:ELLIN01S POLLUTION CONTROL BOARD

August 22~1984

IN

THE

~TTER

OF:

VOLATILE

ORGANIC

MATERIAL

)

R82—14

EM1SSIO~S FROM

STATIONARY

)

Dockets

A

&

B

SOURCES~ RACT

III

PROPOSED

RULE.

PROPOSED

OPINiON

OPINION

OF

THE

BOARD

(by

J~

Theodore

Meyer):

The

petition

which

initiated

this

rulemaking

proposed

to

regulate

the emissions from one type of stationary

source

and

from

five :Lndustriai categories of stationary sources,

The

source

specific

regulation

was

for

storage

tanks

containing

volatile

organic liquidsg

and the

five

industry

specific

regula-

tions

pertained to

(1) wood furniture coating,

(2)

heatset

web

offset

printing,

(3)

synthetic organic chemical manufacturing,

(4) asphalt

roofing manufacturing, and

(5)

petroleum dry cleaning.

The

proposal

also

included

language to amend the

existing

general

ruler

then

Rule

205(f)

of the Board’s Chapter

2:

Air

Pollution

and

now

contained

in

Subpart

K

of

35

Il1~Adm,

Code

215,

which

limits emissions of photochemically

reactive

material

or

odor

causing materials from all stationary air pollution

sources,

This

rule

was adopted

:Ln

the Board~sfirst major

rulemaking

on

air

poI1ut~on,

R 71~23(4

PCB

191,

April

:L3,

:1972).

The principal

distinction

between

the

existing

rule and the

proposed

amendment,

entitled

the

Generic

Rule,

is

that

the

latter~

~

intended

to

reguinte’

volatile

organic

emissionc,

not

just

the

photochemically

reactive or odor causing emissions,

from

those

stationary

sources

not

included

in industry specific and source

specific

regulations,

Nineteen public hearings were held concerning the

entire

regulatory

proposal

and

the

three Economic impact

Studies

(EcIS)

(Doe,,

Nos.,

83/15;

83/31;

83/32)

prepared pursuant to

Section

27

of the Environmental Protection Act (IlL~Rev.

Stat,

1983,

cli.

111½,

par~

1027)

After

four

of

those hearings,

it became

evident

that

the existing definition of volatile organic materials

would

be

expanded under

the

proposed

Generic

Rule.

Thus,

the

sources

affected

by

such

a

rule,

the

feasibility

of

controlling

the

additional

materials,

and the amount of

emissions

involved

were

it

i~ the

wish

of

the

Board

that

the

knowledge

and concern con-

tributed to

this

proceeding

by

Irvin

G.

Goodman

before

his

death,

and his legacy of inspiration he recognized and

remembered,

The

Hoard also acknowledges the work of Marili MrFawn,

hearing

officer

and ~

nictratve

assistant

‘i~ntn~smatter,

---

not

yet

identified

or quantified by

the

proponent,

the

Illinois

Environmental

Protection

Agency

(Agency),

To

provide additional

time

to compile sucn

documentation

without

delaying

the

remainder

of

the

proceeding,

on

its

own

motion

the

Board

established

two

dockets.

The

industry

and

source

specific

rules

were

assigned to

Docket

A~t~ieGeneric

Rule

to

Docket

B

(49

PCI3

247; October 27,

1982)~

~

comprehension

of

the

Generic

Rule

expanded,

Docket A

did

not pr.ceed

independently

as

expected.

it

became necessary

to

compara the Generic Rule

to

certain

industry

specific

rules.

For

example,

the

proposed

heatset

web

offset

rules

regulated

materials

not

included

in

the

existing

definition

of VOM, and not

proposed

for regulation

pursuant

to

the

Generic

Rule,

The

Printing

Industry

of

Illinois

requested

additional

time

to submit

a study

on

the

reactivity

of

the

inks

used

in

that

segment

of

the

industry.

Not

only

was

additional

time

granted,

but

that

action necessitated

preparation

of

and

hearings

on

a

separate

EelS.

Additionally, as

awareness

of

the

Generic

Rule

grew,

segments

of the affected

industries proposed industry specific rules as alternatives

to

the

Generic Rule

which required separate

documentation

and Board

consideration~ Although tine consuming,

both

outgrowths

made

for

superior rulemaking

in the end.

Twice

before

the

Board has adopted industry specific regula~

tions to

control

the emissions

of

volatile

organic

materials

(VOMs)

from existing stationary sources

in Illinois.

Those

regulations, adopted subsequent to the rulemakings known as

RACT

I

(R78—4~5 and

RACT II

(R8O—5)

are now

contained

in

35

Iii. Mm.

Code

215:

Organic Material

Emission Standards and Limitations,

This

ruiemaking,

commonly known as PACT

III,

is

obviously

the

third

in a series.

The

acronym

for

all

three

rulemakings is

derived

from the statutory mandate found

in Part

D of the Clean

Air

Act

(42

U~S,C.A.

7401 et

~

Section 172 therein requires

that

freasonably

available control technology” he implemented

at

existing

stationary

sources

in

the

nonattainment areas of those

states

needing

an extension from

the

1982

deadline until

1987 to

achieve

the

air

quality standard for ozone,

Illinois

is

such a

state,

having requested the extension

in

its 1979

and

1982

State

Implementation

Plans

(SIP), and as of the date of this Opinion,

having

nine

counties

designated as nonattainment areas

for

ozone,

The definition of “reasonably

available control techno1ogy~’

is contained

in 40 CFR 51,

along

with

the requirements for a

federally

acceptable

SIP.

However,

the specific parameters of

what

constitutes

reasonably available controls, and,

therefore,

the

parameters

which the states must adopt to insure that RACT is

implemented,

are

not,

instead, the United States Environmental

Protection

Agency

(USEPA) publishes a series of documents

entitled

‘~Control

Technique

Guidelines”

(CTGs).

Each of the CTGs,

which

are

summaries of

industry specific case studies, contains the

means

and

the degree of control

which the

USEPA requires the

state to

adopt

categorically as a part of its SIPS

in

order

to

have

an acce table SIP.

Failure to adopt rules

identical to

59~514

---

3~

those pres~atedin the CTGs

or other ones demonstrated by the

inda~rthaa state as comparable,

can mean that the state will

have

~ aideq

~te

SIP,

which

in

turn

can

trigger

the

sanction

provi~

sioric uf

the Clean Air Act found at Sections 110,

113,

and

176

(42 ~

~.

7413,

7413,

7506).

While the mandate for sanctions

is cc t~. ~d in the Clean Air Act,

the mandate to

adopt

the CTGs

or ot eni~sedemonstrate a state rule to be comparable

is not.

it

a

~

even contained in the federal regulations, but

instead

is

i’~

thed in the “General Preamble for Proposed Rulemaking

and tepr

a~of State Implementation Plan Revisions for Nonattain~

ment

Arsa~

(44 FR 20372).

Thi~federal policy statement includes yet another require~

ment which is relevant to this rulemaking.

The USEPA allows the

StaC~,

UI

.~J

cue January after one year fcom the finalization of

a CTG .~oadopt either the “rules” contained therein, or comparable

rules,

f

sources

covered

by

that

particular

CTG

are

within a

stat&

nnattainment areas,

Also of interest is the unstated

policy

of Lhe USEPA to publish draft CTGs.

Presumably adoption

of

rules comparable to the information contained within these

is

not

mandatory.

Of final concern,

and possibly of most importance

to this rulemaking,

is the requirement that RACT be

implemented

at exic~.

ic

stationary

sources

in

nonattainment

areas

as

expedth

tiousiy

~.

practicable

in order to demonstrate reasonable further

prograso

~3

U.S.C.A.

7501,

7502,.

These provisions in Part D of

the dc

n Air Act mean that the states are required to adopt

means

~

trol VOM from all stationary sources, regardless of

whetthr

~‘I~’

have

been

published.

The

USEPA,

again

in

the

General

Prear

e

Ii

a

limited

the

application

to

sources

with

the

potential

to cm~t ~c~ethan 100 tons per year,

i.e.

iiaJor stationary sources.

Therc

~

those not covered by a final CTG, the only principle

gove

r

he state’s rulemaking is that the degree of control be

prera~

‘

technology which is :easonably available.

e outset of this rulemaking, only cne the six ca:egories

under

c n~ideration,the petroleum dry clearers, was the subject

of

a

fjflaJ

LTG.

Despite the fact that a final CTG has been

i~su~en the petroleum dry cleaning industry, the Board is not

ad p~ o riles pertaining to it.

The remaining categories proposed

cy ‘~eAcency were the sub~ectof draft CTGs,

Since that tise,

the

~

:s finalized the CTG drafted for synthetic organic

~‘

and polymer manufacturing equipment.

Rules pertaining

co

th~

‘ategory are adopted.

The reasons for these two decisions

ai~ w~i

a~those on the categories with only draft CTGs,

the

Gener

c

Rule, and the additional two industry specific rules

resu:~

r’r

from

the

Generic

Rules

proposal,

are

discussed

separa

ly

n~. 14,

1984

the

Board

adopted

as

final

a

series

of

r1le~ :

the vegetable oil processing industry, which was one of

the

err

t

~c Rule spin—offs, and an exemption for the viscose

casi.g r”n~facturersfrom Subpart K.

Those rules,

along with

---

—4—

rules

pertaining

to

petroleum

dry

cleaners,

had

been

proposed

for

First

Notice on February 9,

1984 and published in

the

Illinois

~~ter

on February

24,

1984

(8 Ill.

~

2407).

The

Second

Notice

version was adopted on May

3,

1984,

but

the rules proposed

for

the petroleum dry cleaners were deleted, because after

closely

reviewing

the record in pertinent part,

the Board determined

those rules to be unnecessary.

The Joint Committee on

AdministratlV’

Rules

issued a Certificate of No Objection to that portion of

this

rulemaking on June 12,

1984.

On June

29,

1984

the Board

amended its June 14th Final Order

in order to have the rules

accepted

by the Administrative Code Unit of the Secretary of

State’s

Office,

The

rationale

supporting

the

final

action

taken

on

the vegetable oil processing rules and the

viscose

exemption

is

contained in this Opinion, along with that supporting the

remainder

of this rulemaking.

It should be

noted

that

the

viscose

exemption

adopted as final

is

repealed.

That

exemption

was

adopted

anticipating that the

general

rule

contained

in

Subpart

K

would

be

amended

with

some

form of the

Generic

Rule,

and would,

therefore,

be

more

restrictive.

Since

it

is

not

to

be so

amended,

the

previously

adopted

exemption

is unnecessary.

Again, this

action

is

discussed

in further detail under the

section

of

the

Opinion

entitled

Viscose

Casing

Process.

GENERIC

RULE

The

Generic

Rule was proposed by

the

Agency

to

be

applicable

to air

pollution

emission sources

and facilities with

the potential

to emit more than one

hundred

tons of volatile organic materials

into the atmosphere annually and not already controlled pursuant

to another Subpart of Part 215.

The proposed rule differed from

the

general rule to control organic material emissions already

found in Subpart K, in that it eliminated the concept of exempted

(non-photochemically reactive) materials as defined in Part 211.

The proposed Generic Rule would require controls at levels deter-

mined by the concentration of hydrocarbons in the

exhaust

stream

and the vapor pressure

of

the

compounds.

These

determinations

would

then be

used to assess

the percentage of control required.

Since

some of the

vapor

pressures

listed

were

lower

than

those

already

contained

in the definition of volatile organic materials,

application of the proposed rule could have been

more

restrictive

than

any of the

industry specific

rules

already

contained

in

Part

215,

or

those

adopted

in

this

rulemaking.

In order to identify the facilities possibly affected by the

rule,

a survey was

circulated

by the Agency.

Approximately 36

facilities were identified which collectively contribute approxi-

mately

30,000

tons

of

uncontrolled

emissions

annually.

Approxi-

mately

25,000

tons

of

the

total

are

attributable

to

sources,

primarily

vegetable

oil

processing

facilities,

in

the

attainment

areas,

In

fact,

vegetable

oil

processing

facilities

contribute

63 percent of

that

total,

and

only

three

of

the

thirteen

facilities

identifed are

in

nonattainment

or

contiguous

counties

(P.C.

47,

---

Table

11

5).

The other types of affected facilities identified

by

tn~

Agency are coke by-product plants

and.

chemical

manufacturing

facilithes

Since industry

specific

regulations

are

adopted

for

the

~r’q~

able oil

processing

industry

and

coke—by

product

recovery

I~iar~o

~nth only

ieaves

the

chemical

manufacturing

facilities

and

~

~elyother

parts

of steel—making plants

subject

to the

proposed

Generic

Rule,

As noted in the part

of

this

Opinion

portai~J.

i~

th

coke by—product

recovery

plants

regulations

under

the ~‘~~~aal Emission Standards fo~Hazardous Air Pollutants

proqrai

~‘eobeen adopted to control benzene and other emssions

frorr.

~.

l~makingplants.

Need to reaulate them as major sources

~

i~~y

mooted because benzene emissions constitute the

maJoL

pot

~aon

of

their

hydrocarbon

emissions.

As

for

the

chemical

manufa’tsring

industry,

it

was

determined

that

further

control

at

thes~

~c~i

mica

was

not

reasonably

availab

e

in

the

RAC~ II

proceoa:~n

No

additional

evidence

was

presented

in

this

rulemaking

to

deny

~~trate

that

the

proposed

Generic

Rule

differed

from

that

proposed

i~fl

RACT

Ii,

and

that

it,

therefore

represents

measonably

availabie

control

technology

for

these

facilities.

Finally,

uncontmo~ed

emissions

from

three

of

the

nine

chemical

facilities

identir~eJ

in

nonattainment

and

contiguous

counties

are

estimated

to be ~ess than 100 tons per year

If the rule were adopted

La

control

th~i

remaining

six

major

sources,

presumably 96 percen~of

3,465

tu~sof uncontrolled emissions would have been controlled

However,

the part

of

this

total

already

controlled

as orgar~c

materfrl~

under

the

existing

general

rule was

not

quantified.

~.

hoard

finds

that

the

Generic

Rule,

as proposed

m~nt

nece~o~to

control

major

sources

in Illinois, since the m~

of

tLtr

cc

subject

to

industry

specific

regulations in Part 215.

For chc~eonly subject to the existing general rule, the Board

~irYi

~

amount of emissions to be controlled does not ju~tfy

adop~

of the Generic Rule without some quantification of the

estimatod uncontrolled emissions already controlled as organi~

maceta~.

under that rule.

Furthermore

since most of these

socices are in the chemical manufacturing industry,

an industr~’

spec~t~crule premised on reasonably available technology ~ould

be prefrrable.

As mentioned above,

this showing was not made

tc

reverc~ethe decision made in RACT II that technology is not nw

r~ctsoram1yavailable,

1FDUS’~~ALCATEGORIES ADOPTED

Subpar’~ F’

Coating Operations

Sect~ou~

204(1):

Wood Furniture Coating

ye the other subparagraphs in Section 215.204, subpara-

Grapf

l~regulates a surface coating operation.

In this

irstanec.

the pr~’ribed VOM limitations are applicable

to operatiors for

coat:~o t:’tnishings made of wood,

wood

composition or simulated

59~~517

---

6—

wood materials.

Five facilities

in Illinois were identified to

be operating coating lines possibly affected by this rule.

A

typical

furniture coating line includes application equipment,

flash—off areas, spray booths, ovens and conveyors.

Examples

of

products coated are cabinets, tables, chairs, beds,

sofas, shutters,

doors, paneling and floor coverings.

The rule establishes the

maximum VOM content for seven types of

coatings

and the minimum

transfer

efficiency

to

be

achieved

in applying each type of

coating.

Traditionally, reducing the VOM emissions from coating

operations

has been premised on substituting existing coatings

with coatings containing higher solids content or higher water

content.

In this instance, emissions are not expected to be

reduced by reformulating existing coatings.

Instead,

the seven

VOM limits included in the

rule

are based on the percentage of

solids

in the materials

reportedly

used by the five Illinois

facilities or elsewhere.

For example, the percent of solids by

volume

in scalers reported to be currently in use

is between 14

and 16

percent.

The emission

limit

required

at

Section

215,204(1)

for scalers

is 0.67 kilograms of VOM per liter.

When converted

to a minimum solids content, this

is equivalent to 15 percent

solids per volume

(R,

275,

Ex.

14,

Table

II),

The rule does not

anticipate that the five identified facilities will have to

switch from coatings they currently use.

In the event that they

do,

the reported use of coatings with VON contents at or below

those which the rule is based upon, evidences that compliance

coatings are available,

As for waterborne coatings, compliance

with the rules is not dependent on their availability.

In drafting the CTG for this industrial category, the USEPA

anticipated

the availability of waterborne coatings

Ex,

29(b).

That CTG, however, was never finalized because the USEPA did not

document the availability of such coatings

(R,

280),

The Agency

on

the other hand,

testified that it believed such coatings to be

available

(R,

276),

If they are, then these five facilities will

be able to substitute waterborne coatings for those currently in

use, and thereby achieve compliance directly, or indirectly

through the internal offset provision found at Section 215.207 or

under the rules for Alternative Control Strategies contained in

Part 202.

In addition to the VOM limitations, the rule adopted by the

Board requires that two transfer efficiencies be achieved at

surface wood coating operations.

The rule requires a 65 percent

transfer efficiency in applying all coatings, except for tIe

repair coat.

A transfer efficiency of 30 percent is required for

the repair coat because it generally involves a small amount of

coatings and is a spot application process.

In

adopting

the 65

percent transfer efficiency,

a reduction in VOM emissions from

surface wood coating

is anticipated,

not from the coatings them-

selves, but from the application process.

Improving the transfer

59~518

---

—7—

eff~cienci~sgenerally reduces the amount of coating used, which

in

turn ieduces emissions of VON.

Conventional spraying operations have at least a 40 percent

transfe.

~lticiency.

Electrostatic spraying processes typically

have a 55 percent transfer efficiency.

More sophisticated electro~

static equipment,

such

as

disc and bell equipment can achieve a

transfer efficiency between 85 and 95 percent (R.

277).

Evidence

indicates that the industry has used this type of equipment since

1978

(Ex.~ 29(b),

pp.

2—9,

2—10).

Given

the effectiveness of

these syrtems, VOM emissions should be reduced by

35

percent

from

operations equipped with electrostatic systems, and by 53 to

58

percent

from those equipped with the more sophisticated systems.

The

affected

facilities can install simple electrostatic spray

systems

LO

achieve the 65 percent or a combination of sy.~emsto

demonstrate

compliance under Part 202:

Afterburners or reformu-

lated coatings, mentioned above,

are also available to achieve

compli’~nceeither alone or in combination with improved coating

application systems.

It is noted,

however,

that these two tradi-

tional metnods are not necessarily available

to

wood surface

coaters,

Based on information contained

in permit applications,

uncontrolled emissions from the five Illinois facilities were

estimated to be 2,900 tons annually

(R.

274,

Ex,

14:

Table I).

Three of the five facilities are located in nonattainment counties

and account for 845 tons per year.

Assuming a 44 percent reduction

in emisstons,

as was done by the Agency, emissions would be

reduced annually by 1,265 tons statewide, and by 372 tons

in the

nonattamn.ient areas,

Based on annualized capital costs and

annual o1~eratingand maintenance costs,

the EelS estimated the

total

annual costs to range between $512,200 and $1,304,200.

Using

thU above emission fiyures, cost effectiveness would then

range

between

$1,291 and

$5,581,

However,

this

cost

is

considered

to

be

o~rst~ited

because

the

annual

operating

and

maintenance

costs

used

in

calculating

it

included

the

existing

operating

afld

maintenance

costs

(R.

1899).

It is

noted

that

both

the

Eels

and

the

Agency

allowed that any

increased

operation

and

maintenance

costs

are

urobably

offset

by

the

paint

savings

(Ex,

48,

p.

4—’~,

R.

282)

That

would

mean

that

the

cost

of

compliance

would

e

the

annualized

capital

cost

divided

by

the

tons

reduced

annuaily.

If

used

in

developing

a

cost

effectiveness

figure,

the

range

is

then

reduced

to

$400

and

$3,600

(Ex,

48,

p.

4—11).

On

the

average,

the

Ec13

found

the cost effectiveness

to

be

$725

per

ton

of

reduced

emissions

(R.

1889).

In

adopting

this rule for the surface coating of wood r.~rnish-

ings,

a definition of the process is provided as well as

definitions

of

tIe

eafings

used

by

the

industry.

These

are

found

in

Section

215,~J~.

it

is

also

necessary

to

amend

Section

215.211

and

213.212

‘hich

address

compliance

dates

and

compliance

plans,

resp~’c

ively.

The sources located

in nonattainment counties or

59~519

---

—8—

those

contiguous thereto are given until December 31,

1985 to

comply; the remainder until 1987.

In amending Section 215.211

from

the language adopted in RACT

II, no substantive change is

made.

The rule now recites all counties currently designated as

nonattainment

by

the USEPA and the counties contiguous to those

counties, instead of a partial list of the nonattainment counties.

This framework is intended to facilitate more ease in identifying

those sources with a 1985 compliance date,

The rule pertaining

to the compliance plan

is amended to delete non-essential language,

and to specify that compliance plans are not necessary if redesig—

nation of

a county from attainment

to nonattainment occurs after

December

31,

1986.

Subpart

12:

Vegetable Oil Processing

Sections

215.340

—

215.347:

Hexane Processing of Soybeans and

Corn

The rules

in this Subpart set out the emission limitations,

the

methods for determining compliance, the compliance plan

requirements and the compliance dates for two types of vegetable

oil processing.

There are thirteen facilities

in Illinois engaged

in this industry, only three of which are located in nonattainrnent

counties.

Twelve process soybean oil,

three of these also process

corn

oil, and the thirteenth processes only corn oil,

Under the

original proposal these facilities would have been subject to the

Generic Rule.

As already mentioned, these facilities accounted

for nearly two thirds of the total uncontrolled emisssions originally

estimated for sources covered by the Generic Rule,

At the September

12,

1983 hearing, the Agency proposed an industry specific rule to

regulate the vegetable oil processes.

The industry, most specifically

the

soybean

industry,

found

the

Generic

Rule

and

the

Agency-proposed

industry

specific

rule

unacceptable.

At the December

8 and 16,

1983

hearings,

a

joint

proposal

applicable

to conventional soybean

processing was entered into the record

(R.

2869;

Ex,

79).

Subsequently,

similar regulations applicable to specialty soybean processing

(which four of the twelve soybean processors

are

equipped to do)

were proposed

(P.C.

40),

Rules for corn oil processing were

also proposed by the

affected

facilities

(R.

2930,

2937;

P.C.

37),

Extracting oil from soybeans is a complex, multi-step process.

Prior to extraction,

raw soybeans are cooked and pressed into

flakes,

The flakes are then saturated with a solvent, usually

hexane,

and the resulting oil and solvent mixture is separated

from the soybean meal,

The solvent laden flakes are then desol—

ventized and toasted to drive off residual solvent for recovery

and reuse.

The oil and solvent mixture

is simlarly stripped and

condensed to recover the solvent for reuse,

In addition to extracting the oil, these facilities process

---

the mosu.it~.ngsoybean meal.

So that the protein contained in the

meal can be utilized by animals,

including humans,

the proteins

must

broken down by heat and/or exposure to moisture,

This

is

known a’ denaturization,

and the extent to which

it

is achieved

is

~n~c’~ lunction of temperature, moisture,

and the length of

time

~cposed to elevated temperature and moisture.

Increasing

any a~‘n~sevariables increases the denaturization process.

~ho~-e re two types of soybean meal p~ocessing’

c iventional

and ape

alcy processing.

Conventional process~ngutilizes

a

desolv~ntizer—toaster (DT) to denature soybeans in the aracess of

extract... ig nexane from crushed and oil ~extracted neal

om

flakes.

The UT, therefore, both desolventizes the neal and denat~esthe

protein c)rtained

in

the meal,

It does so by exposing

cc

meal

to hit

t~rnparaturesin the preserce of at::,

for a

~nrn,~

‘y

period.

The meal

or flour produced

using conventional

processing

is suataole for animal, excluding humans,

conaumpti n.

iho

second process, specialty processing,

produce”~ teal

suitable for human consumption by beginning with un—deratumed

soymeal or flour,

This means that the oil must be extracted and

the mes’ Iting meal desolventized and cooled without exposure to

excess~veheat or moisture for lengthy periods of t me.

~r

that

reason,

a DT cannot be used,

Instead vapor and flash des...~ventizing

are used

:o desolventize soymeal

or flakes with a minirum of

proteir

denaturization,

Neither

system

is

as

effective

.

driving

off the lexane

as conventional desolventizing because the neal

is

not subjected to high heat or elevated moisture levels

fom

long

period~,the conditions necessary for the most efficient lesolventizing

(R.

~°2

P,C,40;

Ex,

81).

It

is for

this r~”sonthe

‘ass

balance J’nitation, discussed later in detai’

for spec~4 ltv

pr(ce’

.~,nc ms double that adopted for c

vettional

coybea;

~rocessing.

2nere are three major sources of emission at these ~ro~essing

planL:

the main vent, the dryer vent, ard the cooler

‘e

t

The

main

ent

IS

usually

controlled

by

condensers

or

nineral

ci1

scrubbera which are capable of achieving

90 percent or greater

contro

efficiences,

If used together,

a 99.9 percent control

efficiency can be achieved, which is even more than required

under U e Generic Rule,

The other two sources are not cont ‘dIed

by a

J~or.equipment.

At conventional processing facilit”cs

control

s instead achieved through efficient operation of the

DT,

nec’

dryers or coolers are used with flash or vapor

desolv~ising systems at the specialty processing facilitie.,,

soire lexane will be lost to the atmosphere.

Based on inaistry’s

inforn’at~on

no

dryers are used at Illinois specialty facilities.

Meal

..oter~may or may not be used.

There are addition~l~‘seurces~

presert a~all these facilities.

There are fugitive hexane

emisr~:cr~ard solvent losses through retention in finished soybean

meal

c~!~

Ci.’

(Ex, 81,

P.C.40).

TI

‘ndustry investigated using ada—on controls at the dryer

---

—10—

and cooler vents.

Incinerating or using the hexane vapors as

fuel proved too dangerous, inefficient, and even if the vapors

could be used the energy rendered would be three times the

amount

needed by the facilities.

Carbon adsorption was rejected to

technical

and safety problems.

Finally,

due to costs, oversized

mineral oil scrubbers proved impractical

(R,

2906

—

2912,

2944).

Due to the escalating cost

of

hexane

over

the

past

eight

years,

the

industry

has

been

installing

control

and

recovery

equipment~and more efficient DT~sand flash or vapor desol—

ventizing

systems

(R.

2599,

2911).

Nevertheless,

since

it

is

difficult to control most of the vents with add-on equipment and

to quantify emissions from the several types of sources at these

facilities,

the industry would prefer a mass balance limitation

for both the conventional and specialty processes.

A mass balance

limitation is premised on two parameters easily measured

—

total soybeans crushed and total hexane

lost.

Framing the rule

this way eliminates any requirements to test the vents at the

dryers, coolers, mineral oil scrubbers, condensers,

and the

residual hexane in the meal or oil.

Instead, the total

hexarie

and

inventory loss at a plant is used to measure emissions.

Based on seven years data, mass balance limitations were

developed

(Ex.

81,

P.C.

40,

50,

51).

On the average it was

determined that the conventional soybean facility processing

could

lose no more than 0.0026 pounds of heaxane per pound of

conventional soybean crush,

and no more than 0.0052 per pound of

specialty soybean crush,

Industry testified that by over control-

ling at the mineral oil

scrubbers, rather than at the dryer or

cooler vents it could operate within these mass balance limitations.

For those facilities unable to currently meet the 0.0026

limitation,

it may be necessary for them to upgrade or install

a

modern

DT

or

new

mineral

oil

scrubber,

The

latter

would

cost

approximately $100,000 plus annual operating costs of $43,000.

A

modern DT would require capital expenditures of $1.4 million, but

would provide lower operating costs

(R.

2912

-

2913;

~x.

81).

Furthermore, since the regulations are based on mass balance

limitations it may be necessary for plants to reduce start-up,

shut—down and other non—operational

losses.

Since

these

improve-

ments are proven and allow for the recovery and reuse of hexane

solvent,

the industry had no objection to the associated costs,

The corn processing industry operates in much the same way

as the soybean industry.

It agreed that a mass limitation was

preferential to the Generic Rule and adequately documented that

the affected facilities could limit emissions to no more than 2.2

gallons of hexane per ton of raw corn germ processed

(P.C.

37).

Recordkeeping

and compliance

determination

with

the

adopted

mass balance limitations

is to be done on a

180 day rolling

average.

Each day the facilities must recalculate the decreases

59-522

---

—11—

in snlme~t.Lr.ventor’y and the total conventional and specialty

soybean

cruch or raw corn

germ

produced

over

the

previous

180

days.

rI

the sum of soybean or raw corn

germ

processed

multiplied

by

L,h’~:e~pectivemass balance limitation is greater than the

tota.~ cc..,. esae of the solvent inventory for the same preceding

180

da’

4~hefacilities

will

be

deemed

out

of

compliance

It

is

noted

J..~t the

solvent

loss

data

for

the

specialty

facilities

was

gatherec~ aur~ng periods

of

no

overlow,

and

when

the

attacled

ocuventa

na’

facilities

were

inoperative.

Therefore,

the

average

loss

caur

be

measured

for

the

specialty

facilities

apart

from

the ror,ien :onal facilities,

The rule provides for the

.wo to

be

combi.

ed ~nr purposes of demonstrating compliance.

Mindful of

the

zc’~s.~ason,quarterly averaging was censidered ire e”d of

biannual averaging.

However, the Agency and the indus~ryadequately

demon~tr

~cd

that

emissions

during

the ozone period woul~be

no

greater

Itan those experienced during the remainder of the

lear.

In fact

‘~mmasmonswould probably be less during this critical

period

d

e

to

the

seasonal

nature

of

the

industry.

Therefore,

quarterly

averaging

was

considered

unnecessa

ily

cumbersome.

Subpart

U’

Printing and Publishing

Sections 215,402

—

215.408:

Heatset Web Offset Lithograp:~c

Printing

Hea1n~etweb offset printing is a lithographic prmntinc

procec~ ~hmch means it involves printing from a flat surf;oe.

Maintair

ny the distinction between image and non—image areas

is

done

cit

a.ly.

The non—image areas are receptive to wa.ex

or

the

f’:

~a

..

~olution,

The image areas are wa er repell~e-d’

or

ont receptive,

so that the ink stay..~on the

ii

mc~

areas.

tach printing unit of a press has a series of vsr’ically

arranged rollers and cylinders above and below the web,

~e

the

paper

1ne fountain solution and the inks are transfer’

‘d ~y

comp’cx./ rrranged rollers to the ulate cylinder

The inago ~s

then transferred from the image plate to a rubber covered blanket

cy.ir dci

and then to the web.

The infeed section of the press

allowe the rolls of paper to be mounted, aligned, unwound,

and

fed :hrigh the press.

In a typical process—color heatset web

offs~t

I

enographic

printing press,

each printing unit simu~ta—

r.eou3i~ya,,~pliesa single color to both sides of the web.

~ogether

an

pr rt~ngunits can overlay colors for

a full color

ir’tmge

withcut irving between printing units,

After the la”t pu

itir.g

unit

tha printing web enters the dryer.

The most commot tjpe of

dryer ~s a high velocity, hot air blower.

Air temperatures c n

be a-

~

as 5000 F.

Much of the heated air is recircu

at .d.

with only enough being discharged to prevent the buildup

f

exploa~

~

~lvent

vapors.

The web leaves the dryer with surface

temperat.re

between 266°F and 329°F and travels over an assembly

of driv..; sneel drums with chilled water circulating through them

wnfcit col

..~heweb to a maximum

860

F,

This

cooling,

in

combina—

59-523

---

tion with the evaporation of the ink in the dryers, prevents

the

ink

from transferring to adjacent sheets when the printed web

is

cut, folded, and stacked

(R.

667—668

2713;

Ex.

29(e)).

The two major sources for organic emissions are the printing

units and the dryers.

If highly volatile organic compounds,

such

as isopropyl alcohol are used in the fountain solutions, signifi-

cant quantities evaporate at the printing units.

The draft CTG

in this category assumed that most of the evaporation take place

in the fountain pan

(Ex,

29(e)),

Since much of the total isopro—

panol used in the system reaches the dryer, the Agency reasoned

that evaporation occurs from the thin film on the rollers or that

transferred to the web, and then finally in the dryer.

Evaporation

of the ink solvents is considered to occur primarily

in the dryer

(R.

669—670).

Accordingly, the regulations proposed by the Agency

required capturing and controlling organic material emissions

with afterburners by at least 90 percent, or reducing the volatile

fraction of the fountain solutions to no more than

5 percent and

utilizing a condensation recovery system wIth at least a 75

percent removal efficiency for organic materials.

A provision

allowing any alternative method equivalent to either of these

was

also

proposed

(Ex,

1).

Fountain solutions usually contain isopropyl alcohol

as a

dampening solution, an etchant such as phosphoric acid,

and gum

arabic.

Isopropyl alcohol,

a volatile organic material,

is

usually maintained at a

20 percent concentration in fountain

solutions where automatic controls are used, and 15 to 25 percent

where manual make—ups are used,

Alcohol

substitutes are available,

such as ethylene glycol, and generally have vapor pressures less

than 0.0019 psia at 70°F and are typically used in concentrations

of about

2 percent, by weight,

or less in the fountain solutions,

The minimum practical l~ve1of alcohol

in dampening solutions,

however,

is

5 percent because older,

less flexible rollers require

more pressure to print properly and the alcohol

substitutes may

have too high a viscosity to properly pass through the roller

systems

under

the higher pressures,

If the isopropyl alcohol

concentration is reduced to 5 percent,

a control efficiency

between 67 to 80 percent should be achievable

(R.

666—671).

Industry agreed that alcohol substitutes with vapor pressures

less than 0.0019 psia at 70°F are available, and offered that,

in fact, the industry is voluntarily switching to them.

One

company testified that by substituting materials at four plants

its emissions were reduced from 81.8 tons of VON annually to 11.5

tons

(R.

2198,

Ex. 59),

Ink solvents are primarily mixtures of narrow cut petroleum

fractions, having an average molecular weight of about 206.

C1

C

hydrocarbons have been identified as ink solvents, and ~

c~monlyused one has C

and C1

hydrocarbons.

Most ink formula~

lions contain 30 to 50 ~rcent,

~y weight, hydrocarbons.

Magiesol

47

is frequently a major component of the ink solvents and has

---

the

au~ve men~,iored characteristics.

It

and

any

other

cariponents

usually

rave

vapor

pressures

less

than

0.0019

psia

at

700

F

and,

therefor~

under

the

existing

Board

definition,

are

not

volacile

organic

~t

~rmals,

Nevertheless,

the

proposed

languaqe

required

all

orqnni

u’terials

to

be

controlled

by

either

an

afterburner

or

a

cond~~u

tition

recovery system.

Catalytmc

or

direct

flame

afterhum

.a

a

can

serve

to

control

the

emissions

associated

with

the

ink

a

vents

and

the

fountain

solutions

by

90

percert.

If

an

afterbu~r

is

used,

reduction

of

the

volatile

organic

cont

nt

of

the

fours

dii

solution

was

not

to

be

required

Preaumabl’~ t

Ls

was

beca’s~

orly

25

percent

of

the

VON

in

the

fountain

soiutan

evapora

~

at

he

roller

and

web

areas

so

that

it

is

suI~icient

that

t

~

fdnalning

75

percent

is

controlled

by

the

d

ytr~s

nifter—

burner

R.

611)

A condensation

recovery

system,

on

.he

r.tneL

hand,

cc

not

control

isopropyl

alcohol,

but

can

only

serve

U

control

the

organic

ink

solvents,

Therefore,

the

require’reit

to

install

and

operate

the

same

was

coupled

with

the

require:ent

to

reduce

ne

molatile

organic concentrations

of

“he

fount

un

alutions,

As

stated

above,

alcohol

substitutes

should

provide

67

to

“)

percent

~eductions,

so

only

a

75

percent

con

r,l

efficieriry

‘nas

sought

frau

the

condensation

recovery

systen

(Ex.

1;

~.

072,.

Ind~..y

did not disagree with

the

Agency

about

Ue

cit.

J.ty

to

contron

ti’e

volatile content of

the

fountain

solution

a

titer

it

objcctcct

‘~othe

requirement

that

organic

components

o’

‘

.,nk

be

control_ed.

Industry

argued

that

the

solvents

conaaar’ed

thereir

~

exempt

from control under

Subpart

K,

given

the

cirrent

defiro,tiont

of

photochemically

reactive

matenia

s,

and

for

reason

t

.r

fndustry

had

switched

to

these

so~.venta’

ro

othe

industrit

are

required

to

control

comparao’e

material.~

vrd~.

Part

zli,

‘~rd

f

the

proposed

Generic

Rule

~a

ppl’caoie

a

these

co~upcurds would

not

have

to

be

controlled

~urtthermore,

this

segnent

of

thc

printing induetry

ir

competitive

with

the

rotogravar~’

and

flexographic segments

which

are

not

requrrad

to

contro

‘~

y~nic material

emissions

other

than

under

Subpart

~

arid

are,

..n

fact,

encouraged to use the

same

ink

solvents

tias

segment

would

be

required to

control

(R,

720—724;

P.C.

4)

Finalny

.ndustry

argued that

ink

oils

used

are

not

any

mo:e

photo

I’o~t:al1y

reactive than ethane and,

therefore,

th~r’.

need

to

.trol

them

as

ozone

precursors.

ir,

t~k.ttion

to disagreeing

with

the

p

opriety

of

r~’onl~t’ g

organc

erials,

industry

disaqreed

with

the

Agency’s

euiz,.

‘

estimd~

..,

costs

of

retrofitting

existing

sources

and

the

cost.

‘t

.

selling condensed ink

oils

as

fuel

R

725—”~34

“~14 705).

Fira2~

‘rdu~trybelieved it would

be

difficult

to

reduce

‘o

VON

coa :~

~ration

of

fountain

solution

to

as

low

as

5

pe

..e

I

becau~e .~y

of

the substitutes

for

isopropyl

alcohol

have

vuor

pressures

treater

than

0.0019 psia at

700

F,

and

are,

therc~ore,

volatiit

rganic

materials themselves,

To

resolve

tnir

dilemma,

industr,’

‘goested

that

the

percentage

be

increased

to

8

percent,

or

the

,~a ~t

of

the fountain solution

be

limited

to

5

percunt

---

—i 4—

isoproeyi

alcohol,

as

opposed

to

5

percent

volatile

organic

material

~R. 702—704)

In

support

of.~~tsargument

that

the

organic

materials

need

not

be

contro11ed~

industry

presented

two

parts

of

a

five

part

study

underway

at

Battel

Ic

concerning the reactivity of

the

ink

solvents

used

in heatset web offset printing

(EKe.

22,

39).

The

first

part

evaluated

the

volatility of heatset printing oils and

the

canabil

i,ty

of

conducting

tests

within

smog

chambers

to

determine

their

p.hc~techemical

reactivity.

The second

part

evaluated

the

ink

oils

reactivity

in

comparison

with

the hydrocarbon ethane.

A

third part,

not

finished

at

the

time

of

hearings,

is

to

collect

stack

samples

and

compare

them

to

the ink

oils

in

order

to

determine

if

the

printing

and

drying

alters

their

composition

in

such a

way

as

to

increase

or decrease reactivity

(R.

1615—1618; 1650~.

Two

ink

oils

were

studied,

Magiesol

47

and

470.

The latter

is

predominantly

paraffinic,

but

also

contains just

in

excess

oe

10

percent

o.efinic

and aromatic hydrocarbons.

Magieso:L

47

±s

its

counterpart

having

similar

paraffinic

content,

but

is

without

the

olefinic

and

aromatic

hydrocarbons.

Using

a

high

ratio

of

hydrocarbons

to

nitrogen

oxides,

a

photochemical

aerosol

appeared

within

approximately

two

hours

of

irradiation

during

the

experiments

with

Magiesol

470,

but

did not with those

conducted

with

Magiesol

47.

The

importance

of the first task~s

results

was

that

it

demonstrated

that it

is

technically

feasible

to

proceed

and

evaluate

the

same materia1s~:photochemicalreactivity

under

ratios

of

hydrocarbons to nitrogen oxides known to

lead

to

ozone

formation

(R.

755—758).

The

standard

urban

mix

would

have a rtuch

lower

ratio

than

that

used in the first

part

of

the

Battelle

study

CR.

782),

There

are

a

variety

of

parameters

that

can be

used

to

evaluate

photochemical reactivity.

The

Battelle

study identified

eight

and

chose

one,

maximum

ozone

concentration,

to

be

used

as

the

yardstick

for

the

second

task.

One

series

of

experiments

was

conducted

to

compare

the

reactivity

of

the

two

ink

oils

to

each

other

and

with

ethane,

Some

compared

reactivity

on

a

mass

basis,

that

is

parts

per

million

of

carbon,

while

some

employed

comparabi,?

molar

concentration,

that

is

parts

per

million by

volume.

In

both

cases,

the

oils

produced

a

higher

ozone concentration

than

ethane

within

the

first

twelve

hours

of

irradiation,

although

ethane eventually

generated

more

ozone

when

compared by

mass

it

must

he’uoted

that

the

ratio

of

hydrocarbons

to

nitrogen

oxide

was

5/1,

much

higher

than

normally

found

in

an

urban

mixture,

Another

series

of

experiments

used

a

typical atmospheric

Ii

~

r~’m

miiture

composed

of

seventeen

hydruc~ubcns

Re,,

that

part

of

the

purpose

of

the

second

part

is

to

compare

the

oi1s~

reactivity

to

ethane~s,

in

half

of

this

series

of

experimer

the

oi:

a

aere

substituted

in

place

of

the

ethane

used

in

the

other half,

~then

ethane

was

replaced

by

Magiesol

47

the

maximum

59~528

---

—1,5—

ozone

~ceraration

dropped

5

percent.

Phen

it

was

replaced

with

Magieso~~ 0,

it dropped about

13 percent.

So this serie~demon-

strated that replacing ethane with either of the ink oils results

in

a

r~d’ t:(,

in

the

maximum

concentration

of

ozone

formed

in

the

d1~

t~

“~~lvehours,

irou

ry

pointed

out

that

the

concentrations

of

ink

oils

used

were

;reater

than

those

found

in

normal

atmospheric

cnditions.

It

was

ci

argued

that

ethane

is

continuously

in

a

gaseou

state

and

is

“it

~fore

always

available

for

ozone

formation,

w’u~’r.

the

vast

r

jority

of

the

heatset

ink

oils

wher

emitted

condence

to

lore

a

~quid

particulate

droplet,

and,

tnerefore,

arc

r.ct

availab~t~~

ozone

formation,

Setting

these

two

poirt~

a

ide

industry

ta’ed

that

the

ink

oils

used

in

the

heatset

web

.

fact

prir~t_r~,U

adji~dged nonreactive

becau~~

he

~xper2men~

~r

~rstrated

that

the

pitotochemical

reactivity

of

the

oils

to

be

comp

~

‘e

to

ethane

R

1617—1644;

Ex.

39),

The

SLPA

commented

on

the

Battelle

study,

and

agreec’

“let.

the

urban

unvironment

simulated

tests

ft

shoe

the

oils

~o

be

slightly ~ss

reactive, although the

difference

may

net

be

‘ynif

i—

cant

(Er

,iS,

R.

2174—2175),

Not surprisingl~,the U”EIA. ~

some of

“l’s

“est parameters changed,

For ins.,ance,

it

‘~‘u oa’ ed

lower ratios of hydrocarbons to nitrogen oxides,

aria the

aca

“on

of an

EUhUenth

component to

the

backgrourd

urban

mix

~

Benefits

to

be

achieved

by

either of the recommendations

~

questi ra

1~

The ratios used by Battelle ranged betwc,i

and

5

0

1.

which include the urban ratio identified

in

a

~

cal

study

noisced by the Agency

(Ex.

41).

Secondly, sin~

seveUecn

0

“he hydrocarbons in the mixture were voar’1.c

Si

ic

material

‘ddina one more would not seem ~o be neca~ar’

The hgency offered

no

supporting evidence that the m~

oi~s

from the

ad’

oils used

in

heatset

web

offset print’.ng

a

~‘.

d

‘r’P

contro~

is

ozone precursors.

Presumably,

iike the U’,

~

y

seek to control these compounds because they are relea. ~o

the atm~rnplie’re as vapors.

However, the USEPA l’as hot fan~’

ii

its draft

O’IG in this category, and has not changed its sU

~d’td

definit’ in of volatile organic materials to include there

v-

~

or com~arabJeones

(Ex.

24),

Furthermore, muTh of the

a.t,~

rnecha’irsm~now in place at these facilities

is

to control

opacit’

‘qh,,,ch

is

indicative of rapid condensation of

tin

Sho rid the third part of the Battelle study demonstrate

thnt

ink

oi

s are changed in some fashion by tht printing

anc

di

proces”

o tiat they are more photochemically reactive tlar

etnana

may be necessary to reconsider the decision ~o

reguia~ero~atileorganic materials,

At thiF time, howev

r

evidence ,irdicates that the present approach ~s all thab

nece~s~

‘,:

air quality purposes.

Ac aU

idy mentioned, industry and the Agency had di”

x’

t

total

~i

,~arbonemission estimates.

Originally, the dia~u~c

59~527

---

—16—

centered around the hours of press operation.

It became clear

that

tine maximum hours of operation listed in the Agency~sTotal

Air System represented the number of hours the presses are manned,

rather than actually operated.

Through cooperative efforts,

the

maximum hours of operation were revised downward accordingly, and

average operating hours were also provided by industry.

It

is possible that these figures should be reduced even

further.

Pursuant to conditions contained in its permit, one

company has performed stack testing as

a part of a yet incomplete

study,

to quantify actual hydrocarbon emission amounts.

The test

results indicate that actual emissions are 53 percent less than

originally calculated using a mass balance formula.

That formula

assumes that only 20 percent of the hydrocarbons applied to the

web are retained.

The test data suggests that much more is

retained.

For example,

using

mass

balance

calculations

the

annual hydrocarbon emissions

from the facility where the tests

were conducted are estimated to be 347 tons, while based on the

test data annual emissions would only amount to 163 tons

(Ex.

75;

R.

2680—2682;

P.C.

49).

Of course the disagreements on total

estimated emissions and

the possible error in the mass balance formula would color the

control cost estimates,

Industry and the Agency also disagreed

on

the costs and practicality of afterburners and condensation

recovery systems.

(Exs.

21;

47;

59;

71;

74

R.

1714—1716;

2182—2188;

2650—2660;

2672—2674; 2689—2694).

A detailed analysis of tine

rightfulness of these disputes

is not necessary.

Since only

volatile organic materials,

i.e.,

primarily tine fountain solutions~

must be controlled,

the total emission estimates at issue are

greatly reduced.

In fact, one company estimated that use of

isopropyl alcohol and its substitutes account for only 4.6 percent

of

its

annual

emissions

(R.

2194).

For

the

same

reason,

tine

types of control and, therefore,

the costs are also greatly

reduced,

Affected facilities will only have to change

tine:tr

fountain solutions.

Afterburners or condensation recovery systems

will be optional, or only necessary if the ink formulas contribute

sufficient amounts of volatile organic emissions to necessitate

control.

The

Board

notes

that nearly half

of the presses at

facilities

in

nonattainment

and

contiguous counties are already

controlled in some

fashion,

Therefore,

compliance

expenditures

should

be

minimal.

Originally an

exemption

for

facilities

emitting less than

100 tons per year of organic materials was proposed.

That level

of exemption becomes inappropriate since the affected facilities

will only have to consider controlling volatile organic materials.

Furthermore, only two facilities

in Illinois use sufficient

amounts of isopropyl alcohol,

i.e.

30,000 gallons or more, to emit

more than 100 tons annually.

(Ex,

21),

Industry suggested a 40

ton per year exemption level in keeping with the levels of signifi-

cant impact contained in the new source review and the prevention

59-528

---

of

significant

deterioration

programs

(R.

727,

Ex.

21),

However,

an

exemption limit

of

25 tons per year is adopted which means

that

facilities using

more

than

7,500

gallons

of

volatile

organic

materials,

generally isopropyl alcohol,

will

be required to

control

emissions either

by

reformulating

the fountain solution

or

by add-on equipment.

According to an industry survey

o’niy

about

fifteen of 52

companies

contacted

will

be

affected,

and

56

of

their

68 presses

are

already

controlled

(Ex,

21,

Table A).

Three final comments

on

the rules adopted.

First, the ink

reformulation

and

afterburner

requirements

are

included

as alter-

natives to reducing the

VOM

content

of the fountain solution.

However,

if the ink

solvents

themselves

contain

volatile

organic

materials,

one of these alternatives would also be necessary

if

uncontrolled emissions are greater than 25 tons a year.

Secondly,

it is the VOM in the fountain solutions which is required to be

reduced rather than the isopropyl alcohol, and the reduction must

be to

5 percent instead of

to

8

percent,

because testimony indicated

that nonvolatile organic substitutes are available

(R.

703).

Finally, the condenser recovery system requirement is included,

but assumed to be applicable only to those facilities

using inks

containing volatile organic materials.

Installation of

tine same

will otherwise not be required.

Subpart

Q:

Synthetic Organic Chemical

and Polymer Manufacturing

Sections 215:420

—

428:

Leak Inspection and Repair Program

The synthetic organic chemical manufacturing industry produces

high volume intermediate and finished products from chemical

feedstocks derived from petroleum, natural

gas,

and other raw

materials.

Approximately 375 intermediate and finished products

have been identified for this segment of the chemical industry.

The polymer manufacturing segment of this industry includes

operations which convert monomer or chemical intermediate materials

from the synthetic organic chemical

segment into polymer products,

namely polyethylene, polypropylene, and polystyrene.

The regula-

tions contained in Subpart

Q

only apply to facilities manufacturing

these chemical products which are listed in full in Appendix

D of

Part 215, and only serve to control fugitive volatile organic

material emissions.

Given that,

these regulations establish a

leak inspection and repair program for these manufacturing facilities.

The Agency had proposed that resin and synthetic rubber manufacturers

falling under SIC codes 2821 and 2822 be regulated; they are not

under the rules adopted.

Sixty~~’fourfacilities manufacturing synthetic organic chemicals,

polymers,

resins

and plastics were identified in Illinois; sixteen

in

counties

designated

as

attainment

for

ozone,

and

the

remaining

forty—eight located in nonattainment counties or counties contiguous

thereto

(Sw.

8,

Table 12),

Since this accounting was done under

59-529

---

—18—

the

‘r&:rri’ti

a

that

all

four

types

of

manufactiring

facilities

would be regslated, some of the sixty—four may only manufacture

resin

ai.d synthetic rubber products.

If

so, they are not affected

by

the

rcrlaLions

adopted.

For

those

remaining,

the

regulations

only

app

~

if

the facility contains more than 1,500 ~p~nerits

in

gasec’ni’

or light liquid service, and monitoring

is only required

at

equiprr~

~

t

containing

more

than

10 percent volatile organic

material s

Ta,. ~rocess units in the synthetic organic ciemical and

polyme

uanufacturing segments of the industry are similar,

They

include miacerial handling of feedstocks and finished prodacts,

heat

transfer,

reaction,

sorption,

distillation,

evapora”ion,

crystaii7ation, and separation.

Likewise the equipment used

in

both

segments is similar, each utilizing pumps, compressor.~,

in—line

valves,

pressure

relief

valves,

open—ended

valves

(includ~

ing

proc se drains),

sampling connections, flanges, agitators and

cooling U~wers, Fugitive emissions generally result when. either

gaseous or liquid process fluids leak from the equipment b~cause

the sealing materials deteriorate or the sealing effect is ‘educed

due to

in’proper design, construction, installation, mair,~enanee

and operathon.

The most commonly used seals are compressc~d

pack,rngs

gaskets, finely machined surfaces

(mechanical seals)

valves seat~,,ball valves and plug valves,

Th’~defrnition of “component~contained

in Section

21”.

04

lists the tjpes of equipment,

primarily sealing mechanistic

air ch

are stb’a,~,tto the regulations.

Excluded fra~mthe defin,+,

r

are

any p’eoe’~of equipment in heavy liquid service since

fur’.

emission’ of volatile organic material would not be mat”rl’

Four ~ea~

r

mechanisms are also excluded’

va’ves not ‘~xernally

regelatea

~anges, ball valves and plug valves.

Either

at

is

not cost effective to control

fugitive emissions from these

sources or the emission amounts are not significant.

Twc

specific

piece. of eq~,apment, agitators and cooling towers, are aia~

exciu ,ed

~cause

emissions from these two sources cannot be

quantified*,

Since all of these sources are excluded from the

defin,tion, none are to be counted in determining whether or not

a fac’lity nas more than 1,500 components,

and none of these

sourc~s~trc~mubjectto the inspection program.

i

les proposed

by

the Agency for this industriat ‘u~er~ory

were or~really premised

in the draft CTG ~Control of Voiaf’,lc

Organ

c

hem,cal,

Polymer,

and Resin Manufacturing Equipeen’

publis~.ecny the USEPA

in

January,

1981

(Ex,

29(a)),

Thi~

t~

‘~ment

presumed

Jr

t the equipment,

process materials and emissions

~rns “agitators”

and

“cooling

towers”

were

inadv?’

.artl~

omitt d

:romi the definition of “component”

in Section 215 ~04.

At

Sec~~d.tiotice,

both

will

be

added.

59-530

---

involved are similar to those

found at petroleum refineries,

The

rules suggested in the draft CTG and,

therefore, those submitted

by tire Agency.

were patterned after the leak inspection program

for petroleum refineries.

In fact, the Agency proposed that the

rules be incorporated. iSto Subpart

R:

“Petroleum Refining and

Related Industries; Asphalt Materials”

(Ex.

1,

R.

186).

The

petroleum industry objected to this on the grounds that the

process materials used in each industry are different, and should

Subpart R be amended as proposed, the petroleum refineries would

be subject to additional requirements

(R.

596

—

599).

In April

of 1982 the USEPA published “Fugitive Emission Sources of

Organic

Compounds

—

Additional

Information on Emissions, Emission

Reductions

a~dCost”

(AID), which studies

in more detail the synthetic

organic chemical and polymer manufacturing facilities

(Ex,

8).

Most importantly, this study changed the emission factors

assigned

to the various sources used in the industry.

The draft CTG had

identified the types of sources to be regulated based on the

industry, but had premised emission factors on information

pertinent

to both the petroleum refineries and the synthetic organic and

polymer manufacturing.

The revised em~ssionfactors contained in

the

AID

document were developed after further investigation of

the latter

(R.

191),

The Agency amended its proposal and support-

ing documentation pursuant to the petroleum industry’s objections

and the new information provided in the AID document

(R.

1314,

1322;

Ex.

8).

In order to estimate emissions, control costs and environ-

mental impacts for process units, the USEPA studies developed

three model units.

Each model plant is defined according to the

number of components in volatile organic material use it contains,

b~causefugitive emissions were found to be proportional to the

rumther of potential fugitive emission sources,

i.e. the number of

components,

in the plant,

rather than related to plant capacity

Cr

throughput

(R.

190),

Model A was defined as having approximatel~r

1 020 components, excluding cooling towers and agitators; Model B,

~,C60 components;

and. Model C,

19,495 components.

These

examples of plant complexity were considered representative of

facilities nation—wide by the USEPA,

with

52

percent of existing

facilities to be similar to Model A,

33 percent similar to Model

B, and 15 percent similar to Model C (Ex.

29(a),

p.

2

—

18).

The

rr

incy he~ievedthese models representative of the industry in

llinoas, and were accepted for use in developing the ECIS

(R,

~,0”;

Em.

48,

p.

3—10).

The Agency proposed an exemption for facilities with less

than 100 valves in gaseous or light liquid service

(R. 1314).

The background documents indicated that this type of facility,

having

so few potential sources of fugitive emissions, would have

OrilY approximately

5

tons

of

fugitive

emissions

to

be

controlled.

The Agency testified that the cost of a leak inspection and

repair program

is not warranted for this small of a return,

estimating

it would cost more than 400 times that instituted

at a

larger plant

(R. 1320

—

1321).

While it is

understandable

that

---

20~

the ~xe~

~

i~

~)~T~5(~J

Jfl

the p~tntc ~omplexaty, it

is

perpiexinq

r~.

rni

based only on va’ves, and

cr01

the total

number of arnporerrs

:~.

a

facility.

~urthermore

this level of

exemncica~

~n:~

‘~er~sentsccntroling

5 tons annually,

appears

cc

be

r”~ji~~~

lo’.

(todd

P plcnns Lace ap :oxLmtei~ 388 components, excluding

agttaton,

c ~ing

cower..

fThaçer, ana valves and pumps

in heavy

l~a~&d

selva;L~

1ppro.ciirr~

,

2s.

co- ‘onents

)f The tatal are

esti~

c~

tD

I

~

.a

fir

~.

git

~tquid

service.

As

dipcsscr ao~

“,

~

~nr

r

~

1~t

:o

be incThded

in

acter

a n~

~i

oThi

~‘

~o

a

rat

li~y It

is

d

cf_0

I. t~

~

ziTh

ot

a.

2..

~

are or

the

raIl

nd o~~gcar’e ~

LIe

~e,

~v

.

Th

u5:rg

cc

I

itd ~

valves

is d:scna cc bn~v

dot’n~~c

~

Dates

nat Lb percent

of the valvr~

in

ft

lic

s

d plug valise

(R,

1215)

:~

ThS

pe:ceThac

Th

a

rn

~.

Uc ind~stry,it

is

likely that

todd

p

.~n

e

ia~ton)f haL

and plug

valves from th~dLira.

-

a

of

cc~p~ert

-

~atcfewer thin 103

valves and,

therefore,

‘odd be

axempt from regulation under the

Agsncy~ssagjested exempt o

level.

Model

B plants contain

approxirrately 1,525 conponanta,

~xThuding agitators, coLl~ng

towers,

flanges, and valves ard pacins

in heavy liquid service.

Again,

the number of ball

arid plu

valves

s uncertain~

If 95

percent of t1~eestimated 925 ~a.crr are hal±and plug

valvet, tha

number of valves at Mcdel B plants would also fall below the

suggested 100 valve exemption leveL,

If the same components

excluded for Mdes A and B o~artsare not counted,

Model C plants

contain ~pproxirrao.~y1,69

ccnpor.ents.

AgLn,

aosuming the 95

perceat ftrure,

Ir’ of ~e

ratfmared

2

6a0 ‘aThes contained in

th’s. fcaljfreo

aTh

.ct

cc..

nd p1 g nl a~ (AL fie.e~based

cdi

8,

Teble

1.

‘2hercf ~cc, I

the A~ercy a exemption

Level

s’as adopted, along with a hell and plug valve exclusion

only

Model C planls woura be suLjec~

regu1att a.

Tire

,cOC

4omroae

are 1tLn

level

adapted by the Board

takes ante consideratior.

thc er.cnL conplexity of tne plant,

a as ~r

Lban

just the an

‘e~

rY

vaI cc cents lb ating to

‘ts potent

i

f~gi~~c.

emj-~ions, Furtheran..e, ~t link0 .~stimatedannual

eLsa

re to

:acility compA~

t

eDtabIiThir.g a limit.

lugi

ssiona

‘rcm Mode

~ plardi ar~estThated Th be 42

rors

pc

year,

from Model B plant~,

~‘

tons per year;

and from Model

C

uI~n90 520

LOPS

per year

(ha

.

6,

lable 5)~ These emission

~t:o~ctesinduce potential fugitive emission amounts from ha

a

anc

p uc

calves,

which the dccrd tas chosen to eliminate

Eve

so, ~t is only at plants as eonplex as Model

B, which contain

~f0 cmponents before bat

arc’ plug valves are excluded, tba~

p~’tenkialemisciors are greater than 100 tots per year

Assum’ng

mat ball and dug valves d

not contritvte greatly to annual

ftgat~mc

cm

ssacn’

it

iS

r

ieredsary to include them when asse~sLa

bhe ~1cnt

a complexity,

and once their contrihation is deleted,

It

Ic

procable that amiseL

frorc Model

13 plants are less than

-)

532

---

—21—

100 tour per

‘rear.

If the 95 percent figure is not representative

of the amount of bal

and plug valves used. in the industry, then

the number of comlonents

(i.e. non—exempt valves) to be counted

will rise acc~’oinglyand tinore facilities with the potential to

enit more than 100 tons per year will he subject to regulation.

Although the Board is unable to unravel the 100 valve exemption

level

and the potential to control only

5 tons per year, the

exeention level adopted

is

based. on the plant’s complexity and

the

potential

of

its

emissions

to

harm

the

environment.

Premising the exemption level on plant complexity coupled

with the potentIal to emit more than 100 tons per year,

still

iea/~’-~one problem unresolved.

Some facilities are operated year

round

while

sore use the equipment subject to regulation inte~-

iritteutly foL bat~uprocesses

(R,

1382).

It would be ~‘erycostly

to undertake a monitoring program, even once a year,

if that

equipment is ur~donly for a few hours.

It was suggested th~tan

exemption for facilities limited to emitting less than 100 tons

per year either by operation

or

by

permit

be

adopted.

Having

adopted an exemption level different than that proposed,

it

Li

difficult for the Board to determine whether an annual emiscion

exemption is still advisable.

Furthermore,

it is not clear how

many batch process facilities are affected and the amount of

annual fugitive emissions each contributes,

Comments are invited.

Ball and plug valves

have

been

excluded

from

the

definition

of

“component”

and,

therefore,

from

determining

the

complexity

of

a plant because they have extremely low leak rates,

Unlike

Fine

globe and gate valves commonly used at refineries,

they do n~

have packing seals.

There is no packing gland mechanism for

leakage.

They are used by the synthetic organic chemical and

poiyler manufacturers for just that reason and to comply with

occupational Safety and Health Administration

(OSHA) regulatio~s

(R,

1806).

Should the process materials be allowed to leak~the

cnt~ide

seat

of

the

valve

would

deteriorate

and

freeze

the

vai.’~e

the emission factors for ball valves is 200 times smaller than

that for gate or glove valves,

(P.

1215—1217).

For the same

reasons, ball and plug valves are also exempt from monitoring

orogram even if a facility is subject to an inspection and —ep~r

prgrdm because it contains more than 1,500 components.

Pursuant to Section 215.422(h),

storage tank valves, pump

equipped with mechanical seals and pressure relief devices coni~

*At First Notice the term “inaccessible valves” was also

theted at Subsection

(in),

Subsection (a), however, also appl~’~

o ~‘inaoceoeiblevalves~and requires that they be tested annuai:,.

To eliminate this contradiction, the term shall be deleted flow

0ubsectio~ (in) unless public comments indicate that subsection

~

shouth be deleted instead.

---

72~-

to

an

r

~a~-inc frare

header

or

vapor

recovery

device

ar—

so

exempt

Loa

rorftcring

reqrrements.t

The

Agency

suggeeted

that

all tLe—e d~’

acer-

ni’ exempt, except for pumps equipped with

menhar.

r

on

Ls

“1’-c anicas

seals

are

used

in

place

of

p

cring

gsnuc

~e

ThnLrrl

emisThoad

rR.

1217).

Based on studie0 a~fi’e

typtea.

chatacal

ompourds used in the industry, the avenge

Ui$

ha

atc from trecharical seals was determ~nedto be

11

)ran~e1-our

as opposed to the 120 grams/hour estimated by tie

USEPA

nd the Agency

lB.

i.i33; Er

33).

It should be net’e

that

the

-

a

I’

~cer-

tc

~

-he

(JOEL

enirsion rate for all

a’

scat’-

di

~c ~rams/hour, The emission factors are still

-

r

an a

Sc

-ry suggested,

ano tae Board

qrces

that

‘f

.1\’San

ically

se

e..

ups

arc

exemot-di

fro

he

m~itor

ag

reqa

~ovS

,

the affected

laciliLies are encouraged to ewitch to th~e

yr-.. cf

seaflr~d”The, .I~thhwi’l more erfecLv~5yrnducn

fajit’.

emisern.

han

‘

.1 ants

in pection and repafr

(A.

J9.t

c~ I’

ha0 then date -rilnad

ti

t a facsi~tyL

bj

~

the regusa..iens,

two types of inspectiors mast be cordic

First,

all pumps

(except those witi mechanical

seals)

n at

visually inspccted week’y.

If found to be dripping, they

~-

t

cc

repaired w’tha.n 22 days,

unless such a repair trust await J

th am,

or thc a ailability of a repair part.

In these events, ta

leaki’ g oump must be repaired as soon as possin1e.

Secon~y

non—exempt

components must be tested with monitoring eaufpa

approved by the Agency at least once a year.

Equipment con,

J~

‘-c

inaccessible, that

is out of reach or unsafe to test witno t

special

irecautions, must be monitored only once a year.

remairing oonponents must be tested once a year ~mmed~ate

-,

preco

11v

or

at

the

outset

of

the

ozone

seasor

If

fornd

5

drJ~

ug or

~o m

thokinq

ifl amounts groater than 13~000o

per

m Then ~f volatile organic material when tested, th

-,e

components, usually valves, must be repaired within 22 days

or

as

soot

a ooosible.

If more than

2 percent

of

the

total nun

af

conç ~ertc tested ‘re found to be leaking, than a secon”

an

no

‘caL program must be conducted during the ozone sea n

This second in°pection

as

triggered because it was generaL.

fount that

if the percentage of leaks

ic

below

2 percent

on ;

one a’ercent

of the componencs leak during the course of

a

y’-a.

yo’cvcr

ti-’ second inspection only involves testing pressir

rtdic

aid

ppelane

valves in gaseous service and comprc°’m.

acetIc

and, of course, any of those components found leakin-

A

f,rst

time.

the tgency proposed a quarterly monitoring program

wa.

pen ds provided if one quarter’s tests indicated that on y

pe’~cen’

r

ess of total valves tested were leaking.

The lr~a

dercrib’d above, however,

is

adopted as sufficient to redu’

emiss’

ic

¶rom leaking equipment during the ozone season

F

period of nest concern for public health and welfare,

Earth..

a

a

1,

i,.

ccnp’ring the number of leaking components found in reLic

t

arid

~oae in

the

syntnetic

organic

chemical

and

polymer

manet

a.

59-534

---

—23—

facUitho~ the latter is found to have 50 fewer percent leaking

componeuts than refineries when the materials are subject to OSHA

workplace concentrations

(R.

1211—1212).

Since most of the

affected facilities are subject to OSHA, it would be incorgruous

to r~guiieguarterly monitoring at these industiial facilities,

and not at petroleum refineries.

The Agency did offer that

quarter~y~nspections are more cost effective than annual inspec—

~ions heoause leaks are detected earlier, which causes co ~rective

measures and product savings to occur earlier. Consequently, this

would entan~ecost effectiveness

(R.

1314),

Affected facilities,

are; -f course,

free to monitor more frequently than jusF ~ ring

the ozone sea~on,

Tine Agelcy proposed that all open—ended lanes be aou~~ied

with ~loublebLicI’ scaling mechanisms

(Ex,

1, Rule 205th,thJ,;

R.

195).

This would mean that the ends of sampling lines b~ ~ocble

sealed

with

a

cap,

blind

flange,

plug

or

other

sealirg

devr”es

According to industry’s testimony, many line erds are alrethy

equipped with ball or plug valves, the efficiency of which has

already

been discussed.

Industry further offered that installation

of

a

second

sealing

device

would

mean

that

a

second

union

craftsman

would be required to remove the second seal in order to takc

e

sample.

Industry added that plant safety procedures ofceo :~~otibit

installation of a

second sealing device,

Finally,

one cipany

estimated that replacing

existing valves on sampling lir~e”v

Fl~

double block valves could cost

over $75,000, or approximately

$225 per valve,

for what

it considered negligible emission recuctions

(R.

1215

—

1216),

The Board did not adopt

this requirement

Rather it anticipates that the affected facilities will be

~

c

rsged

to replace ordinary valves with the more efficient plug ox i~a’l

valves,

in turn reducing the potential for fugitive emis’ ion

much ~n tine same way as was argued for mechanical seals

H

qe

it does recognize that many affected facilities may contrnue to

use

i langes which are also not counted in determining th~ ~x

level

and are not subject to the monitoring re~uireinents~

The variety of chemicals

used

and

manufactured by this

Industry is diverse and difficult to quantify.

The control

mechanisms alruady in place at the affected facilities

is aln

l&~fic~

It to ascertain,

Therefore, the authors of the Ecic

f~j~

it is difficult to estimate expected annual reductions in

~i~&tx~

emissions and future control costs,

The information cont’~o~

r

the EelS was developed using

the Agency~sproposal.

The appicc~

ibthity of the leak

inspection and repair

program adopted b~t~

Bc& ~lis very different than that originally proposed.

Lb e~

turn, makes

it

even more difficult to quantify expected

reduct

c

and costs,

Assuming quarterly inspection

and

double blocked seals,

net a~nuarzed costs

and

cost

effectiveness were

estimated for

each of the model

plants

in

the

Eels,

It

was

estimated

that

Hode~A plants would have to incur net annualized cost of $L,130

59~535

---

at a cost per ton of $742; Model B plants, $22,410 at a cost per

~on of $412; and Model C plants,

$39,000 at a cost per ton of

$191

(Ex.

48,

p.

4—6).

Industry found these investment costs and

average

cost of $334 per ton

far too low

(R.

2137,

2165).

Much

of the disagreement centered on the emission factors used, the

emission re:ducti.ons assumed,

and that some of the facilities

affected were rever studied federally.

For instance, industry

argued

that

the

USEPA

did not study

plastic or resin facilities.

However, these are not affected in the regulation adopted by the

Board,

Likewise, the number of inspections required and the

number of facilities affected,

and number and types of components

to be tested have been substantially reduced

t.han originally

proposed.

Therefore,

the estimates for the model plants are

askew,

Nevertheless, some of the information

contained

in the

EcIS as well as other information presented at hearing can help

e~t.imatecosts for individual facilities

.

The capital cost of a

monitoring instrument was estimated to be $11,990, which, of

course,

can

he

annualized.

Labor,

r•epair

work

and

parts

would

require

additional

outlays by a facility undertaking

its own

inspections.

Another alternative would be

to hire outside con-

tractors.

An independant contractor testified that his company

performs inspections charging $1.50 per component on original

testing, and $1.00

for the follow up

(R.

1968).

This would mean

that a

plant containing

the 1,500 components

would have to pay

$2,250

for an

initial annual

inspection,

and less for retesting

or a

follow—up

inspection in the ozone

season.

Of course, repair

work

and parts would

be at an additional cost.

Since the investment

for

double block valves is

eliminated and

the cost of inspection,

repair

and recordkeeping is reduced to once or twice a year,

the

cost

and administrative burden to affected facilities should now

be

r.easonably

related

to

the

potential

to

emit

fugitive

emissions.

Subpart

U:

Coke Manufacture and By—Product Recovery Plants

Sections 215.500, 215.510

—

215.517

Coke By—Product Recovery

Plants

Three

coke

by—product

plants

were

identified

in

Illinois;

each located in a nonattainment area,

Originally the

VON

emis~

sions from these plants, and any other possibly unidentified coke

by—product recovery plants, were to be controlled under the

proposed Generic Rule, the proposed

storage

tank

regulations, and

tine regulations proposed for the synthetic organic chemical

manufacturers

(R.

355,

356,

406).

After visiting a

coke by—product

recovery plant, the Agency,

however, agreed with

the affected

industry

tinat a limited, industry specific rule was more appro—

priate~ Accordingly, at

the December

7,

1982 hearing the Agency

submitted proposed language for a leak inspection program

(R.

1324).

Those process components dedicated to light oil liquid

service would be subject to a visual

inspection and

repair program

to reduce fugitive emissions, Light

oil

liquid

was

defined as a

---

—25—

liquid

condensed or absorbed from coke oven gas and composed of a

mixture

of benzene, toluene

and xylene.

In

later testimony

the

Agency suggested

that in

addition to the inspection program,

and

instead

of the all

encompassing Generic Rule, only the

uncontrolled

emissions from four

types

of sources would have to be reduced

by

85

percent or more

(R,

3041—3045).

At hearing,

description of the processes involved

and the

means

of control was

minimal.

A paper,

~‘BenzeneEmissions

from

Coke

By—Product Recovery Plants

-

National Emission Standards

for

Hazardous Air Pollutants~’, authored

by

L,

L. Beck

was introduced

R,

2586,

Ex.

69(c),

It contained a description of the coke

by-product recovery plants and possible control methods for

benzene,

a VOMO

Industry disputed the emission amounts represented

therein,

but did not dispute

the ability to

control

the

same four

types of sources identified by

the Agency or

the feasibility of

the leak inspection program

CR0

3054—3055,

P.C.

42).

Further-

more, although

it

did not agree with

the emission figures, industry

found

the

rule acceptable

(R,

1402), and

believed that,

as revised,

the regulations would provide emission reductions

equal to or

greater than

that anticipated under the

Generic Rule

(R.

3055).

The Agency proposed different compliance dates

for the two

programs involved, allowing the affected industries

until

tine end

of 1985 to implement

the

inspection program,

and until the end of

1986

to

install any equipment necessary to reduce emissions

at

the four emission sources by 85 percent.

Industry requested

until

the

end

of

1986 to comply with both programs

for capital

planning purposes

(R,

3056).

Given the short

compliance deadlines

under the

federal National Emission Standards for Hazardous Air

Pollutants

(NESHAP)

programs for benzene,

as adopted and proposed

(discussed further below), and since these facilities are located

in

nonattainment areas, it seems unnecessary to delay

compliance.

Therefore, compliance for both programs will

be required to be

achieved no later than December 31,

1985.

Generally coke by—product recovery plants are a

part of a

steel—making facility,

Coke, which is derived from coal,

is

a

necessary material

for converting iron ore to

iron,

The coking

process

(converting coal to

coke) takes place in

coke oven bat-

teries, producing a gas with a high Btu value,

This gas is,

therefore, used to

underf

ire those very

same ovens and other

parts of the

facility,

However, before it is used,

the various

chemicals evolving from the

coal

are

separated

and recovered by

passing the

coke oven gas through the coke by—product

recovery

plant.

It is

the cleansed gas which is used

as

fuel

to make more

coke

(H,

3052—3053;

Ex, 69(c),

According to the Beck paper, four areas of the coke

by—product

recovery

plant,

involving approximately a dozen major

sources,

emit the VOM benzene,

The four areas are:

(1) napthalene

separa—

tion and processing,

(2)

tar separation and processing,

(3)

light

59~537

---

—26—

oil r’~ v~ry,and

(4) leaks from pumps, valves, exha~stex~ard

other equipment components.

Only three of the four area~are

sub~ec~tto this regulation;

the napthalene process is rot

r~wo

of

.;e ~eTOic~~g three areas are processes, each

invrlvinç, ~‘~oof

the ~cur cores! n sources subject to Subpart

U.

In the

irst

prccEs~aur ic’t to this regulation,

tar is removed

frorr the aas

by ~

d~cant~rsand is then held

in storage for later use or

sae~

The two largest emission sources in this process axe the

tar d canse~and the tar interceptirg sump

Each type

s sub~ect

to 85

rcert reducticn requirement under Section 215.~ 0

~r

the

s

r. p~o~

eus,

‘igint

c

I

is reco

~red I

the

o

g~r~Light oil

is composed prurar~llof benzene

Th~I

x

emr~s~unsource

in

this

process

is

tine

vent

attached

L

Jrcth

oil. ~~arator/condenser,

used to exhaust non—condensab

trtuents.

Also evolved in this process is a sump, which receives ~i?

proces

‘

wastewater,

Uncontrolled enissions from botl~

f~

oil

cordenser/separator and the light oil

surcp are req r~

tr

be

reduced

by

85

percent,

According

to

the

Beck

paper,

tthe

ouctrons

can

be achieved by sump covers, condensers, and closed v et

systems.

Technical

or safety problems were acknowledgea

C

possibly preclude closed vent systems at light oil sumps

The f~nalarea subject to regulation is the leakirg

rc~

process

equipment such as pumps, exhausters and pipeline ~

and otl~erequipment components,

As stated above, only

tY. cc

involved in light oil

liquid service are subject to an in re ‘~ion

and

repaix program.

Furthermore, those components serv~circ

ke

oven ass lanes, operating flare headers, and vapor rec

x

devic

arc exempted,*

Under Section 21~.5l2the non

er

r

corrpore ~

must be visually observed weekly.

If found

1

tr~’

tz~E

co oncr~tsaus~be repaired within twenty—tw

day’

U

is i:iio~srIeto repair it for

lack of

a

part or while th.

nents

unit is in service0

Then the repair can be delavec~

ir ~il

the p

t

is

rec~ivedor until process turn—around.

Rec

i~

Iaks

repairs, and delays must be maintained for two ys~~

These requthements are parallel to those for petroleum rc

A; the outset, actual emission data for these soureeci w

~parc~

Beginning with the total emission estimate for all

~‘

p~

uc

recovery plants in the United States,

as give~ it

to

Beck ‘D-~per and estimating that 40 percent of the total

1.3

able

t~i fugitive

emissions,

the Agency

arrived at an es~irt~

I 85~ton’ per year of fugitive volatile organic emission I on

I 1rnors facilities.

The 40 percent figure was premi.~e

-

S nes those components servicing coke oven gas lines

ex~.pt a~ light o~lliquid is the condensate from coke ov~i

‘~.

o require leak inspection of the components

in l~~n

liqoid

~o‘ice rather than light oil liquid service?

---

—27—

information documenting the synthetic organic chemical manufac-

turing category

(R.

417—418).

Industry disagreed with

the exten-

sive extrapolation exercised

by the Agency to arrive

at an

emission

figure for fugitives at coke by—product recovery plants.

After the industry

specific rule was developed, emission

figures specific to th~four

types of sources

to be controlled by

85 percent were provided.

Using the

benzene

emission

factors,

presumably

those used in the Beck paper (as provided by industry

in Ex,

85), along with a multiplier of six to arrive at total VON

emissions,

tine Agency developed total VON emission estimates for

the four specified emission sources at two of the three identified

facilities.

Total uncontrolled estimated emissions amounted to

2,575 tons per year.

If

85 per cent control requirement is

achieved, fugitive emissions from these sources

should be reduced

to 386

tons per year

(P.C.

47).

No cost data was provided by the

Agency or in the EelS.

(H.

3060;

P.

C.

47; Ex.

76, pp.

28—29)

However,

at hearing one facility estimated that compliance may

cost

it $1 million

(R,

3059).

Based on the emission amounts

attributable to that facility, cost effectiveness per

ton

is

estimated to

be

$643.

Emission data pertaining to

the inspection

program

was

not quantified

further,

Exact

figures are not critical,

however, since only visual inspection is required.

Product

saving and

plant safety should be sufficient incentive

to justify

this program.

In the

interest of cohesiveness, two interesting

aspects

about the regulating of

coke by—product recovery plants

are

noted,

Industry

testified that numerous sources

at coke by—product

plants

are subject to

other

existing rules,

but are exempted by

the terms of those rules

(H. 3053—54).

For instance, the storage

tanks used in the three processes

are

subject to Subpart B:

Organic Emissions from Storage and Loading Operation,

but are

exempted presumably

due

to

size

or material content,

The sepa-

rators used in the processes are subject to Subpart C:

Organic

Emissions from Miscellaneous Equipment.

Finally, other operations

may be subject to the general

rule for organic

materials found at

Subpart K.

It may be prudent to

group

the exemptions and applicable

regulations under this new

Subpart U,

Suggested language would

be

welcome,

Secondly, on June

6,

1984 the USEPA adopted final rules

under Section

112

of the Clean Air

Act

to

control benzene as

hazardous air pollutants,

except from process

units located at

coke by—product

recovery plants

(49 FR 23498).

That same day ii

proposed standards for

benzene emissions

from the coke by—product

recovery plant

(49 FR 23522).

The proposed

rule encompassed

emission

standards, equipment, work practices and operational

requirements,

The sources considered were more numerous

than the

four named in this rulemaking, and the inspection and repair

requirements more stringent

in

parts,

On

June

29,

1984 the Board

---

—28—

adopted

these NESHAP

regulations under its preemptory

ruiemak~ng

authority

(H 34—24)

iii.

Rev. Stat.

1983,

eh.

111½, par.

1009.

1(c)

~.

Should the proposed NESHAP standards for the

coke

by—product

recovery

plants

be

finalized

by

the

USEPA,

the Board

will

be required to

adopt

the

same

pursuant

to

Section

9.

1 of

the

Act

(id~

(It may then

he

necessary

to

have

a

rulemaking to

delete

the rules

adopted

herein.)

In the

meantime,

this new

Subpart

Ti

will

require

a

reduction

in

all

VOM,

including

benzene,

from those emission sources

at

coke

by—product

recovery plants,

albeit fewer

sources,

where

it

has

been

demonstrated that controls

are

reasonably available,

Comments

on

the necessity of

this

Subpart

should the NESHAPS program be

finalized

by the USEPP ste

INDUSTRIAL

CATEGORIES

NOT

Storage

Containers

pursuant

to

the

concurrently

adopted

definition of volutti a

organic

material

(VOM),

the

storage

container

regulations adopted

in

H

71—23:

Emission

Standards

(4

PCB

191,

at 239;

April ~L3~

1972)

and

now

found

in

Subpart

B

of

35

Iii,

Adm.

Code 215 are

applicable

if

the

material stored

is

an

organic

material

with

a

vapor

pressure

of

2.5

pounds

per

square

inch

absolute

(psia) or

greater

at

7O~

F.

If used to store such

material

those containers

with

storage

capacity

greater than 40,000 gallons

or

more

most. be

equipped

with

either

a

floating

roof

or a vapor recovery systom.

At the outset of this rulemaking

it

was

proposed

that

the existing

rule

he amended to

include

those

storage

containers

used to store

volatile organic materials

with

vapor pressures of 1.5 p~is

or

greater at

storage

temperature0

To

implement

the same,

a definition

of

volatile

organic

liquids,

as opposed

to

an

amendment

to the

existing

defanition

of volatile organic

material,

was

prcpossd.~

A volatile organic

liquid

was

to

be

any

material, other thans

petroleum liquid

(already

defjned)

with a vapor pressure of

;Lth

psia

or

greater

when

it

is

at

equilibrium

with

its

own

vapor at

storage

temperature

(Ex,

1).

That

definition,

therefore

the

applicability

of

the

rule,

would

have

included

1,1,1—trichiorosohona,

a compound

exempt

under

the

definition

of

volatile

organac

mares’.

Twenty~fiveexamples of

other

compounds

qualifying

as volat±lo

organic liquids,

due

to

vapor

pressure

and

storage

temperature

were

provided

(Ex.

13,

Tables

I

and

II),

As

originally

proposed,

those

tanks

storing

volatile

orpurric

liquids would have been

required

to

be

equipped

with

full cont~ci:

floatin

roofs and

secondary

seals

•

Due

to

revisions

in

thu

emission

factors

for

contact

roofs

and

secondary

seals,

the

proposal

was

changed

at

hearing

to

only

require

internal

I ioatin:~’

roofs

and primary seals

CR.

362—363,

922—924).

Finally,

it

was

proposed

that thrLks

storing

organic

liquids

be

subject

to

record

keeping

and

inspection

programs.

An

organic

1

iquid was to

59-540

---

~29~

defLiet

as

any

organic

material

other

than

a

petroleum

liquid,

whica

has

a

vapor

pressure

of

0~9psia

or

greater

when

it

is

at

equtlibrnr~

wthh

its

own

vapor

at

70°

F

(Ex0

1).

Au

vu

‘tth

case

when

the

original

rule

was

adopted

in

R

7i~23,

the

facilitet

most

likely

to

be

affected

by

the

proposal

are

the

bulk

storage

operations0

Others

affected

are

those

facilities

which

maintain

sufficiently

large

storage

containers

as

a

part

of

the

overCall

operation0

Five

bulk

terminals,

owning

and

aperating

36

tanks

in

nonattainmcrt

areas,

were

identified

a° po~s

rly

being

nffecned

due

to

the

redef

ning

of

VON

for

purpace~

of

‘his

Section,

ssum’ng

that

these

36

utorage

containers

are

equ

pp~d

only

with 1ixed roofs,

annual emissions were estimated

hr tth

Agency at i30 tons per year

(Ex0 17, Table IlI)~ Using

~‘-

~arage

control efficiency of

95 percent, the snnaax emissions neduct~’ons

were estimated to be

i24 tons per year, or in other wct~, th

annual emiss~onsfrom these sources,

if regulated, would

c

~l~c—

tively be s~xtons0

Industry estimated annua’ eniss~osafrom

tanks equipped with fixed roofs to contribute only 70

a

o~r

year, and if required to install floating roofs, emi’si-ns w

be reduced by 6i tons per year,

or emissions after reulat~oi

would be

9 tons per year

CR0

446).

The difference in the twc

sets of emissions estimates

is that industry~sfigures d~c~t

include tanks containing i,1,1,~trichloroethane (R, 44?,

Si.

~,,

As noted above, this compound

is exempt unden the deiiniL~oi

~‘.

volatile organic material0

The Board exempted this compo~~iu

in

RACT

II ~R 80~5),finding that

it did not appreciably cnto~th~o

to osone formation and, therefore,

is not pronerly regula ~i

unde’- ±‘an213

(49 PCB 76’ October

5,

1982).

Nevertns~&

~.

£

Agency would have it included under the definition of v~

a

o’-3~ricliquids

and have containers storing

sthject

;~

Is

tion

(h,

455;

Ex.

17

Table

IV),

The initial objection to the proposal by iodast:~v

~

the emipmeth spec4 fications and the recordkeep4ag and

requiraments were mon

stringent than those required at Th

storage tanks originally regulated and containing materie~

~i’~

volatile,

Installation or full contact floating roofs w a e;

a

on the average to cost double than for pontoon style ~nta

exThrra~roofs

(Ex.

18).

Secondary seals were estimated ~o

S6~b~Th4for the affected tanks in Illinois, and only ~c

~

reduced emissions of 6.7 tons per year

CR.

450,

Ex,

18).

~Thoidy mantioned, due to revisions in the applicable emL

~t(torn, and acknowledging that contact floating roofs law

CO~Xc.

~tonai.difficulties than pontoon type roofs, the laeac*

amerCad its proposal, eliminating these two requirements.

q~5

t.rkv

would only be required to install floating roofs wi h

single seal or a vapor recovery system, which

is the same ~a

require-i of tanks storing VON under the existing rule,

Rves after the proposal was rewritten,

industry ohje~ted

ararr’ng that the costs incurred were not justified given the

59~54

I

---

small awunt of reduced emissions involved.

The Agency est~1rated

that

installatior

of internal floating roofs would cost in tIe

range

or

$?,80t

and $7,800 and anticipated that this figure 4ould

be

offset

by pvod~~tloss savings (Ex~17, Table III)~

fly

Agency

estimated the cost effectiveness to range Letw en $45 art

$5,600

pen ~or

On the other hand,

industry estinated

ti’s cost

of

installing non~contactfloating roofs~ i.e., pontoor typ~

roofs,

to range between $12,000 and $26,000, and cost offec~-

tiveness

to he $8,975 pe~ ton overall,

Since the affeet~v I

ities generally do not own t~tecommodity, product loss

a~

were not included to offset the installation cost,

whi~

revenue count instullation was included in estimating t

costs

(P. 44T,

Ex. 18),

Finally, industry’s figures alec

*

~uded

costs for cleaning the tanks prior to installation.

The’

ferences, along with the conflicting emission estImates a~

to account for the wide differences between the two estir~-~n,

The

EelS

estimated cost effectiveness, on the average

tc

Fc

$2,328

(Fx

48, p

4~15;.

Since this figure is closer to t.~

Agency’s

estimate, and since the figure does not appear

n

~

ate

lost revenue

and cleaning

costs,

it

is assumed that they ~

~enot

included.

Based

on the Agency’s estimates and emission racto~~r

trolled

these sources collectively contribute no more thnn ~3

tons

per year, or or

the average only 3.6

tons indiviciu~n

/

year.

It

must be remembered that these figures assume

tIns

-~ls

storing

1,1~trichloroethane,would be regulated.

S~ncc a

w

evidence ~raspresented to support controlling this excr~

c

as

an

ozore precursor, that assumption must be rsje~ThC

*

estimated unsoatrolled emission figure of 70 tans per y~

accurately

represents the amount of enissiors at issue

the Agency’s

cost figures

are accepted

as accurate,

,~

*

installation of internal floating roofs is not jU~tIfi~c

a

other

industrial categories, sources individually emitci~

than 100 tons per year are exempted from regulation.

~

*

time these sources shall be subject only to Subpart~P.

*

be demonstrated that such floating roofs are already in pJa

a majority of the facilities,

so the cost figures are evis:o~’d

to be non~existent,then reconsideration would be proper

As

aside,

tne Board notes that this rulemaking

is

~ot nece’~’vav

‘

r

a

satisfactorj State Implementation Plan as required oy o~

since the sources identified in this rulemaking are rot,

U

ally or collctively, emitting more than 100 tons of ~o

t

a

organic

materials per year.

Two other questions are outstanding:

~hether recori’U

should

be required of these tanks storing organth liqus

whether

the proposal,

as amended, encompassed tar storage

at steel—making facilities,

including

those at coke by—prcdin

recoverj plants.

While they both are mooted with the de’

~.

not to udpt the proposed rules, some

discussion is warm

To impiei~en~

these regulations,

a definition of orgaile

---

—31—

in addition to the definition of volatile organic liquids, was

proposed.

That definition set 0.9 psia at 70°F as the minimum

vapor .pressure for organic liquids.

Use of this term,

in conjunc-

tion with the definition of volatile organic liquid

(as well

as

in lieu of

volatile organic material)

causes problems.

First,

it

is

unclear why

the applicability of the recordkeeping requirement

was premised on the

term organic liquid.

Did this mean

that

those containers storing volatile organic materials were

not

included?

Did it mean

that those containing volatile

organic

liquids were required

to keep records?

If it was intended

to

encompass

all

storage tanks regulated and proposed

to be regulated

under Subpart

B

of

the Board~s

rules,

it was never so stated~

If

it

was intended

to only require records from tanks which

are

heated,

this

likewise was never explained.

Secondly,

the possible

conflict between the

two

definitions caused

problems for the

steel

industry.

Only at hearing did the possibility

arise

that

the storage container rules were to

be

applicable to

storage

tanks containing tar at steel facilities,

Industry argued that

tar

is

not an organic liquid under the proposed definition.

At

equilibrium with its own vapor at 70°F,

tar has a vapor pressure

less than 0.9 psia.

Reading the definitions together, industry

argued that since tar is not an organic liquid,

it cannot be a

volatile

organic liquid,

and therefore the containers

are not

subject to the rules,

The Agency argued that the

rules must be

read independently, and should tar~svapor pressure at

storage

temperature exceed

1.5

psia,

the regulations would be

applicable.

(R.

1401;

P.C.

26),

Both problems apparently are caused because

the proposed

definition

of

organic

liquid

did

not encompass all

volatile

organic

liquids in the

same

fashion

that

the

definition

of organic

materials encompasses volatile organic materials,

This demonstrates

the difficulty in regulating pursuant to a definition

of volatile

organic

liquids instead of amending the term volatile

organic

materials.

Both problems remained unresolved and although

they

are

mooted,

it

is troublesome,

Has the Agency,

in its permitting

capacity,

found it desirable that the storage tanks

now subject

to Subpart B maintain records?

If

so, this was not

made clear to

the

Board

at hearing.

Secondly, did the Agency choose

not to

provide information documenting the need

to control tar storage

tanks

either

under this

category

or

in the industry specific

proposal

for

the coke by—product recovery plants because

it

assumed

that

they would be regulated hereunder?

If either be the

case,

the Board

Subpart K

Section 215.305:

Viscose Casing Process

As the

Generic Rule was proposed, two Illinois manufacturers

of

regenerated cellulose

casings would have been required

to

control

the

carbon disulfide emissions from

their plants.

One

59-543

---

—32—

company

is located

in

a

nonattainment

area~

the

other

:Ln

an

attainment area,

Based on

the

descriptions

offered

by

each, the

viscose

processes

of

both

companies

appear

to

be

similar,

c;aseous

e~issions

of carbon

disulfide and hydrogen

sulfide

occur

at the

coagulatiori,~

regeneration

and

purification

stages

of

the

proceSS

(R.

933—934, 975—976).

The

company

in

the

nonattainment

area

estimated