0
jj

eu
i
o
u








4,1
•5 I
a LI
o ja
u ra
I.
O "C
U B
eo
to
u
2
C9

LI a
jj 45
u re
CO 
•H
GO
O

, . . 3



i n
t 1-4 eu
a S
0,-n
S! *"
0
en

v -
eu to
a —> u
a IM u a
0 -H >. 0
Ll ,J 0 >,
•0
eu
3
to
CO
L,
CU -H CU
S.-5
a. tS **

-
o
en

A
v \ Lt Jj O JJi^CJ*O
ocoiiie a o OJ:XE
&MC*CE}CM cu u zucdcc



£ rt*
•o /— > "5 •— v -a
tH eu ^^ eu o
.•rt > -rt > . -0 •= LIO-OT3-0
B •« E -H O O,, • -rt 00 O O O

O O CS CS. b, Ll CS CS CS

o u c LI a
O O O O CE<
cs u cs u





B
eu
^HOO O "H-H-HOO
eu o o o eoeocooo

X






E B S • e
eu cu cu ^^ eu

C O *H r* O •— "C O -ft O P-l
O O Ci CU OCU-HOL1OCU
X X X CO w X
U Cd Ed "—• Ld-

"
3 3 eu cu eu eu eu
33CM CM J3 Ji & £1 £1
•1**^<^. "v^ eocQcoeoco
E B B B>> JS JS Jl ^ Ji
a cu cu cj co uuuuu
cc BS eg eo co
CM CM CM c-4 — a< cu a. o. 3-


1
E
•H
o -» m o eu
i-. M -a V vo^jvovovo


Jf ' J£
3 3 cu cu cu eu QJ
_3^3CM^ C^ J3 W2 J3 J3 ft
BBS B JSJSJ3J5JS
euoeu cu uuuuu
CO CO CO CO CO
CN CM CM CM CUCUOjCUCU
,





^f >O NO ^jO VO \O
i i-« *-< m i i i i i
en CM CM CM CM CM






B CU rl O
O ^ JJ B OJ — <
uui-iB u o B'mo
— 4 n .a cu LI LI-H uja ooiJj

co-Hcoeu J3 jsto eucoi-<9B-H
1-4
CO U
0 Ll
Ll CO
L> S
rcu

CO
T5 E
S -rt
CO r-H
0
l-l 14
CU CO
CU U
JJ
cu o

£ jtT

eu Li
CO (3
0.
B
o eu

K M
OJ B
•J* *^
> Ll
a) s-

O L.
Ll CO
u eu
6 CO
o eu
u cc

OJ •
4J >.
IU U
i-4 B
3 CU

•m ^
4J
U G
£.2
(^4 U
O eu
eu O
U Ll
B CU
CO
B ^
CU CO
Jj JJ
B B

•e LI
s

o

JJ CO
«
Li •
cu a
a.
o
r*-
OJ O
JJ CO
Ll 1
CU CN
»s
^ ^
1
is cu
Id
•c
B
CO •
CO
. eu
• to
S 10
cu
•• a
o o

ra a.
N
CO

tt
eu
u
u
3
o
CO
66
-------
5.3.2.1  Preoperational Checks and Startup—
     Once started, the operation of a fabric filter system is often completely
automatic.  However, preoperational checks, startup, and shutdown are criti-
cal. As  with  the  scrubber,  fabric  filter operational problems  can often be
avoided with a detailed preoperational check.  This check should ensure that
the bag cleaning mechanism  (air line or shaker) and dust removal system are
operating properly.  Bag  installation should be checked and the baghouse
compartments cleaned.  Finally, all control instrumentation should be checked.
A more detailed complete checklist is included in Appendix D.

     At the first startup of the  system, and  also whenever new bags have
been installed by the maintenance  crew,  the bags should be checked after a
few hours of  operation  for correct tension, leaks, and expected pressure
differential.   Initial temperature changes or the cleaning cycle can pull a
bag loose or  burst it.  It  is wise to record at least  the basic instrument
readings on new bags during this startup period for ready reference and com-
parison during later startups.10

     During any startup, transients in the dust-generating process and sur-
ges to the filter house are probable and ought to be anticipated.   Unexpected
temperature, pressure,  or moisture has often badly damaged a new installation.
In particular, running almost any indoor air or combustion gases into a cold
filter can  cause  condensation  on the walls and cloth,  leading to blinding
and corrosion.  Condensation in the filterhouse,  in fact,  may void'the manu-
facturer's guarantee.  Condensation can be avoided by preheating the filter
or the gas.16  Another  problem associated with cool baghouses is sulfate
condensation.   Some  of  the  sulfur in the coke is emitted from the cupola
as gaseous  sulfate.  This sulfate condenses at 320°F  to 350°F and reacts
with water vapor in the gas stream to form "acid dew."   This "acid dew" can
result in both bag blinding and rapid corrosion of metal  surfaces.   Since
most cupolas  operate on an  intermittent basis, it is necessary to heat  the
cupola gases  above the  water and sulfate dewpoint in a bypass  mode during
each startup.   The filter can then be brought on line.

5.3.2.2  Shutdown—
     The main precaution  in shutting down the filter system is prevention
of moisture in the filterhouse.   Condensation can occur due to cooling of
gases containing  moisture,  particularly  combustion gases,  if they are not
completely purged from the filter system and replaced with drier air before
the filter  cools  down.  This can also happen with air  at  ambient moisture
levels if the  filter is in a colder location.   To prevent  condensation, the
systems should be purged carefully on shutdown and then sealed  off completely.
Alternately, a flow of warm air can continue to pass through the filter dur-
ing the  shutdown,  which also helps prevent condensation when the system  is
started up again.   A shutdown procedure is summarized below:

     1.  After the process has  been stopped and emissions  have  ceased,  allow
baghouse to track through one complete cleaning cycle;  this  will purge sys-
tem of process gas and collected dust.
                                  67
-------
     2.  Stop main fans. '

     3.  Stop separate reverse air fan if used.

     4.  Allow material  removal  system to operate for 1 h or until system
is purged of collected material.  This is imperative for a fabric filter on
a shakeout as the combination of moisture and binders may result in bag blind-
ing or hopper plugging if the systems not cleaned prior to shutdown.

5.3.2.3  Maintenance During Normal Operation—
     Maintenance of  fabric  filters  in the iron and steel industry centers
around the bags  and  the  moving mechanical parts in the hostile interior of
the baghouse  (i.e.,  dampers,  screw conveyors, and shaker linkages).  The
same maintenance procedures can  be applied to  baghouses operating on elec-
tric arc furnaces or cupolas in ferrous foundries.

     Plant personnel must learn to recognize the symptoms that indicate po-
tential problems in their fabric filter, determine the cause of the problem,
and remedy it either by in-plant action or by contact with the manufacturer
or another outside resource.

     For example, high  pressure drop across the system is one symptom for
which there  could be many causes, e.g., difficulties with the bag cleaning
mechanism, low compressed-air pressure,  weak shaking action, or loose bag
tension.  Many other factors can cause excessive pressure drop, and several
options are usually available for corrective action appropriate to each cause.
Thus, the ability to locate and  correct malfunctioning baghouse components
is important and requires a thorough understanding of the system.   A detailed
list of- troubleshooting  and corrective measures is given in Appendix  D.

     T-able 5-5 presents the frequency of failure of basic fabric filter parts,
including the frequency  of  inspection, the inspection time, and the time
required for repairs.

     Some of the major fabric filter components requiring routine maintenance
and the problems frequently encountered include:

          Inlet ducting  - abrasion,  corrosion, and plugging of the duct.

          Blast gate and flow control - Hydraulic  system  failures  and bad
          seals.

          Fans - Wear from corrosion or abrasion,  balance,  and bearing fail-
          ure.

          Hoppers - Plugging caused by caking or bridging of dust.

          Bags - Collar wear,  poor tension, burnout.

          Shaker mechanism - Bearing failure, improper amplitude or frequency.

          Reverse air mechanism - Line blockage from moisture.

Maintenance procedures  for  these components  are described in Appendix D.

                                  68
-------


V.
4™
d
Ol


o
CJ



T) S-l
<*-) d -H
0 0 to
CO &
Ol w CU
Q* CJ ^
r^» OH
°
•H
CO
P.
01
to

O
M cu
cj s
S ^
EH d
£25 o
H-( -H
^4 4^
2 -o o
01 0)
O -H CO
EH 3d
CJ fH-H
w oi
O 01 0
cj s  co
H cu d
^£ &pj
EH
O
>» 1
U 0
d TU
Ol **t
3 to
rH > S
O CO
CJ CM CO
d d d
CQ CQ CQ
"to ESS
01 ...
S-l 4J 4-1 4-1
o d d d
XI . -H -H >H
tO CO to CO
r-3 O 0 u 22 2
00 00 00
(Q CQ (Q
X> X) Xi
d d d o ca
•H -H -H to S-l to
S S S 0 -^, ca . . 3
o d M o
o m o "o -H 3 xl
co --d" vo ^**. s o
1 .1 1 XI XI «*
o m o o i
rH rH CM rH CO CM CM
CO CO
co ca
00 CO 00
CO CO CO
Xl OOXl
CO
O Xl O
o o
rH O rH
f Cf
d xl d o 3
CO 1 to . O O
S d S *a >*
,
CO
PH O
o S
4-1 Ol

01
S 00
3 to
ca xi
ca

d d d d d d
CQ CQ CQ CT3 CO A
S S S £ E S
.... QJ . .
4-> -4-> 4-1 4-1 S-l 4-) 4->
d d d d o d d
•H -H -H -H XJ -H -H
2^ S 2 r3 ^^

00
CO
ca xi
co to ^,
to 3d
3 ca co o -H
£ S g * s -S g
7^ ^ 7 7 o*
rH OO OO rH VO CO rH



CO
00
CO
Xi
_
CU 0
^ c.
to xl 33
> I O O
-•^ fi i-i u
_g "i-'i'1- s-idi4dd to, -^,-v^
co s 10 d d -H 3 -r-i 3 -H 'H E d d
A vr A -g "g e 5 S ^ S S vJ 11
i, • in o o m .
rH CMCM in lOrH rHCOrH rr) p^ ' ' ' O CMCM



r-l rH rH rH " rH rH
Xi Xl XI XI XI XI
4-> 4-) 4-> 4-) 4-> 4-> O
d - d d d d d s
O 00 0 0 O
2222 22 vo v



>> >> >>
rH rH r-4 J^ to
X! X) Xl S» >,
4J 4J +J O S-l



>> >» >» >%
	 xl • ' 3 xi xl
0 0 M 4J tl S 5
s s >> d d d d
o o o o
O ^O rH ^J J2J jgj yj




S-l ^4 1-4 rH

Lj i i t_i t_i
ddd> CMCMCO^CO d >,>,
OOOI III 1 S r-'K-'
2 22 CM"\OCN rH,

o
EH S §
U CO 4-> CO
P3 -H ca to
t-fl 4J 4-> 4-> d >> O
HH *W H-! H-4 CO CO ^»
O xl to 01
CJ 
rH CM CO 0) CO CO d
 to
rHCO !-IBJFdCU4->CJ-
co oqin OOCM o oicoes coco-.
13 tO - rH O rH 3
rH P3| Q CJ Q
H CM vr CM 2 CM CM
S5
O
rH
EH
SCO
•H g

o to d
CO CJ Xi CU
0)0) CJ rH
COCJ 4-> >O OI4-1P-
irf-r-ICOi— 1< SO)
CJ4JrHCOP3 -r-jOO
O eo PL, > CO OO d
HE W-00 dCU-H
3 CU S-l O (Q -HCOCO
MCUrH CU r-l -XI drHrH
HdM-lftCO 1033
«!P-I "4-t S EH rH Olp-iCM
0} tO J3 rH rH
'« Q O < CJ
                                                         cu

                                                         'd
                                                         •H
                                                         4->

                                                         o
                                                         CJ
69
-------
                 r-   S
                                                          
                                                          4J W i-t   0)
                                                          (0          S
                                                          H   • Jd  -H
                                                          s en  u  4->
                                                                             CO
                                                                     CO
                                                                         co
                                                                         oo
                                                                      cu  r>~
                                                                      u   i
                                       CO

                                       £

                                   •H 43
                                   rH  U
                                   o  d
                                                                             0)
                                                                             00

                                                                             CO
                                                                             43
                                                                             U

                                                                             (SO
                                                                      d  CM  «
                                                                      CO  r-l CJ
                                                                      d  o
                                                                      0)  O 43
                                                                         VO
                                                                     •H
                                                                   -  00  >EH
                                                              O     O  r»  T3
                                                          '   rH 43   •   U   d
                                                           • rH  O CO   CO   CO
                                                          0)  CO  !-l     rH
                                                          rH  O  CO   "43   00
                                                                             (U 43
                                                                                         CO
                                                              S-l  W  CJ  T3  43
                                                              (U  -»
                                                             fn PS  (U   CO  4J
                                                                     'M      3
                                                                          ca  o
CS   CO  CJ  U rH  00
        (3  d  U  S
    rH   4J   S  0)
                                                                           4-1  d  O
                                                                         4J  O
                                                                                     -i
                            OrH  O43     (U      CO
                                          "4-1      
                            U           •      O  43  O
                                                                                      43   CJ  13
                                                                             70
-------
5.4  FOUNDRY, EXPERIENCE WITH OPERATION AND MAINTENANCE OF CONTROL EQUIPMENT

     Since most  state  agencies and some industry personnel indicated that
malfunction of control equipment was the major foundry control problem, one
facet of  this  study was to contact well-controlled  foundries  to  identify
the types  of procedures they used to minimize malfunctions.   In  order  to
identify  exemplary  foundries,  13 state and local agencies were contacted.
These agencies identified 36 foundries that had exemplary control.  Of these
36 foundries, 30 were contacted by telephone and six which appeared to have
well-developed operation and maintenance programs were visited.

     This section summarizes the results of these contacts.  The section is
divided into three parts.  The first part presents an overview of the results
of the  survey.   The next two parts describe practices at two plants which
have had particularly good success in avoiding malfunctions.

5.4.1  Conclusions of the Foundry Survey

     A limited number of foundries were contacted for this survey and a wide
variety of responses were obtained.  Therefore, it was not possible to develop
general conclusions about technology availability, equipment design problems,
or equipment maintenance procedures.  (For example, the number of hours spent
on preventive maintenance ranged from 0 to 160 h/week.)  The discussion that
follows is  a digest of the information gathered  from these foundries.

     •  The majority of foundries disclaimed the need for improved technology.
However, the following concerns were addressed: (a) a need for more data on
the possibilities of combining gas streams from two processes  (e.g., hot
cupola gases with cool shakeout gases) in a single collector; and (b) a need
for a cupola control system that would allow heat recovery for space heating.
This same  concern has  been raised related to the use of heat  in  captured
fugitive streams.

     •  Malfunctions of  control  equipment  are frequently a result of lack
of training of  operators in the proper operation and maintenance, or the
inability to motivate workers to properly operate and maintain the equipment,
rather than the result of inherent problems in the process or control equip-
ment system.

        Generally,  larger  foundries practices  greater surveillance of the
control system during operation and do more preventive maintenance. However,
one of  the  best operated and maintained cupola control systems identified
during the  study was at a small jobbing foundry that only melts  3 h/day.

     •  At least one foundry contacted considers the operation and mainte-
nance requirements of the equipment in comparison to the capability of foundry
personnel as a  factor in choosing both the control equipment and  equipment
supplier.

     •  Maintenance procedures usually evolve in-house and make use of ini-
tial input from control device suppliers.   The degree of follow-up help varies
widely among different suppliers.  Maintenance people are generally trained
                                  71
-------
on-the-job although,  some  larger  companies with multiple  facilities have their
own  training centers and some obtain help  from  the Cast Metals Institute.

        The  sources  of malfunctions  frequently identified are:  (a) plugged
water  spray  nozzles  on both wet scrubbers and evaporative  coolers;  (b)  bag
blowout and  overheating;  (c) fan bearing failure;  and (d)  corrosion of  the
throat and mist eliminator. Each of these  problems  can be minimized with
the  proper  design, operation, and maintenance  of  the control equipment.

        Some of the  features at  the  foundries visited during the study which
might be incorporated to  minimize malfunctions are:   (a)  automatic air dilu-
tion and filter by-pass controls to  avoid bag overheating;  (b) periodic check
of all fan motors with an amprobe to detect possible failures; (c) periodic
calibration  of monitoring and control instrumentation;  (d) daily check of
bag  hoppers  and weekly to biweekly  internal  checks  of  bag condition;  (e)
periodic cleaning of  quencher spray nozzles; and  (f) post-startup and shut-
down procedures in operating room similar to those described in Appendix D.
It is  apparent that the  practices in  these foundries include  many  of the
elements described in 5.1 and 5.2.

     *  Most contacts  indicated  that stainless  steel  parts  in  the scrubber
throat, mist eliminator,  and fan are essential  if malfunctions are  to  be
minimized.

     The results presented  in the above paragraphs  lead to the following
conclusions:

     1.  Malfunction of control equipment can be minimized  (to as little as
1 to 2%) through proper operation and maintenance.

     2.  Training of personnel  is a prerequisite  to proper operation and
maintenance.

     3.  Such training is available  through some vendors and,  if not avail-
able from the vendor, can be supplied by the Cast Metals Institute.  The foundry
should consider the cost  of such training as a part of the cost of the control
equipment package.

     These conclusions (especially item 1) are supported by information pre-
sented in the following sections.  These sections describe successful oper-
ation and maintenance  practices  at  two foundries of different types.  The
first is a  small jobbing foundry which has a cupola with a fabric filter,
that operates 3 h/day.  The second is a large production foundry,  again with
a cupola and fabric filter.

5.4.2  Operation and Maintenance on  Cupola and Shakeout Controls  at  a Small
         Jobbing Foundry

     One of the six foundries visited was a small jobbing foundry that pro-
duces primarily grates and manhole covers for streets.  Although the foundry
was  the smallest and least  mechanized that was visited  during the study,
                                  72
-------
 the  preventive maintenance program for  the  control  equipment was more  ex-
 tensive  than at any of the other foundries.  -The paragraphs below describe
 the  foundry control  equipment  and operation and maintenance procedures.

     The  shakeout had a side draft hood which vented to a pulse-jet fabric
 filter.   Based on visual observations  during the visit,  the  hood adequately
 captured  the dust from  the  shakeout; no  emissions were visible  from the ex-
 haust.  The  system was  designed with an  easily accessible manometer and pulse
 indicator to allow frequent monitoring.  Key elements  of the  maintenance
 protocol  include  these  steps:

     •  Pressure  drop,  fan, and rotor  lock are checked daily.

        Bags are  checked and manually  cleaned every  3 weeks.

     •  The  bags  are pulsed manually each morning before system startup to
 eliminate blinding.

     •  The"only  spare part inventoried  is a spare motor for the screw con-
 veyor which has caused problems in the past.  One-hour service  is available
 on the fan motor.

     The  muller has a shaker-type fabric filter.  There were no visible emis-
 sions at  either the muller or the stack.  The maintenance protocol was similar
 to that on the shakeout.

     The  control  system for the  cupola included an evaporative  chamber  for
 gas  cooling  and a Pangborn shaker-type  fabric filter.   The  system  design
was  conducive  to  good preventive maintenance with easily accessible moni-
 toring instrumentation and a double backup system for the evaporative cooler
 to avoid  overheating of the bags.  Features of the system which aid in pre-
ventive maintenance include:

          Automatic temperature-controlled damper which admits dilution air
          to the  exhaust stream  if the evaporative cooler exhaust is above
          450°F.

          Automatic temperature-controlled damper system which results in
          bypass of the fabric filter if inlet gas temperature is too high.

          Pressure and  temperature guages  located  inside the foundry in a
          control room for easy monitoring.

          A flow  meter  on the evaporative cooler water line  located in the
          control room.  A  booster pump is  used to  increase flow in case
          water pressure on the city line drops.  "

     The  maintenance  foreman monitors  the cupola emission control  system
continuously during melting  to  ensure  proper operation of the  equipment.
The foreman can manually operate the dampers and booster pump described above.
In addition, a rigorous preventive maintenance program is followed to ensure
continual compliance.  Features of the  maintenance program include:
                                  73
-------
          Startup procedures include heating of the system and manual shak-
          ing of all bags before melt each day.

          Fan and  grease  bearings are checked daily before start of melt.
          Fan is checked periodically during melt.

     *    Honeywell calibrates  electrical  controls and monitoring devices
          once every 3 months.

     •    Nozzles in the spray chamber are cleaned monthly.

          Bags are checked for leakage, once every 2 weeks by injecting agri-
          cultural lime into  the  exhaust duct  30  ft upstream of  the  fabric
          filter inlet.

     The foreman agreed that  the  procedures described  above are  expensive.
However, he felt that the procedures are cost effective from the standpoint
of equipment  life  and minimized downtime.   During the time the system has
been in  operation  (about  2 years)  the only downtime was caused by a  frozen
air line which locked the inlet to the fabric filter closed for two shifts.
This is  judged  to  be important as the company is  fined $150 per day when
operating with  the filter down.  The procedures have also resulted in ex-
cellent bag wear with the lifetime expected to be 2-1/2 to 3 years.

     The foreman did indicate that they have had problems with shrinkage of
the Nomex bags; some bags have shrunk as much as 5 to 6 in.

5.4,3  Operation and Maintenance on a Cupola Fabric Filter at a Large Foundry

     Another foundry visited was a medium- to large-sized foundry which melts
on one shift at the rate of 30 ton/h.  In contrast to the foundry described
above, this foundry  tended  to react to problems rather than to spend time
on preventive maintenance.  However, the process and control system is care-
fully monitored, and  maintenance  is performed  quickly  to  limit downtime to
about 1-1/2%.   The paragraphs below describe the control equipment and mon-
itoring practices at this foundry.

     The effluent  stream  from the cupola passes  through  a conical bottom
chamber, gas-fired afterburners, air/air heat exchanger, two water prequench-
ers and into the baghouse.  The heat exchanger heats ambient air to approxi-
mately 900°F  at 11,000 CFM, which is recycled  to  the cupola for  hot blast.
No space heating is attempted. The water lines to prequenchers are equipped
with filters  to avoid plugging the nozzles.   The cupola gas stream enters
the baghouse at approximately 500°F and is filtered through fiberglass bags.
Bags are replaced as needed (approximately five bags/week) and can be changed
with the system on-line by shutting off a given .baghouse segment.  Mechanical
shakers clean the bags approximately every 3 minutes.'       •  -

     The control room .consists of two large panels.  One panel is a schematic
flow diagram for the system showing the location of numbered thermocouples.
The second panel consists of a circular chart'recorder for flow rate, a se-
lector switch and digital readout for the thermocouples, and a bank of lights,
                                  74
-------
each labeled with  a  given operation segment.   In  the  event of trouble, a
warning buzzer  sounds  and the appropriate light flashes  to  indicate the
source of the problem. Amp meters for the various motors are also displayed
on the panel, which is under constant surveillance by the operator.

     The system is designed in such a manner that both broken bags and mal-
functioning nozzles can be changed with the system on-line if a malfunction
occurs.             ••-
                                  75
-------
1.



2.


3.



4.
5.


6.



7.

8.
                                REFERENCES

     Davis, J., E. Fletcher, R. Wenk, and A. Elsea.   Final Report on Screen-
     ing Study on Cupolas and Electric Arc Furnaces  in Gray Iron Foundries.
     Battelle Columbus Laboratories, Columbus, Ohio.   August 1975.   pp.  IV-21".
     American Foundrymen's Society.
     114.
Cupola Handbook, 4th Ed.  1976.  pp.
     Electric Arc Furnaces  in  Ferrous Foundries - Background Information
     for Proposed Standards  (Rough Draft).   U.S.  Environmental Protection
     Agency, Research Triangle Park, North Carolina.   April 1980.   pp.  4-35,
     36.

     Fennelly, P. F.,  and  P.  D. Spawn.  Air Pollutant Control Techniques
     for Electric Arc Furnaces in the Iron and Steel  Foundry Industry.   U.S.
     Environmental Protection Agency, Research Triangle Park,  North Carolina.
     Publication No. EPA-450/2-78-024.  June 1978.  pp.  221.

     Bates, C. E.  Profit Potential in Permanent Mold Iron Castings.  Foundry.
     November 1972.  pp. 49-52.

     American Conference of Governmental Industrial Hygienists.   Industrial
     Ventilation.  A Manual of Recommended  Practice, 16th Ed.   1979.   pp.
     5-15.

     Kane,  J. M.  Air Pollution Ordinances.   Foundry.  October 1952.

     Design of Sand Handling and Ventilation Systems.  American Foundrymen's
     Society, Des Plaines,  Illinois.  1972.

 9.   Szabo, M., and R.  W. Gerstle.   Operation and Maintenance  of Particulate
     Control Devices on Selected Steel and Ferroalloy Processes.   EPA-60012-
     78-037.  U.S. Environmental Protection  Agency.  Research  Triangle  Park,
     North Carolina.  March 1978.

10.   Billings, C. E. and J. Wilder.  Handbook of Fabric Filtration Technology,
     Volume I.  Prepared by GCA Corporation  for National Air Pollution  Control
     Administration.  Contract No.  CPA-22-69-38.  December 1970.
                                  76
-------
          6.0  STATE AND LOCAL AIR POLLUTION CONTROL REGULATIONS
                 AND IMPLEMENTATION POLICIES
     An important  aspect  of an analysis  of factors  affecting compliance  of
ferrous foundries is the identification of applicable regulatory strategies.
These include both legal requirements and enforcement policies; as they re-
lated intrinisically to the structure of those state and local agencies with
major enforcement  responsibility for  foundries.   One  subtask of this  study
comprised an  analysis  of  both the regulations  available  to  state and  local
agencies and the implementation policies of these agencies.

     This subtask was directed to two specific objectives:   (a) the identifi-
cation of state  and  local  regulations that establish  emissions limitations
for ferrous foundries  (in particular particulate emission limitations for
cupolas, electric  are  furnaces,  pouring  and cooling operations, and shake-
out and sand handling); and (b) the identification of state and local imple-
mentation policies with specific reference to enforcement problems and solu-
tions used by enforcement  officials in applying  regulations.  These objec-
tives were addressed through  three basic activities:  a  survey of relevant
regulations for all states and selected local agencies;  a simultaneous tele-
phone survey of agency personnel assigned implementation responsibility for
those regulations; and a subsequent analytical phase to identify how agencies
actually apply regulations to ferrous foundry processes, areas that agencies
perceive to be problems in apllying the regulations, and solutions to these
problems.

     The results of  these  activities are summarized in four sections.  The
first section identifies  the  regulations that  are available and the degree
to which they are applied.  The second section is a brief discussion of the
strengths and weaknesses of these regulations.   The third section identifies
some major problem areas and, when possible, describes solutions identified
by some agencies for these problems.   The final section presents some issues
raised by the surveys which deserve further study.  Each of these topics is
discussed in greater detail in Appendix E.

     The data presented below should be examined and used-with some caution,
as they result  from  single telephone contacts with agency personnel.   The
conclusions presented  in  these sections  are based on the authors analyses
of the data compiled during these contacts and were not tested through fur-
ther contacts with either control agency or foundry personnel.

6.1  REGULATIONS APPLIED TO PARTICULATE  EMISSIONS FROM FOUNDRY PROCESSES

     Two distinct  aspects  of emissions  control must be addressed by state
and local regulations applied to ferrous foundries.  First, the regulations
                                  77
-------
must  result  in,the  installation of  appropriate  control equipment  on  foundry
emissions  sources.  Second,  the regulations must be  capable  of  ensuring  oper-
ation of the installed equipment  in a manner  that  results  in continued com-
pliance.   In general, agencies surveyed  felt  that  regulations were adequate
and had been used sufficiently to require  the installation of all necessary
control equipment.   Statements  about the  status of  foundries with respect
to continued compliance were not  as positive. Several states identified  mal-
function of  foundry control  equipment as a major regulatory  problem.

      State and  local regulations  used to require installation and operation
of controls  for ferrous.-. foundries vary widely both from state to  state  and
within states. - Factors  such as  foundry location,  type  of process,  foundry
size, date of startup,  available legal  structure, and previous experience
with  the public, the foundry industry,  and the  court systems all  influence
the choice of regulation.

      The four general  types of regulations used to  control  emissions from
foundry processes which  are surveyed during  the  study  are mass emission,
visible emission, fugitive emission, and nuisance-related  regulations.   Other
important  regulations identified  during the study but not surveyed  are  the
malfunction  regulation operation  and maintenance (OSM) regulations, and  oper-
ating permit regulations.  These  regulations  may be  used individually or in
combination  to  ensure appropriate control  of  foundry emissions.

      The paragraphs below briefly describe each of the major types of regula-
tions.  These descriptions are followed by a  summary of the  availability of
the regulations surveyed  to the  states and the degree to which the states
apply the  regulations.  More detail is provided in Appendix  E.2.

     Mass  emissions regulations  specify the  quantity of  particulate that
can be emitted  from a foundry process  or group  of  processes. As  such they
allow for  the relatively precise prediction of particulate emission necessary
for the calculation of  air  quality impact and, thus, are  valuable in the
development  of  (SIPs) for the attainment and  maintenance of  National Ambient
Air Quality  Standards (NAAQS).  Three major types of mass emissions regula-
tions were identified during the study: (a)  the process weight regulation
which limits  the total mass  of hourly emissions based on the hourly raw ma-
terial input; (b) the concentration regulation which limits  the mass of par-
ticulate in  a specified volume of undiluted gas; and (c)  the removal  effi-
ciency regulation which  specifies the efficiency that must  be attained by
the control  device on a foundry process. .  Some of the regulations  identified
were written for general industrial processes while  others were written for
specific foundry processes.

     Visible  emissions (VE)  regulations  generally  limit the  opacity  of the
emissions plume.  (Opacity is the degree to which the plume  limits an observer's
view of the background.)  Unlike mass emissions  regulations,  VE regulations
cannot be  used  to limit precisely the quantities of  particulate emitted  to
the atmosphere  and  hence are not  as valuable  in developing SIPs.  However,
VE regulations have the  advantage of being more  easily and economically en-
forceable  than  mass  emission regulations.   This is especially true  in the
ferrous foundry  where many of the emissions are fugitive and are  difficult
to test precisely.

                                  78
-------
     Fugitive emissions regulations differ from the above categories in two
major respects.  First, they often involve no specific, quantitative stand-
ards; rather, they invite the discretion of the responsible agency official
to determine the  levels  of fugitive emissions and control to prevent such
emissions that are reasonable in a given situation.  A second difference is
that fugitive emission regulations vary according to significantly different
models: some are preconditioned on the determination that a nuisance exists;
some are preconditioned on the determination that ambient concentrations at
the property line exceed an established limit; others  require that  reason-
able precautions be  taken  to  prevent any fugitive emissions; and finally,
some restrict visible emissions at the property line.  Because fugitive emis-
sions are an air pollution problem commonly associated with foundries, fugi-
tive emissions regulations may play an important role in enforcement strate-
gies .  They may be used in establishing in-plant capture systems for fugitive
total suspended particulates  (TSP),  as well as aiding enforcement  of the
performance of those in-plant capture systems.

     Nuisance-related regulations have their basis in common law.  The three
general types of nuisance regulations surveyed during the study were: (a) a
general proscription against emissions that harm persons or property; (b) a
proscription against air pollution which causes a nuisance; and (c) regula-
tions which  proscribe  air pollution  causing odors.   These  nuisance-related
regulations have proven  helpful in some states because  state courts have
acted favorably on action brought under these regulations due to their his-
torical basis  and also because they provide an avenue for action based on
citizens' complaints.
                               %
     Malfunction regulations were not originally included for consideration
in this  study.  However, it quickly  became apparent  that these  regulations
are a two-edged sword which might be used on the one hand to excuse excessive
emissions as a malfunction, and on the other hand to better ensure continuous
compliance at ferrous foundries.  Essentially, most malfunction regulations
require that a source report a malfunction (defined differently in different
states) to the local agency.  The foundry must then present a plan for cor-
recting (and sometimes preventing) the malfunction.  Much greater detail is
provided in Appendix E.2.6.

     The paragraphs  below  and Table 6-1 summarize the degree to which the
states apply the regulations described above.

     Mass emissions regulations applicable to ferrous foundries are primarily
process weight  and  concentration  (grain loading)  limitations.   Forty-three
states have  process  weight regulations,  and 23 states have  concentration
limitations that are or could be applied to one or more of the ferrous foundry
processes.  Only two states (New Mexico and Utah) have no general mass emis-
sion limitation that would be applicable to ferrous foundries.

     Of the  43 states  with process weight regulations, 41  have  limitations
that apply to sources in general, including foundries, but only 14 have limi-
tations that apply  specifically to foundry operations  (usually  the  melting
process).  Similarly,  of the  23 states with concentration limitations, 19
have regulations that apply to  sources in general, including foundries, but
only  9  have regulations  that  apply  specifically to foundry processes.

                                  79
-------
TABLE 6-1.  APPLICATION OF REGULATIONS TO FERROUS FOUNDRIES
States where States
Type of authority not inter-
authority exists viewed
Process weight
Regulations
Grainloading
regulations
Visible emission
limitations
Fugitive emission
limitations
Nuisance-related
authority
General pro-
hibition
Odor
Nuisance

43 7

23 . 3

50 9

43 9

38 6

20 - 4
25 3
16 1
States inter- States
viewed with which use
no foundries authority

4 31 (97%)

1 18 (95%)

5 36 (100%)

4 23 (77%)

3 11 (38%)

1 5 (33%)
2 5 (25%)
2 5 (38%)
                           80
-------
     Mass emissions regulations have generally been used to require installa-
tion of  control  equipment,  particularly on the melting furnaces.  Because
of the cost  of emissions testing, mass emissions regulations are not used
frequently to ensure continued compliance.

     Virtually every state  has a visible emissions regulation  that would
apply to foundry operations; and a large majority (43) have a fugitive emis-
sions regulation that  could be applied.  However, very  few  if any of  these
regulations are specifically designed to regulate foundry emissions.   Fewer
states (37)  have nuisance-related authority (nuisance,  odor, or a general
prohibition against air pollution) specified in their air pollution statutes
and regulations  (although  it  is  possible that additional nuisance-related
authority exists elsewhere  in the state code).   This authority also seems
designed to  aid  in the regulation of air pollution sources  in general, but
not foundries specifically.

     VE regulations have been used both to require the installation of con-
trol (particularly on shakeout and sand handling) and in the enforcement of
continued compliance.  In fact many of the states use the drive-by VE inspec-
tion as their primary enforcement tool.  Little information was gathered on
the specific uses  of nuisance-related authority.  However, theoretically
this authority can be  used to require installation and regulate continued
compliance.

     Since the malfunction, O&M,  and operating permit regulations were not
included as a part of the original survey, data on the extent of their avail-
ability and  application  are not available.  However, it is  known that in a
few states,  these regulations  do  form  the backbone of continual compliance
efforts.

6.2  STRENGTHS AND WEAKNESSES OF VARIOUS TYPES OF REGULATIONS

     Each of the regulations described has certain strengths and weaknesses
when applied to the foundry processes that were identified during the tele-
phone survey.  The paragraphs  below briefly summarize these strengths and
weaknesses for four basic types of regulations:  process  weight,  concentration,
fugitive emissions, and visible emissions.

     As a mass emissions standard, the process weight regulation was  perceived
by respondents to have two  advantages:  (a) it establishes a fixed quantity
of allowable particulate subject  only to a change in the process weight;
and (b) it varies stringency with the size of the source.   This second factor
is particularly important for the foundry industry where many small jobbing
foundries have a very low profit margin and cannot afford sophisticated con-
trol systems.

     On the  other  hand,  two major criticisms were leveled  at the process
weight regulation.  Foremost was  the practice of applying a single process
weight curve or table to a wide variety of processes.   The intrinsic  differ-
ences in quantities and controllability of emissions may result in an inequi-
table burden on  some industrial  categories.   A  second  criticism  that is
particularly true of  foundries is the difficulty of determining the  input
                                  81
-------
weight.  For example is the shakeout input weight simply the weight of cast-
ings or is it the combined weight of castings and sand.
          *
     The advantage most  often cited for concentration regulations is that
they rely only  on measured  emissions and do  not  depend on determination of
input weights.  As a result it is easier to determine compliance on many of
the fugitive  emissions  sources.   In particular, emissions from roof vents
and ventilation ducts can be monitored.

     The two  major criticisms of the concentration  regulation are  that the
regulations do not generally vary allowed emissions according to source capa-
city (a  few states  do have such variance) and that the  regulation may be
subject to circumvention.  The importance of the first criticism, especially
with respect to jobbing foundries, was described above.  The second criticism
is based on the assertion that it is possible to infiltrate large quantities
of air and circumvent the regulation.  However, it  seems that if the process
were properly monitored during testing, the potential for such circumvention
is slight.  Survey respondents,  in fact, did not indica'te that this potential
problem actually occurs.

     In  discussing fugitive emissions  regulations,  respondents had  fewer
strong statements about the strengths and weakness  of these regulations than
they had about the mass emissions regulations.  This is probably the result,
at least  in part, of a  lack of concern about fugitive  emissions by many of
the respondents.  For those persons concerned about fugitive, emissions, the
primary advantage of the regulation is that it often provides a better vehi-
cle to  control  fugitive sources than either mass emissions regulations or
the visible emissions regulation.

     The major disadvantage of most fugitive emissions regulations is their
subjectivity.  This subjectivity invites dispute on the part of the industry
and makes violations much more difficult to prosecute.  Although many states
have retreated from enforcement  of such regulations because of their subjec-
tivity, other states have attempted to reduce this  subjectivity by providing
more objective criteria.  Fourteen states, for example, prohibit any visible
emission  at the property line;  four states prohibit any fugitive emission
that exceeds 20% opacity; and six states establish  ground level ambient con-
centration  standards.   In addition,  several  states  shift the burden  to the
source to demonstrate the need for fugitive emissions  or the reasonableness
of the  in-plant control equipment installed, thus  providing the  state an
edge in any dispute that may  result.  Another disadvantage of fugitive emis-
sions  regulations that  require property line measurements,, is  that  these
measurements  are  difficult to obtain and consume a large amount of agency
resources.

     As  indicated earlier, visible emissions regulations form the basis for
ferrous  foundry regulatory  activity in many  states.  Visible emissions regu-
lations are often used because they are more easily applied than other regu-
lations available to the agency.  Visible emissions "tests" can be performed
with almost no planning and with a minimal commitment  of resources.
                                  82
-------
      Survey respondents  had two  major criticisms  of visible emissions  regula-
 tions as  applied to  foundries.   First,  it is  likely that a visible emissions
 standard  (especially a 40% opacity standard)  is significantly less restric-
 tive  for  emissions  stacks  than  is a mass emissions standard.  Thus, sole
 reliance  on the  visible  emissions  standard results  in an overly lenient en-
 forcement posture.   The  second disadvantage of the  visible emissions stand-
 ard is that it often does  not adequately apply to foundry fugitive sources.
 Since the emissions from these sources exit from a large number of windows
 and doors,, acceptable method 9  readings are  often  not possible,  and it is
 not likely that  the  emissions from any one exit will exceed the opacity limit.

,6.3  SELECTED PROBLEMS  AND SOLUTIONS INVOLVING FERROUS FOUNDRY REGULATIONS
        AND THEIR APPLICATION BY  STATES AND LOCALITIES

      One  major objective of  the regulatory analysis in this study was to
 identify  solutions  to particular regulatory problems that inhibit effective:
 enforcement of ferrous  foundries.    Most state contacts reported,  however,
 that  there are few,  if any, significant problems  encountered in the regula-
 tion  of ferrous  foundries.   Foundries thought generally to be in compliance
 with  applicable  emission limitations; when out of compliance they have been
 willing to comply voluntarily; and the development  of specific investigation
 or enforcement strategies  to deal  with ferrous foundries has been considered
 unnecessary.   While  most survey  respondents could recall isolated instances
 in which  compliance problems had  occurred, few were willing to state  that
 theoretical problems in the  applicability or effectiveness of particular
 types of  regulations were  to blame.

      In contrast, a  few  respondents  related their concern over a multiplic-
 ity of compliance and regulatory problems involving ferrous foundries.   Fed-
 eral  regulators  and technical experts  associated with private consulting
 and engineering  firms also expressed-many of the same concerns.  This dis-
 crepancy  in opinion is  not fully  explainable.  In  some cases,  such as the
 control of fugitive  emissions, it  is conceivable  that the majority of  state
 regulators do not yet recognize compliance problems that must be addressed
 in the future and,  therefore, have  not yet experienced certain regulatory
 and enforcement  strategy problems  that are inevitable.  In other cases, it
 is possible that certain matters  are not considered problematical because
 the problems  have already  been resolved.

      Whatever the case,  it  has  been possible to  identify a limited number
 of regulatory and strategic  problem areas of actual or potential signifi-
 cance and explain how these have actually been resolved by some states.   As
 a general observation, problems  tend to fall,  into the following broad  cate-
 gories:  (a)  the unavailability of an appropriate regulation; (b) problems
 involving the type  of emission  limitation included in the regulation;  (c)
 problems  involving vagueness or  overbreadth in the  regulation;  (d) problems
 involving a lack of adequate resources to implement regulations; and (e)
 problems  in the  design of  surveillance or enforcement strategies.   Solutions,
 also, tend to fall  into  broad categories:   (a) the  adoption of new types of
 regulations;  (b) changes in existing regulations; (c) reliance on alterna-
 tive  regulations;  (d) reinterpretation of existing  regulations;  and (e)  formu-
 lation of new surveillance or enforcement strategies.
                                   83
-------
           These problems  and solutions  are identified and discussed briefly
 in the  following paragraphs.   It should be noted that certain of the problems
 may be  less prevalent or less substantial and certain of the solutions may
 be less desirable or even  inappropriate in particular states.   No attempt
 has been made to analyze, or evaluate either the problems or the solutions
 presented in this section.

 6.3.1   The Process Weight Regulation Problem

     Major problems  voiced  and solutions identified  during the  survey  include:

 Problem;   The rate  was not designed for foundries and therefore results in
           an inequitable  burden on specific segments of the foundry industry.
           Specifically, certain small or intermittently operating foundries
           must meet  limitations that are too restrictive;  and certain  large
           foundries  escape  with a more  relaxed emission limitation.

           Solutions  Mentioned

           1.   Adopt  regulations specifically designed for ferrous foundries.
 These may be process weight rates (see,  for example,  Pennsylvania's regula-
 tion. Appendix E, Table E-3)  or concentration limitations  (see,  for example,
 Michigan's regulations, Appendix E,  Table E-3).

           2.   To  prevent  certain foundries from  taking advantage of a  more
 relaxed mass  emission limitation,  as  well as  to  relieve  the burden of  a more
 stringent mass emission limitation, a collection efficiency regulation may
 be  superimposed  (in  the first case,  to  apply if  more  stringent;  in the sec-
 ond, as an alternative if less stringent) (see,  for  example,  Connecticut's
 and New York's regulations, Appendix  E,  Table E-3).

 Problem;   General process weight rates  result in grossly inefficient
           control of shakeout and sand handling  emissions.

           Solutions  Mentioned

           1.  Adopt  a  separate regulation for shakeout sand handling (see,
 for example,  Pennsylvania's regulations, Appendix E, Table  E-3).

           2.  Rely on  the visible emission  regulation  if its application
would result  in more stringent  control.

           3.  Rely on  the process weight  regulation to  obtain initial in-
 stallation of control equipment;  then rely on other regulations that pertain
primarily  to  operation and maintenance to  ensure effective, continuous  con-
 trol.   Such regulations may include visible emission regulations and collec-
 tion efficiency regulations as well as operation and maintenance regulations,
permit  regulations,   malfunction  regulations, and even nuisance and odor regu-
 lations .

           4.  Rely on nuisance,  odor, ambient air quality,  or  some other
authority  to  obtain  initial control, then  on operation and maintenance  regula-
tions to ensure continued-control.

                                  84
-------
Problem:
 Problem;   Process  weight rates are  difficult  to apply because of problems
           in estimating input and ensuring representative testing conditions.

           Solutions  Mentioned

           1.   Adopt  a  concentration limitation.

           2.   Use  the  process weight rate  to  obtain initial control that
 should  achieve the emission limitation  with an ample cushion;  then rely pri-
 marily  on permit regulations, operation  and maintenance  requirements,  and
 similar regulations  to ensure that the  control equipment  is  operating properly.

           Process  weight rates are difficult to  apply to  the cupola because
           of  problems  in accounting for  fugitive  emissions  when  measuring
           emissions  during  stack  testing.

           Solutions  Mentioned

           1.   Adopt  either  of the solutions  described in  the problem immedi-
 ately above.

           2.   Measure  fugitive emissions  from  appropriate points  (e.g., roof
 vents)  and/or  during appropriate  portions of the operating cycle  (e.g., dur-
 ing  charging,  melting, and  tapping). These emissions are then included as
 a part  of  the  allowable  emissions from  the melting operation.  At  least one
 state has  derived  a  similar  factor  for pouring  and  cooling  emissions  and
 has  included  these emissions as a part  of the  allowable melting emissions.

 6.3.2   Regulating Fugitive Emissions

     Major problems  voiced and solutions  identified  during the survey include:

 Problem:   Existing regulations do not allow  for adequate control of
           fugitive emissions.  Fugitive emission regulations are too vague,
           too  subjective, or  too  complex.

           Solutions  Mentioned

           1.  Adopt  a  fugitive emission regulation that prohibits all fugi-
 tive emissions unless  reasonable  control  measures  are adopted;  then define
 reasonable  control measures  in terms specifically responsive  to  typical
 foundry problems.  The state should have  the power to insist on any of  the
 enumerated control measures at any point within the foundry that contributes
 to fugitive  emissions  whenever fugitive emissions  are observed or  measured
 at a point of exit  from  the  foundry enclosure without regard  for  whether
 they are observable  or measureable at  the propejjty  line.  The source has
 the burden to  demonstrate that the measures are,  in fact,  unreasonable.

          2.  Require  as  a  condition of an operating permit that measures
be adopted to  prevent  fugitive emissions.  These measures should be speci-
fied in the permit as conditions  for issuance and would be specifically en-
forceable without regard to whether it could be demonstrated that there was
an actual violation of the underlying fugitive emission regulations.

                                  85
-------
Problem;  Mass emission regulations do not allow for the control of
          .fugitive emissions.

          Solutions Mentioned

          1.  One state estimates  the contribution to fugitive  emissions
from certain  activities associated with melting (e.g., charging, tapping,
pouring, and cooling) and accounts for these emissions during stack testing
by adding them to the measured stack emissions to determine whether the melt-
ing operation is in compliance with a process weight rate.

          2.  An emissions  concentration  limitation may be applied at any
exit point depending on the definition' of "source" in the state regulations
or air  pollution statute.   If defined broadly  so that  roof vents, windows,
and other openings are incorporated, the emissions concentration regulation
could be applied to require further control of in-plant processes; however,
the effectiveness of  such  a regulation would depend  on its relative  strin-
gency compared to the actual emissions experienced  at individual points.
It would also depend greatly on the susceptibility of the exit point to emis-
sion measurement techniques.

Problem;  Visible emissions  regulations do not allow for  effective control
          of fugitive  emissions because they are not easily applied  around
          buildings or from nonrectahgular stacks.   They also do not effec-
          tively address the problem of emissions from multiple roof vent
          emissions.

          Solutions Mentioned

          1.  There was disagreement relating to the feasibility of conduct-
ing accurate  readings  exiting from the sides of buildings.  A spokesperson
explained that such readings are feasible if certain precautionary measures
are taken and adequate allowance for background opacity is made.

          2.  It may be necessary to revise the test method for determining
compliance with visible emissions  regulations  to use these regulations as
part of a strategy to respond to fugitive emission problems.
      i
6.3.3  Post-Installation Enforcement

     Major problems voiced and solutions identified during the survey include:

Problem;  Regulations  do  not  provide effective  authority for  ensuring
          continued compliance after  the  initial compliance demonstration.
          The primary  problem is  that to demonstrate a violation requires
          time-consuming, and  expensive stack testing which is usually non-
          representative of actual operating conditions.

          Solutions Mentioned

          1.  Adopt operation and maintenance regulations which are independ-
ently enforceable.  The most- effective of these regulations allows for the
state to  require preventive steps  without documenting an actual emissions
violation.

                                  86
-------
           2.  Adopt  and  enforce malfunction  regulations which require  self-
 surveillance  and reporting whenever control  equipment  (including capture
 equipment)  is down.  The most  effective  of these  regulations  requires  imme-
 diate  corrective action, reporting to the  state  within a reasonable time
 frame,  and subsequent preventive  action  according to  a plan approved by  the
 state  and  enforced as an independently enforceable requirement (in  the form
 of a permit condition, variance,  or enforcement order).   Malfunction reports
 should be  used  to assist in the development  of investigation  and  enforcement
 priorities, and they should be constructively evaluated  when determining
 whether the source has made good  faith attempts to comply.

           3.  Adopt  and  use operating permit requirements.  The most effec-
 tive of these requirements  allows for the state to impose reasonable opera-
 tion and maintenance conditions to prevent violations of  all  applicable  re-
 gulations  (including fugitive emission  regulations)  according to a plan
 prepared by. the source and subject to the state's approval.   Permit condi^
 tions  should  be independently enforceable without a need to  demonstrate
 actual emissions violations.   The permit should  be renewable  on a periodic
 basis,  and the  state should have  the authority to impose  new  conditions  prior
 to renewal.

 Problem:   States may not have  effective  investigation strategies.

           Solutions  Mentioned

           1.  Proper detection of actual and potential fugitive  emission
 violations  require  in-plant inspections.  Drive-by inspections are  an in-
 sufficient indicator of fugitive  problems because of the  difficulty of ob-
 serving significant  fugitive  emissions  at a distance.  It is  also  thought
 that any significant potential for fugitive emissions may be  detected  dur-
 ing an in-plant inspection  and effectively prevented.

           2.  States tend  to  rely primarily  on complaints and secondarily
 on inspection prioritization  as  a basis for dealing  with the  lack of  sur-
 veillance  resources.  Effective additional strategies are:  (a) to  "capture"
 other  investigation  resources by  coordinating with the Occupational  Safeguard
 and Health Administration  (OSHA),  or the state equivalent, and local health
 inspectors; and (b)  to educate the public and solicit public  assistance  in
 surveillance.

           3.  States that  must rely on  stack tests to  initiate effective
.enforcement should  explore the potential for an  abbreviated  version that
 would  be considered  equivalent within the meaning of  state regulations that
 allow  for  equivalent alternatives to specified test methods.   An  abbreviated
 version may be  sufficient  to  shift the burden to  the source  to demonstrate
 compliance  or may convince the source that a stack test would show noncom-
 pliance and therefore  serve as an incentive to  comply voluntarily with the
 state  requests  in issue.
                                  87
-------
          4.  States may  wish to use other techniques  that  show probable
non.complian.ce, including:  (a) establishing a rough correlation between visi-
ble emissions and particulate mass  emissions  during stack  testing;  and  (b)
inspecting control equipment  (including capture equipment) for signs of physi-
cal deterioration.   Such methods would serve as indicators of noncompliance
and would therefore be used primarily as leverage for pre-enforcement negotia-
tion.

6.4  UNRESOLVED ISSUES REGARDING FOUNDRY REGULATIONS

     As indicated at the  beginning  of Section 6.0, one  of  the major objec-
tives of the control agency survey was to identify enforcement problems and
solutions experienced by agencies in applying regulations to ferrous found-
ries.  Because of the limited effort and relatively broad scope of the study
and the wide  variation in responses from agencies, it was not possible to
conclusively address all  issues raised during the study.  The major ques-
tion which is still unanswered is whether the current methods of evaluating
ferrous foundry  compliance adequately address foundry emissions problems.
Three possible problem areas which impact on this questions are:  (a) fugi-
tive emissions;  (b) malfunction of control equipment; and  (c) the contribu-
tion of ferrous foundries  to NAAQS nonattainment.  The following paragraphs
review the  findings of the study with regard  to  the major  question  and  the
surrounding areas of concern.

     Data in Appendix  E indicate that the vast majority of state contacts
feel that ferrous foundries are generally in compliance with applicable emis-
sion limitations and are  rarely found in violation.  Hence,  these agencies
see no need for the development of enforcement strategies to deal with fer-
rous foundries.  The bases for these views are  threefold.   First,  it has
been the experience of most agencies that foundries have installed the con-
trol equipment necessary  to come into compliance upon request; or, if the
cost of control was not feasible, the foundry closed.  Thus, most foundries
have completed the installation phase of control, and that phase appears to
be perceived by  state  agencies as most  important.  Second, most states  use
process weight rates to regulate foundry emissions.  For some foundry pro-
cesses such as shakeout and sand handling, the input weight is so large that
the allowable emissions can be attained by equipment operating inefficiently.
Thus, these  sources are not perceived to be compliance  problems.  Finally,
many of the foundries which have inadequate controls are small jobber found-
ries located in economically  deprived rural areas.  These foundries are seen
as essential to the economy of the community and not as major emissions prob-
lems.  As a result they are not given, enforcement priority.

     The findings described above appear to be in conflict with the opinions
voiced during the  study by federal  regulators who indicated that ferrous
foundries are a  problem source category.   Some of  the  information obtained
from the  state agencies substantiate this  view.   It  is  particularly worthy
of note that almost a  third of the  respondents indicated that the majority
or almost all foundries had inadequate control equipment.  In addition, the
majority of  the  respondents indicated that malfunction of foundry control
equipment is  a problem.  These responses suggest that the current methods
of evaluating compliance are  not adequate.
                                  88
-------
     Given the inconsistent responses described in the above paragraphs, it
is not possible to draw conclusions about the importance of ferrous foundry
compliance problems.  Additional information'from three areas (fugitive emis-
sions, malfunctions,  and  the  impact of foundries  on attainment of NAAQS)
may help resolve the issue.  However, unanswered questions remain about each
of these areas.   The following paragraphs highlight available  information
and unresolved questions regarding each of these issues.

     As described in Section 4.0, many ferrous foundry processes are sources
of fugitive  emissions.  The responses of  the majority of the agencies  con-
tacted suggest that these fugitive emissions sources are not a major concern.
But this conclusion is not supported by the limited fugitive emissions data
in Section 4.0 and Appendix B, which suggest that fugitive emissions sources
may account for the majority of .emissions from ferrous foundries.  Further,
the data gathered during the  survey  indicate that  the  regulatory authority
for addressing fugitive emissions problems  (particularly continual  compli-
ance) is problematical and sometime totally inadequate.  Even if the regula-
tory authority for  control of fugitive emissions exists, Section 5.0 indi-
cates that .control  technology for these  sources is  not always  adequate.
Given the above concerns,  it appears that a more detailed analysis of fugi-
tive emissions quantities, controls, and regulations would provide a better
indication of  the impact  of  fugitive emissions on foundry compliance.

     One problem that was  identified by almost all persons contacted, both
control agency and  foundry personnel, is  the malfunction of control equip-
ment.  However, the attitude most frequently conveyed by those agencies con-
tacted was that  the malfunction of control equipment  is inevitable.  The
resultant position of the agencies is that enforcement actions are not war-
ranted.

     The assumption that malfunctions are inevitable is not consistent with
the  findings presented in Section 5.3.   These findings indicate that the
incidence of malfunction can be minimized with proper operation and mainte-
nance of control  equipment.   However,  even though practices appear to be
available to minimize malfunctions, the study does not clearly indicate that
states have the regulatory structure to prevent malfunctions.

     Because of the costs of testing, process weight and concentration regu-
lations are  not  a  good tool  for the prevention of malfunctions.  Opacity
regulations may be  of some value if the  malfunction results in excessive
concentrations at a control device outlet,.  However, if the malfunction causes
decreased capture efficiency  at a .fugitive emissions  source, the opacity
regulation may not be violated.  Three types of regulation were identified
during the study which may be useful in reducing malfunctions  and promoting
continued compliance.  These  regulations  include  malfunction regulations,
operation and maintenance  regulations,  and operating permit regulations.
However, because these regulations were not originally a part of the survey,
data are not sufficient to determine the  degree to which these  regulations
are available to  states  and their effectiveness in reducing malfunctions.
Thus, the degree  of reduction of malfunctions that  can be attained is an
unresolved issue.
                                  89
-------
     The  final  factor which affects the question of whether ferrous  found-
ries are  a problem source  is  the  impact of uncontrolled or noncompliance
ferrous foundry emissions on the nonattainment of NAAQS.  Since almost all
states indicated that SIP revisions would have no  impact on  foundry controls,
it is assumed that states consider  foundries to have minimal impact on non-
attainment.  The  scope of this study did not permit examination of the im-
pact of  foundry emissions on  ambient air quality.  However, since Section
3.0 does  indicate that the majority  of foundries from a sampling of six
states are located in particulate  nonattainment areas,  the  issue deserves
further study.

     In summary,  the  following unanswered questions were raised during the
survey:

     1.  Do foundry emissions  resulting  from lack  of control (either
         uncontrolled sources,  insufficiently  controlled sources, or mal-
         functioning  control equipment)  contribute significantly to the
         nonattainment of NAAQS?

     2.   Are  current technology and regulations  sufficient to control
         foundry fugitive emissions?

     3.   Are  current regulations  sufficient to reduce the  incidence  of
         malfunction to an acceptable level?

Answers to the  above  four questions would allow a much clearer determina-
tion of whether important compliance problems exist in ferrous foundries.
                                  90
-------
                                    TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing}
  REPORT NO.
                              2.
                                                             3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
  Summary of Factors affecting Compliance by Ferrous
  Foundries - Volume I, Text
                          5. REPORT DATE
                            phniaw 1 081
                          6. PERFORMING ORGANIZATION CODE
  AUTHOR(S)
  D.  Wallace, P. Quarles  the  Research Group, P. Kielty
  the Research Group and  A. Trenholm
                                                             8. PERFORMING ORGANIZATION REPORT NO.
  PERFORMING ORGANIZATION NAME AND AOORESS

  Midwest Research Institute
  425 Volker Boulevard
  Kansas City, Missouri   64110
                                                             10. PROGRAM ELEMENT NO.
                          11. CONTRACT/GRANT NO.

                            68-01-4139, W.A.15
12. SPONSORING AGENCY NAME AND AOORESS

  EPA,  Office of Enforcement
  Division of Stationary  Source Enforcement
  Washington, D.C.  20460
                          13. TYPE OF REPORT AND PERIOD COVERED
                            Task Final 2/79 -  12/80
                          14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
                    DSSE Project officer is Robert L. King  EN 341,  (202) 755-2582
16. ABSTRACT
  The report presents an overview of the ferrous foundry  characteristics and state and
  local regulatory practices which affect the evaluation  of  foundry compliance with air .
  pollution regulations.  Ferrous foundries are described with respect to size, location,
  investment trends, and process equipment.  Particulate  emission factors are developed
  for cupolas and electric  arc furnaces as well as the process fugitive emissions sources.
  Techniques are described  for controlling emissions from cupolas, electric arc furnaces,
  pouring and cooling,  shakeout, sand handling and the cleaning room.   Emphasis is placed
  on identification of  malfunction problems associated with these control measures, and
  operation and maintenance practices that can be used to reduce the incidence of mal-
  functions..  The regulations  which are applied to ferrous foundries by state and local
  agencies are identified.  Problems which have been encountered in regulating foundries
  and solutions which some  agencies have found for these  problems are described.
17.
                                 KEY WORDS AND DOCUMENT ANALYSIS
                   DESCRIPTORS
                                               b. IDENTIFIERS/OPEN ENDED TERMS
                                        c.  COSATI Field/Group
  Air Pollution
  Iron and Steel Industry
  Foundries
  Furnace Cupolas
  Electric Arc Furnaces
  Air Pollution Control Equipment
 Emissions
Regulations
Pollution;Control
Stationary Sources
Particulate
Operation and Maintenance
 13. DISTRIBUTION STATEMENT
  Unlimited
             19. SECURITY CLASS (ThisReport)
               Unclassified
                                                                           21. NO. OF PAGES
                                                                                98
             20. SECURITY CLASS (Thispagt)
               Unclassified
                                                                           22. PRICE
EPA Form 2220-1 (Rev. 4-77)    PREVIOUS EDITION is OBSOLETE
-------
-------