BEFORE THE ILLINOIS POLLUTION CONTROL BOARD

)

)

)

)

)

R 05-20

IN THE MATTER OF:

)

)

PROPOSED AMENDMENTS TO

EXEMPTIONS FROM STATE

PERMITTING REQUIREMENTS

FOR PLASTIC INJECTION MOLDING

OPERATIONS

(35 III. Admin.

Code 201.146)

)

NOTICE OF FILING

TO:

Ms. Dorothy M. Gunn

Clerk of the

Board

Illinois Pollution

Control Board

100 West Randolph Street

Suite 11-500

Chicago, Illinois

60601

(PERSONS ON ATTACHED

SERVICE LIST)

PLEASE TAKE NOTICE that

on July

1,2005,1 filed with the Office of the Clerk

of the Illinois Pollution Control Board, by electronic mail,

the attached Chemical Industry

Council of Illinois’

First Errata Sheet, a copy of which is hereby served

upon you.

Dated:

July

1, 2005

Respectfully submitted,

CHEMICAL INDUSTRY COUNCIL

OF ILLINOIS

By:

/5/

Patricia F. Sharkey

One of its Attorneys

Patricia F.

Sharkey

Mayer, Brown, Rowe

&

Maw LLP

71

South Wacker Drive

Chicago, Illinois

60606-4637

(312) 782-0600

\~

---

BEFORE THE ILLINOIS POLLUTION CONTROL BOARD

IN THE MATTER OF:

)

)

PROPOSED AMENDMENTS TO

)

EXEMPTIONS FROM

STATE

)

PERMITTING REQUIREMENTS

)

FOR PLASTIC INJECTION MOLDING)

R 05-20

OPERATIONS

)

(35

III. Admin. Code 201.146)

)

CHEMICAL INDUSTRY

COUNCIL OF ILLINOIS’

FIRST ERRATA SHEET

The

Chemical Industry Council of Illinois

(“CICI”), by its attorneys Mayer,

Brown, Rowe & Maw LLP,

hereby submits the following corrections

and amendments to

documents previously filed in this

proceeding:

AMENDMENT TO PROPOSED REGULATORY LANGUAGE

CICI

proposes to amend the text of its proposed regulatory language as follows:

TITLE 35:

ENVIRONMENTAL PROTECTION

SUBTITLE B:

AIR POLLUTION

CHAPTER 1:

POLLUTION CONTROL

BOARD

PART 201

PERM1TS AND GENERAL PROVISIONS

Section

201.146

Exemptions from

State Permit

Requirements

Construction or operating permits, pursuant to Sections 201.142,201.143,

and 201.144 of

this Part, are not required for the classes of equipment and activities listed below in

this

Section.

The permitting

exemptions in this

Section

do not

relieve the owner or operator

of any source

from any obligation

to comply with any other applicable requirements,

including the obligation to

obtain

a permit pursuant to Sections 9.1(d) and 39.5 of the

Act, Sections

165,

173, and 502 of the Clean Air Act or any other applicable permit or

registration requirements.

hhh)

Plastic

injectjon-eempression, and tranGfer molding equipment, and associated

lastic resin

loadin

unloadin

conve

in

inixin

storage, grinding,

and dr

n

e

ui

ment and

associated

mold release a

ents.

---

CORRECTION TO PRE-FILED TESTIMONY

OF LYNNE R. HARRIS.

On page 5, line

10 ofthe Pre-Filed Testimony of Lynne R. Harris on Behalf of the

Society of the Plastics Industry, Inc.

filed on

June

16,

2005

is corrected as follows:

“Illinois),

particulate matter (PM

10 Total

Particulate. referred to herein as PMJ,

and

a variety of hazardous airpollutants (HAPs).”

Respectfully submitted,

CHEMICAL

INDUSTRY COUNCIL

OF ILLINOIS

By:

/s/ Patricia F. Sharkey

One of Its Attorneys

Dated:

July

1. 2005

Patricia F. Sharkey

Mayer, Brown, Rowe & Maw LLP

71

South Wacker Drive

Chicago, Illinois

60606-4637

(312) 782-0600

---

CERTIFICATE OF SERVICE

I, Patricia F.

Sharkey, an attorney, hereby certify that I have served the Chemical

Industry Council of Illinois’ First Errata Sheet upon:

Ms. Dorothy M.

Gunn

Clerk ofthe Board

Illinois

Pollution Control Board

100 West Randolph

Street

Suite

11-500

Chicago, Illinois 60601

(Electronic Mail)

Matthew Dunn, Chief

Division ofEnvironmental Enforcement

Office of the Attorney General

188

West Randolph Street,

20th

floor

Chicago, Illinois 60601

(U.S.

Mail)

Donald Sutton

Manager, Permit

Section

Division of Air Pollution

Bureau ofAir

Illinois Environmental Protection Agency

1021

North Grand Avenue East

Post Office Box

19276

Springfield, Illinois 62794-9276

(U.S. Mail)

Charles E. Matoesian

Division ofLegal

Counsel

Illinois Environmental Protection

Agency

1021

North Grand Avenue East

Post Office Box

19276

Springfield, Illinois 62794-9276

(U.S. Mail)

Office ofLegal

Services

Illinois Department of Natural Resources

One Natural Resources Way

Springfield, Illinois

62702-127 1

(U.S. Mail)

as indicated above, by electronic mail or by depositing said

States Mail, postage prepaid, in Chicago, Illinois on July

1,

documents in

the United

2005.

Patricia F. Sharkey

Mayer, Brown, Rowe & Maw LLP

71

South Wacker Drive

Chicago, Illinois

60606-4637

(312) 782-0600

/s/ Patricia F. Sharkey

Patricia F.

Sharkey

---

BEFORE THE ILLINOIS POLLUTION

CONTROL

BQARD

HECEIVED

IN THE MATTER OF:

)

CLERK’S

OFFICE

JUN

2005

PROPOSED AMENDMENTS TO

)

EXEMPTIONS FROM STATE

)

STATE OF

ILLINOIS

PERMITTING

REQUIREMENTS

)

Pollution Control

board

FOR PLASTIC INJECTION MOLDING)

R 05-20

OPERATIONS

)

(35 Ill. Admin. Code 201.146)

)

PRE-F1LED

TESTIMONY OF LISA

FREDE

ON BEHALF

OF THE

My name

is Lisa Fredc,and.J am the Dkectorof Regulatory Affairs for the

Chemical Industry Council of Illinois (“CICI”),

a not-for-profit Illinois corporation.

CICI

is pleased to be

the proponent of the rulemaking proposal

in this proceeding.

I would

like

to begin by giving

you an

overview of CICI

and its membership and

then briefly discuss the significance ofthis

proposed rulemaking

to our members.

CICI

is

a

statewide trade association

representing the chemical industry in

Illinois.

CICI has offices in Des Plaines

and Springfield, Ilinois.

We have

198 member

companies

with over 54, 000 employees employed in

745

manufacturing facilities and

975

wholesale and

distribution facilities

in Illinois.

One of CICI’s functions is to

represent its member companies

in the formation of

public policies and programs which

are mutually beneficial to

the citizens of Illinois and

the chemical industry.

In this capacity, CICI

monitors statewide legislation and

regulations in Illinois, including environmental

permitting programs,

and provides

information and

makes recommendations to its

membership.

CICI also often advocates

on

behalfof its

membership for more cost effective and

efficient regulatory

requirements.

(

/

\~_4A’

U

THis

DOCUMENT

IS PRINThD

ON

RECYCLED PAPER

~

~

---

Chemical manufacturers in Illinois produce a wide array ofproducts from

plastics,

pesticides and industrial

chemicals to lifesaving medicines

and household

products.

Workers directly employed in the chemical

industry represent 7.3

of the

state’s manufacturing work force and have an

average

wage over $60,000 per year.

The

chemical

industry generates

an additional 296,000 jobs in Illinois

at industry suppliers,

manufacturers, transporters, trade and business services companies,

and construction

companies.

The proposal

in this

proceeding will amend the Board’s regulations governing

state air pollution control permits

to exempt plastic

injection

molding operations from the

state construction

and operation permitting procedures.

CICI

is proposing this

amendment to clarify the Board’s regulations and achieve efficiencies and cost savings

for its plastic injection

molding company members in Illinois and for the State permitting

program.

As will be discussed by another witness

in

this proceeding,

the emissions from

plastic

injection molding machines

are extremely low

—

on

the order of a few tenths of a

ton of volatile organic emissions per year.

This

is on the

order of

—

and in fact less than

—

the 0.1

lb/ hour or

0.44 tons per year that defines

an “insignificant

activity” under the

Board’s major source regulations

at 35

Ill. Adm.

Code 201.210 (a)(2) and(3).

These emission

levels are also on the order of

—

or less than

—

the emissions

recognized to be

associated with other categories of emission sources that

are currently

exempt from state permitting under Section 201.146.

In

fact, the emission factors

accepted by Illinois EPA

and other regulators across the country for determining

emissions from plastic

injection

molding operations are the same as those that

are used

-2-

THIS DOCUMENT IS

PRINTED

ON RECYCLED PAPER

---

for plastic

extrusion

—

a process which

is exempted from Illinois state permitting in

Section 201.146(cc)

and defined as an “insignificant

activity” in Section 201.210(a)(5).

While many owners

and

operators believe

that “plastic

injection molding” is

a form of

extrusion covered under the existing categorical exemption, the

adoption of the

specific

language proposed in

this rulemaking is designed to

resolve any question.

Here’s what this amendment will do:

•

It will appropriately regulate the insignificant level ofemissions

generated by

plastic

injection

molding operations by treating those operations

in

the same

fashion

as other operations with similarly low

levels of emissions.

•

It will reduce unwarranted permitting costs to plastic injection molding businesses

across Illinois.

•

It will also relieve owners

and operators ofplastic

injection molding operations

from the risk of enforcement actions based upon differences

in interpretation of

existing categorical exemptions.

•

Finally, it will

allow

flinois EPA to allocate

itspermitting

and enforcement

resources to more significant emission sources.

What this

amendment will

not

do:

•

It will

not relieve affected emission units

from any applicable requirement other

than

state construction and

operating permitting.

Thus, for example, a plastic

injection molder

—

like

any otherexempt

emission source under Section 201.146

—

remains subject to

the generic

volatile organic matter emissions

limit of

8

lb/hour

found in

the Board’s rules at 35

III. Admin.

Code

215.301.

•

It will

not result

in

an increase in

emissions and will

not

have an impact on

air

quality in Illinois.

Because this

is only

an exemption from procedural

requirements, it will not affect

emissions to the environment.

Prior to proposing this regulatory amendment, CICI’s Executive Director, Mark

Biel, had several discussions with Don Sutton,

the Manager of the Illinois EPA Permit

Section, about

adding

a categorical

exemption to the list ofexisting categorical

exemptions in

35111.

Admin.

Code

§

201.146

for plastic injection molding and associated

resin

handling and

storage activities.

Mr. Sutton

agreed that this is a category of

-3-

THIS DOCUMENT IS PRINTED ON RECYCLED PAPER

---

insignificant emission sources for which a permit exemption is

consistent with other

categorical exemptions in Section 201.146.

He also agreed that relieving the State of the

burden of permitting these insignificant sources would

be beneficial to the State.

CICI believes that reducing the permitting burden on

the Agency is in the interest

of its members.

Agency resources

should be focused on significant emission sources.

In

the pending rulemaking proceeding, R05-19,

Mr.

Sutton

testified that the Agency still

hasn’t issued 30 of the Title V major source permits that were due to

be issued back

in

1997.

Transcript, pp. 29-30, April

12, 2005 Hearing, IPCB Docket R05-19.

In addition,

CICI

is aware

that many of its members have Title V permit renewals and permit

revisions that have been pending before the Agency for several years.

Mr. Sutton

testified that while

IEPA issues roughly

1,900 air permits

a

year, it has at

any time

a

backlog of 900

to

1,000 permit

applications.

j4,

p.

31.

Yet

the Agency is required to

spend its resources on

a host of construction

and operating permits

for very minor

emission sources.

The transcript of the R05-19 April

12, 2005 hearing reveals that 70

of the Agency’s construction

permits are issued

for modifications involving no emission

increase or increases of less than

I

ton.

Id. p.

12.

At the same

time, 95

of the actual

emissions emitted

in

Illinois are emitted by the top

15

of the State’s major sources.

Id.

p. 53.

Permitting very

small emission sources, while large emission source applications

are backlogged isn’t

a good use of tax dollars, it isn’t good for the environment, and

it

isn’t good for regulated businesses.

That burden

will

be

significantly reduced when the rulemaking

in R05-19

is

adopted.

However, because that rulemaking

only exempts insignificant emission sources

at facilities with other

signiji can:

or non-exempt emission

sources,

it does not

relieve the

-4-

THIS DOCUMENT IS PRINTED ON RECYCLED PAPER

---

Agency from permitting

a

plastic injection molding facility that has

no

other emission

sources.

This is an anomaly with no rationale in terms ofemissions

or

the environment

when it

comes to plastic injection molding.

Given the

limitation in

the proposal in

R05- 19. the adoption

of a clear categorical exemption forplastic injection molding

operations in this rulemaking proceeding will harmonize the Board’s regulatory approach

for a category recognized

by

all to emit at levels

that

do not

warrant separate state

permitting.

CICI

would

like

to thank the Board for

its consideration

o. this proposal, and I

would be happy to

answer any questions

you

may have.

Date:

Ct

(IL1.! O6

Respectfully

submitted,

Lisa Fiede

Director of

Regulatory Affairs

Chemical

Industry

Council of

Illinois

-5-

Tills DocUMENTis PRINTED

ON

RECYCLED PAPER

---

ORIGINAL

BEFORE THE ILLINOIS POLLUTION CONTROL

BOARD

IN

TUE

MATFER OF:

)

)

PROPOSED AMENDMENTS TO

)

EXEMPTIONS FROM

STATE

PERMITTING REQUIREMENTS

)

FOR PLASTIC

INJECTION

MOLDING)

R 05-20

(35 IlL Admin. Code 201.146)

)

PI4E-FILED

TESTIMONY OF

LYNNE 9. HARRIS

ON

BEHALF OF THE

SOCIETY OF THE PLASTICS INDUSTRY, INC.

My

name

is

Lynne

R.

Harris, and

I am

the

Vice

Presideut, Science and

Technology, forThe Society ofthe Plastics Industry, Inc. (“SPI”),

a not-for-profit

501(c)6

trade association

headquartered

in Washington D.C.. predoininaniy

serving

members across the United States.

I have been employed

by SN for over H years. My

current work focuseson

science

and

technology, environment, health

and safety, and

codes

and standards for the

plastics industry. My educational

background includes a

Bachelor ofScience

and

Mastersof

Engineering

in chemical engineering.

My

publications include co-authorship on a

paper

for the development

ofemission factors for

the extrusion processing ofpolyethylene resin.t

I have worked in and

around

the plastics

industry

forover 25

years.

I have been asked

by

the

Chemical Industry Council

of Illinois (CICI) to provide

an overview ofthe plastics Injection molding industry,

a description ofthe plastic

injection molding process,

and

a discussion ofthe types

and

volumes ofemissions

generated during the plastic injection

molding process for various resins.

RECEIVCT

CLEPtCS

~:~‘

-

JUN

16

STATE

OF lWi’aic

po~ut(ofl

Control

BOSEG

4)

---

The

Society of

the

Plastics Industry~Who Are

We?

Let

me

begin

by describing SN

and

the work it

performs

on

behalf

ofits

members. Founded in

1937, The Society ofthe Plastics Industry, Inc..

is the trade

association representing one of the largestmanufacturing industries in the United States.

SN’s members represent the entire

plastics industry supply chain, including processors,

machinery and

equipment

manufacturers and raw materials suppliers. The U.S. plastics

industry employs 1.4

million workers

and provides more

than $310 billion

in

annual

shipments. SN

represents the

entire plastics

industry

arid

has

more

than

1000

members.

SN has been involved in the development ofstate

and federal

environmental regulations

affecting

the plastics

industry

for decades.

As

I will be discussing. SN

has also

coordinated a

number studies of

emissions

generated by the extrusion

processing

of

therrnoplastics.

Backsronnd on the Plastic Iniection ~ol4ipg

i~dnstry

My testimony today is

focused

on

plastic injection molding (“VIM”),

a category

ofplastic

product manufacturing. There

are

over 7,700

NM facilities in the United States

and

approximattly 500 operating in Illinois!

These facilities range in size

from small

facilities with a few machines

and

less

than

20 employees to larger facilities with dozens

of machines employing over a hundred employees.2’4 The tradepublication

Plastics

News

surveys the PIM industry annually

and publishes an

annual

listing of over

600

P1M

companies

in

North America. That listing

indicates the top

NM companies responding to

the survey with

annual

sales ranging from approximately $100,000

to

$1.5

billion, with

median

annual sales

on the order of $10 million. The components produced in

ViM

processes are generally small plastic components

used

in

a multitude of products. For

2

---

example.

RIM products include knobs

and

handles used in the automotive

industry and

hole plugs

used

in household appliances.

VIM products tend to be

molded

to

meet

specific

needs

in customized molds and

made

with resins meeting the temperature,

strength

and

durability specifications required for a specific use.

As

a result, VIM

machines

are

generallydedicated to molding specific component parts and cannot be

used

to

produce other parts without physical modification of the equipment.

Description of

NM

Etuinment and Process

The PIM process essentially involves

forcing molten plastic into a mold cavity.

This

takes

place in several steps.

A diagram of a standard RIM

machine,

attached

to my

pre-flied

testimony, depicts the components ofthe NM process.

Exhibit

I.

As can

be

seen

from that diagram, the essential components

are a hopper from which pelletized resin

is

fed

into the extruder screw, a heated extruder barrel which melts the resin

as

it

is

advanced by

the extruder screw under

pressure, and

a die headthrough which the molten

resin

is

injected into a mold cavity.

Note

that

the fundamental piecc ofequipment involved

in this process

is a heated

screw extruder.

The equipment that is required to extrude resin

into moldsin the

VIM

process

is the same as that which is required to extrude

resin

into a continuous

strand

except that the

resin

is injected into an enclosed mold at the end of thc process rather

than

simply

conforming to

the shape of

the extrusiondie. A VIM machine is essentially a non-

continuous extruder. As

I will

discuss later, this is whythe

emission factors developed for

extrusion processes

arc

appropriate for the

VIM process.

Plastic

injection molding machines,

like other types ofextruders.

vary

in size. A

small

VIM machine may have

a throughput of 10

pounds

per hour,

while

a large machine

3

---

may process as much as 200 pounds

per

hour. These numbers are derived

based

on a

typical injection capacity of4 to

100 ounces andtypical tonnage

of50 to 600 tons.

Injection

capacity can

go

to around 400 ounces

and tonnage can

go

up to

around

10,000

tons.5

These data

are

consistent with product information compiled from several

equipment manufacturers, as

illustrated in

Exhibit 2.

Very large PIM machines

can

processover

1,000

pounds per hour. VIM machines ofall sizes are in use in Illinois and

across the United States. However, the most commonly used machines in the JIM

industry

have an average daily

throughput of less than 100

pounds per hour.

The five most commonly

used plastic resins

in

the

VIM

industry

according to the

2005

survey of

North

American

injection molders

by

Plastics News2

are polypropylene

(PR),

acrylonitrile

butadiene styrene (ABS), polycarbonate (PC), high

dtnsity

polyethylene (1-IDPE)

and nylon (po!yaniide,

PA).

Emissions from Extrusion Processes

Until 1995, little quantitative infoTmation

was

publicly available regarding

emissions from thermoplastic extrusion processes.

While it was

assumed

that

any volatile

organic, particulate or hazardous air emissions were very low,

emission

factors simply

did not

exist. To fillthis gap, SN sponsored a number of studies published between

1995

and

2002 to develop emission factors for a range of plastic resins.

The studies were

intended to provide emission factors for processors who needed Tide V permits

under the

US

Environmental Protection Agency Clean

Air Act Amendments of 1990.

The SF1-sponsored studies were conducted at an independent testiug laboratory

operated by Battelle

in Columbus. Ohio.

Studies were conducted using a

strand extruder

with a

1.5-inch single screw

and fitted with an eight-strand

die for commonly used

resins.

4

---

Resins with basic

additives were

provided

by

a number ofsuppliers and tested as

aggregates;

the

resins

tested were PP, PC, YE,

PA and ethylene-vinyl acetate and

ethylene-methyl acrylate copolymer

(EVAJEMA.

The

extruder

system

was

chosen as the process Likely

to

overestimate emissions.

As a continuous system,

it

was

anticipated to mimic extrusion processes and overestimate

closed mold operations, such as injection molding. This assumption

was supported

by

a

two-year

study

that found extrusion processes generated a

higher level of emissions than

injection

molding.6 Emissions

from the die head

of

the

extruder

system were captured

and analyzed

for volatile organic compounds (VOC; volatile

oTganic material

or VOM in

Illinois), particulate matter (PM-I0), and avarietyof hazardous air pollutants (HAPs).

The

SPI sponsored

studies ofthe commonly used resins PP

PS, FE

and PA

are

attached to my pre-flied testimony as

Exhibits 3 —6

and witS

be

referred to herein as

the

“SPI Studies.”

The

EVA/EMA study

(Exhibit

7)is provided for informational purposes.

A study on ASS, conducted at the same laboratory as

the SPI Studies,

is also provided

for

informational purpose&

Exhibit

8.

That study was not conducted under SN

auspices, and

thus

I

have limited knowledge of the conditionsunder which it

was

performed.

The above-mentioned studies form the basis for the plastics

industry’s

understanding ofemissions from

these processes and are recognized by industry and

regulatory authorities,

as defining emission factors

for

both simple extrusion and the

extrusion processutilized

in PIM.

What these studies demonstrate is that extrusion processing of differentresins

under various operating conditions produces different types and amounts ofemissions.

Exhibit

9 attached to my pie-filed testimony is a

chart summarizing the emission

factors

S

---

devcloped

in the

SPI

Studies for eachofthe emissions ofinterest for the resins studied.

The

information in this

chart was compiled

from irifonnation contained in each of the

SN

Studies

to make

it

easier to review this data in this

proceeding.

As

can

be seen from

this chart,

the emissions of

interest

include VOM, PM and a

variety of HAN.

The type and volume ofemissions varies from a high ofapproximately

0.4

lb

of

VOM per ton

of

resin

processed to a low of approximately 0.1

lb

per

ton of

resin

processed. HAl’s ranged

from

a high ofapproximately

0.3 lb per tort ofresin processed to

a low of approximately 0.02

Sb per thousand tons of

resin processed.

Particulate

emissions ranged from a high

ofapproximately 0.5

lb

PM per ton ofresin processed

to a

low ofapproximately 0.02

lb PM

per

ton of resin processed forthe commonly used

resins.

Exhibit

JO

Based

on

the emission factors developed

in the

SF1 Studies and the

capacity of

PIM machines, across the range from small to large P1M machines discussed

above,

one

can

obtain an overviewof the annual

volume of

emissions associated

with NM

processes.

Exhibit /1

to my

pie-filed

testimony is a chart showing the estimated volume

of VOM, PM

and

HAP emissions in tons

per year, associated

with the various types of

resins studied by

SPI. As

can

be seen from this

chart,

the emissions range

from a high of

0.2 tons

per

year of VOM to a low of 0.002 tons per year VOM. HAP

emissions range

from 0.1 tons

per

year to 0.0004 thousandths ofa ton

per

year. PM emissions

range from

0.2

tons per year

to

0.0004

tons per

year.

That concludes my pre-Illed testimony describing the PIM industry, NM process

and

types

and

volumesof emissions associated with the processing ofvarious resins.

I

6

---

appreciate the opportunity

to testify

and am

available to answer

any

questions

the

Board

or other participants in

this proceeding may have.

Re

cifully submit

On

half

of

The Society ofthe Plastics Industry, Inc.

‘Barlow,

A.;

Cantos,

P.; Holdren, M. W.;

Garrison,

P.; Hartis,

1.; Janke, 8. (1996).

Development of

emission

factors

for polyethylene

processing.

J.

Air &

Waste

Manage.

Assoc.,

46,

569-580.

22002

Economic

Cenans, Manufacturing

lndusuy Series,

All

Other

Plastics

Product

Manufacturing: 2002.

US Census Bureau, ECOZ-3 11-326199 (RV).December

2004; p.2.

‘SPI

Plastics

Data

Source.

(2001).

State-by-State

Guide to Resin

and Equipment,

p.

A-2.

4Sun’ey of

North Amedcan

Injection Molders.

Plastics News.

April

II, 2005.

Rosato,

DV.,

Rosato,

DV. and

Rosato,

MG.

(2000).

Injection Mo/ding Handbook

3’~

ed.

Boston:

Kluwer Academk Publishers.

p.

28.

‘Forrest,

Mi.,

Jolly,

A.M.,

HoLding,

SR., and

Richards,

S.J.

(t99S.

Emissions

Cram Processing

Thennoplastics.

Annol.c ofoccupational hygiene.

39(l),

35-53.

7

---

Pre-Filed Testimony of Lynne Harris

IPCB Rulemaking Docket R05-20

EXHIBITS

1.

Plastic Injection Molding Machine Diagram,

Injection Molding Handbook,

3rd

2.

Plastic Injection Molding Equipment Manufacturer Product Information.

3.

Adams, K.; Bankston,

J.; Barlow, A.; Holdren, M.;

Meyer,

S.;

Marchesani,

V.

(1999) “Development of Emission Factors for Polypropylene Processing,”

J. Air

&

Waste Manage. Assoc., 49,

49-56.

4.

Rhodes, V.;

Kriek, 0.; Lazear,

N.; Kasakevich, J.;

Martinko, M.; Heggs, R.P.;

Hoidren, M.W.;

Wisbith, A.S.; Keigley,

G.W.; Williams, S.D.; Chuang,

S.C.;

Satola,

J.R. (2002) “Development ofEmission Factors for Polycarbonate

Processing”

J. Air

&

Waste Manage. Assoc., 52,

78 1-788.

5.

Barlow,

A.; Contos, D.; Holdren, M.; Garrison, P.;

Harris, L.;

Janke, B (1996)

“Development ofEmission Factors for Polyethylene Processing”

J. Air &

Water

Manage. Assoc., 46,

569-580.

6.

Kriek, 0.; Lazear, N.;

Rhodes, V.;

Barnes,

S.; Bollmeier,

S.;

Chuang, S.;

Holdren,

M.; Wisbith, A.;

Hayward,

J.; Pietrzyk, D.

(2001) “Development ofEmission

Factors for Polyamide

Processing,”

J. Air

&

Waste Manage. Assoc.,

51,

1001-

1008.

7.

Barlow, K; Moss, P.; Parker, E.;

Schroer, T.; Hoidren, M.; Adams, K. (1997)

“Development of Emission Factors for Ethylene-Vinyl

Acetate and Ethylene-

Methyl Acrylate Copolymer Processing,”

J. Air &

Waste Manage. Assoc., 47,

1111—1118.

8.

Contos, DA.;

Hoidren, M.W.; Smith,

D.L.;

Brooke, R.C.; Rhodes, V.L.; Rainey,

M.L. (1995) “Sampling and Analysis of Volatile Organic

Compounds Evolved

During Thermal Processing of Acrylonitrile Butadiene

Styrene Composite

Resins,”

J. Air

&

Waste Manage. Assoc., 45,

686-694.

9.

SPI Studies Emission Factor Summary

Chart.

10.

Estimated Emissions Using

11.

Overview of Estimated Emissions.

THIS

DOCUMENT

HAs

BEEN

PRINThD 0*1 RECYCLED PAPER

---

BEFORE THE ILLINOIS POLLUTiON CONTROL BOARD

IN THE

MATTER OF:

)

)

PROPOSED AMENDMENTS TO

EXEMPTIONS FROM STATE

)

PERMITTING REQUIREMENTS

)

FOR PLASTIC

INJECTION MOLDING)

R 05-20

OPERATIONS

)

(35

III. Admin.

Code 201.146)

)

PRE-FILED TESTIMONY OF LYNNE R.

HARRIS

SOCIETY OF THE PLASTICS INDUSTRY, INC.

My name is

Lynne R.

Harris, and

I am the Vice President, Science

and

Technology, for The Society ofthe Plastics Industry, Inc.

(“SPi”), a not-for-profit

501(c)6

trade association headquartered in Washington D.C., predominantly serving

members across the

United States.

I have been employed by

SPI for over

14 years.

My

current work focuses on science

and technology, environment, health and

safety,

and

codes and standards for the plastics industry.

My educational

background includes a

Bachelor of Science and

Masters of Engineering in chemical engineering. My

publications include co-authorship on a paper for the development of emission factors

for

the extrusion processing

of polyethylene resin!

I

have worked in and around the plastics

industry for over 25

years.

I have been asked

by the Chemical Industry Council of Illinois (CICI) to provide

an overview ofthe plastics

injection molding industry,

a description of the plastic

injection molding process,

and

a discussion of the types

and

volumes of emissions

generated during the plastic injection molding process for various resins.

RECEIVED

CLERK’S OFFICE

JUN

162005

STATE OF ILLINOIS

Pollution Control

Boarc’

z-)C

j’Ll

/

---

The Society ofthe Plastics Industry: Who Are We?

Let me begin by describing SPI and the work it performs on behalf of its

members. Founded in

1937. The Society ofthe Plastics Industry,

Inc., is the trade

association representing one of the largest manufacturing industries in the United States.

SPI’s members represent the entire plastics industry supply chain, including processors,

machinery and equipment manufacturers and raw materials

suppliers. The

U.S. plastics

industry employs

1.4 million workers

and provides more than

$310 billion in annual

shipments.

SPI represents the entire plastics industry and has

more than

1000 members.

SPI has been involved in the development of state and federal environmental

regulations

affecting the

plastics industry for decades.

As I will be discussing, SPI has

also

coordinated a number studies of emissions generated by the extrusion processing of

thermoplastics.

Background on the

Plastic Injection

Molding Industry

My testimony today is focused on plastic injection molding (“PIM”), a category

of

plastic

product manufacturing. There are over 7,700 PIM facilities in the United

States

and approximately

500 operating in Illinois.2~3These facilities range

in size from

small

facilities with a few machines and less than 20

employees to larger facilities with dozens

of

machines employing over a hundred employees.2’4 The trade publication

Plastics News

surveys the PIM

industry annually and publishes an annual

listing ofover 600 PIM

companies in North America. That listing indicates the top PIM

companies responding to

the survey with annual

sales ranging from

approximately

$100,000 to $1.5 billion, with

median annual

sales on the order of$lO million. The components produced in PIM

processes are generally small plastic components used in a multitude

of products. For

2

---

example. I~JMproducts include knobs and handles

used in the automotive industry and

hole plugs used in household appliances. PIM products

tend to be molded to meet

specific needs in customized molds and made with resins meeting the temperature,

strength and durability specifications required for a specific use.

As a result,

PIM

machines

are generally dedicated to molding specific component parts and cannot be

used to produce other parts without physical modification ofthe equipment.

Description of PIM Equipment and Process

The PIM process essentially involves forcing molten plastic into a moldcavity.

This takes place in several

steps.

A diagram of a standard PIM machine, attached

to my

pre-filed testimony, depicts the components of the PIM process.

Exhibit 1.

As can be seen

from that diagram, the essential components

are

hopper

m which pelletized

resin is

fed

into the

xtruder screw,

heatej~de~~~er

band

hich melts the resin as it is

advanced by the extruder screw under pressure, and

a~f

head)rou~h

which the molten

resin

is injected into a mold cavity.

Note that thej~~damental

piece ofequipment involved in this process

is a heated

screw extru~~jThe

equipment that

is required to extrude resin

into molds in the PIM

process is the same as that which is required to extrude resin into a continuous strand

except that the resin is injected into an enclosed mold at the end of the process rather than

simply conforming to the shape of the extrusion die.

A PIM machine is essentially a non-

continuous extruder.

As

I will discuss later, this is whythe emission factors developed for

extrusion processes

are

appropriate for the PIM process.

Plastic injection molding machines,

like other types of extruders,

vary

in size.

A

small PIM

machine may have a throughput of

10 pounds per hour, while a large machine

3

---

may process as much as 200 pounds per hour. These numbers are derived based on a

typical injection capacity of 4 to

100 ounces and typical tonnage of 50 to

600 tons.

Injection capacity

can

go

to around 400 ounces and tonnage can go up to around

10,000

tons.5

These data are consistent with product information compiled from

several

equipment manufacturers, as illustrated in

Exhibit 2.

Very large PIM machines can

process over 1,000 pounds per hour. PIM machines of all

sizes

are

in use in

Illinois and

across the United States.

However, the most commonly used machines in the PIM

industry have an average daily throughput of less than

100 pounds

per

hour.

The five most commonly used plastic resins in the PIM industry according to the

2005 survey of North American injection molders by

Plastics News2

are polypropylene

(PP), acrylonitrile butadiene styrene (ABS), polycarbonate (PC), high density

polyethylene (HDPE)

and nylon ~o1yamide,

PA).

Emissions from Extrusion Processes

Until

1995, little quantitative information was publicly available regarding

emissions from thermoplastic extrusion processes. While it was assumed that any volatile

organic, particulate or hazardous

air

emissions were very low, emission factors simply

did not exist. To fill this gap, SN sponsored a number of studies published between

1995

and 2002 to develop emission factors fora range of plastic resins.

The studies were

intended to provide emission factors for processors who needed Title V permits under the

US Environmental Protection Agency Clean Air Act Amendments of 1990.

The SPI-sponsored studies were conducted

at an independent testing laboratory

operated by

Battelle in Columbus, Ohio.

Studies were conducted using a strand extruder

with a

1.5-inch single screw and fitted with an eight-strand die for commonly used resins.

4

---

Resins with basic additives were provided

by a number of suppliers and tested as

aggregates; the resins tested were PP, PC, PE, PA and ethylene-vinyl acetate and

ethylene-methyl acrylate copolymer (EVAJEMA).

The extruder system

was

chosen as the process likely to overestimate emissions.

As a continuous

system,

it

was anticipated

to mimic extrusion processes and overestimate

closed mold operations, such as

injection molding.

This assumption.

was

supported

by a

two-year study that

found extrusion processes generated a higher level

of emissions

than

injection molding.6 Emissions from the die head of the extruder system

were captured

and

analyzed for volatile organic compounds

(VOC;voiatileorganicmat~th1~r’~’ONja.~......

Illinois), p~culate matter

(PM-l~~ndavariety

of hazardous

air

pollutants (HAPs).

The

Pisponsorcdstüdies of the commonly

used resins PP~PS, PE

and

PA

are

attached to

my pre-filed testimony as

Exhibits 3

—6 and will be referred to herein as the

“SPI Studies.”

The EVA!EMA study

(Exhibit

7)

is provided for informational purposes.

A study on ABS, conducted at the same

laboratory

as the SPI Studies, is also provided for

informational purposes.

Exhibit 8.

That study was not conducted under SPI auspices, and

thus

I have limited knowledge of the conditions under which it was

performed.

The

above-mentioned studies form the basis for the plastics industry’s

understanding of emissions from these processes

and

are

recognized by industry and

regulatory authorities, as defining emission factors

for both simple extrusion and the

extrusion process utilized in PIM.

What these studies demonstrate is that extrusion processing of different resins

under various operating conditions produces different types and amounts of emissions.

Exhibit

9 attached to my pre-filed testimony

is a chart summarizing the emission factors

5

---

developed

in the SPI Studies for each of the emissions of interest

for the resins studied.

The

information in this chart

was

compiled from information contained in each of the SPI

Studies to make it easier to review this data in this proceeding.

As can be seen from this chart, the emissions of interest include VOM.

PM and a

variety of HAPs.

The type and volume of emissions varies

from a high of approximately 0.4

lb of

VOM

per ton

of resin processed to a low of approximately 0.1

lb per ton of resin

processed.

HAPs ranged from a high of approximately 0.3

lb per ton

of resin processed to

a low of approximately 0.02 lb per thousand tons of resin processed. Particulate

emissions

ranged from a high of approximately 0.5

lb PM per ton of resin processed wa

low of approximately 0.02

lb PM per ton of resin processed

for the commonly used

resins.

Exhibit

10

Based on

the emission

factors developed in the SPI Studies and the capacity of

PIM machines, across

the range from

small to large NM machines discussed above, one

can obtain an overview of the annual volume of emissions associated with PIM

processes.

Exhibit 11

to my pre-filed testimony is a chart showing the estimated

volume

of VOM, PM and

HAP emissions in tons

per

year, associated with the various types of

resins

studied by

SPI.

As

can

be seen from this chart, the emissions range

from a high of

0.2 tons

per

year of VOM

to a low of 0.002 tons per yearVOM.

HAP

emissions range

from 0.1

tons per year to 0.0004 thousandths of a ton

per

year. PM emissions range from

0.2 tons per year to

0.0004 tons per year.

That concludes my pre-filed testimony describing the PIM

industry, NM process

and types and volumes of emissions associated with the processing

of various resins.

I

6

---

appreciate the opportunity to testify

and am available

to answer any questions the Board

or other participants

in this proceeding may have.

Resp~ctfully

/

e

R. Harris

On

ehalfof

The Society of the Plastics Industry, Inc.

‘Barlow,

A.; Contos, D.; Hoidren, M.

W.;

Garrison,

P.; Harris,

L.; Janke,

B, (1996).

Development of

emission

factors

for polyethylene processing. .1

Air

& Waste Manage. Assoc.,

46,

569-580.

2002

Economic Census, Manufacturing industry Series, All Other Plastics Product Manufacturing: 2002.

US

Census Bureau,

ECO2-3 11-326199 (RV).

December

2004;

p.

2.

SPI

Plastics

Data Source.

(2001). State-by-State Guide to Resin and

Equipment, p.

A-2.

Survey of North American

Injection Molders.

Plastics News.

April

II.,

2005.

Rosato,

DV.,

Rosato, DV. and Rosato,

M.G.

(2000).

Injection Molding

Handbook.

3~

ed.

Boston:

Kluwer

Academic

Publishers.

p. 28.

6

Forrest, Mi,,

Jolly,

A.M.,

Holding, SR., and Richards, 5.1.

(1995).

Emissions from

Processing

Thermoplastics,

Annals of OccupationalHygiene,

39(1),

35-53.

7

---

PLASTIC INJECTION

MOLDING MACHINE DIAGRAM

Fig. 2-2

In-line reciprocating screw unit with

hydraulic drive schematic.

Source:

Injection Molding Handbook,

3”’ Edition, 2000, Kiuwer

Academic Publishers.

---

PLASTIC

INJECTION

MOLDING

EQUIPMENT MANUFACTURER PRODUCT INFORMATION

A-i

17

0.47

33

0.95

55

1.95

110

6.02

165

10.59

330

31.4

(1)

Cycle

Time

(sec)

10

25

25

25

25

50

(2)

Maximum

Throughput

(Ib/hr)

11

9

18

54

95

141

A-2

990

362

100

815

1100

362

100

815

1500

540

100

1215

1760

769

150

1154

2200

769

150

1154

3000

1054

200

1186

3500

1054

200

1186

4000

1054

200

1186

B-i

28

1.7

40

2.8

55

7

90

9.3

110

9.3

120

12.7

140

12.7

165

12.7

220

20.1

B-2

85

5

120

10.7

170

14.7

230

25.4

300

40.3

400

59.2

500

89.6

C-i

30

3.76

50

6.04

80

11.9

130

11.9

280

34

C-2

150

28

200

28

250

28

300

28

25

25

25

25

25

30

30

30

45

15

25

63

84

84

95

95

95

101

25

45

25

96

35

95

50

114

80

113

100

133

100

202

25

34

25

54

25

107

25

107

50

153

50

126

50

126

50

126

50

126

C

C-3

225

22

45

110

310

54

90

135

450

76

100

171

550

105

100

236

(1)

Typical cycle

time is

from

10 to

100 seconds for injection

molding machines with typical

injection capacity of

4

to

100 ounces and typical tonnage

of

50 to 600 tons.

References:

Typical cycle times

-

Chemical

Engineering

Department, University of Connecticut

~w.enpr.uconn.edu/chea/polymer/inimoId.htni

Typical

injection capacity and tonnage

-

Rosato,

Rosato and Rosato.

Injection Molding

Handbook 2000; page 28.

3rd edition.

Boston,

Kluwer Academic

Publishers.

(2)

Max. Throughput (lb / hr)

=

Max. Shot Weight

(oz / cycle)

x

lb /16 oz

x

cycle / cycle time (sec)

x

3600 sec / hr

Maximum

Shot Weight

Model

Tonnage

(oz)

(3)

Equipment

Manufacturer

A

A

B

B

C

C

NOTES:

(3)

Injection molding machines outside

of the typical injection capacity and tonnage

ranges.

---

TECHNICAL PAPER

ISSN 1047-3289/.

Air&

WasteV

copyrgh,

1999 A~

&

Waste

Development of Emission

Factors

for Polypropylene

Processing

Ken

Adams

The

Society of the Plastics Industry,

Inc.,

Washington, District of Columbia

John

Bankston

Aristech

Chemical Corporation, Pittsburgh, Pennsylvania

Anthony Barlow

Quantum

Chemical Company, Cincinnati,

Ohio

Michael W. Holdren

Battelle, Columbus,

Ohio

Jeff

Meyer

AmocoPolymers,

Inc., Alpharetta,

Georgia

Vince

J. Marchesani

Montell North America, Inc., Wilmington,

Delaware

ABSTRACT

Emission factors for selected volatile organiccompounds

and particulate emissions

were developed during extru-

sion

of commercial grades of propylene homopolymers

and copolymers with

ethylene. A

small

commercial ex-

truder

was

used.

Polymer

melt

temperatures

ranged

from

400

to

605

°F.However,

temperatures in

excess

of

510

‘F

for

polypropylene

are considered

extreme.

Temperatures

as high

as

605

‘F are only used

for very

specialized

applications,

for

example,

melt-blown fi-

bers.

Therefore,

use

of this

data

should

be

matched

with

the resin manufacturers’

recommendations.

An emission factor wascalculated for each substance

measured and reported as pounds

released

to the atmo-

sphere

per

million

pounds

of polymer processed

ppm

(wt/wt).

Based on

production

volumes,

these

emission

factors

can

be used

by processors

to

estimate emission

quantities

from

polypropylene extrusion

operations

that

are similar to

the resins and the conditions

used

in

this study.

INTR0DuCrI0N

The Clean Air Act Amendments of 1990

(CAAA9O) man-

dated the reduction of various pollutants released to the

atmosphere.

Consequently, companies

are being

faced

with

the

task

of establishing

an

“emissions

inventory”

for thechemicals released or generated In their processes.

The chemicals targeted are those that either produce vola-

tile organic compounds (VOC5) and/or compounds that

are on the U.S.

Environmental Protection Agency’s (EPA)

list of

189 hazardous air pollutants

(MAPs). ‘flUe V of the

amended

Clean Air Act establishes a permit program

for

emission sources to ensure an eventual reduction in emis-

sions. When

applying

for a state operating permit, pro-

cessing companies are first required to establish abaseline

of their potential emissions.’

In

response to the needs of the plastics industry, the

Society of the Plastics Industry, Inc. (SPI)organized a study

to

determine the

emission

factors

for extrusion

of ho-

mopolymer and copolymer of polypropylene. Sponsored

by ten major resin producers, the study wasperformed at

Battelle, an independent

research laboratory.

This

work

follows

a previous SPI/Battelle study on

the emissions of

IMPLICATIONS

This

study

provides quantitative

emissions

data that were

collected during extrusion of honiopolyrners and Copoly-

mars of propylene.

These data are directly related

to pro-

duction

volumes and

can

be used as reference

points

to

estimate emissions

from similar polypropylene

resins ex-

iluded on similar equipment.

~k,me49

J.nuary 1999

Jown&

of

the

Ao

& i4~ste

Mansgrant

Association

41

---

Adams et S.

polyethylene’

and was

performed

In conjunction

with

emission studies

on ethylene-vinyl acetate and ethylene-

methyl acrylate copolymers.’

A review

of theliterature

reveals that thermo-oxida-

tion studies have been performed onpolypropylene.t5The

primary concernsaboutthese previous emissions data are

that they were generated using static, small-scale,’ or oth-

erwise

unspecified

procedures.’5

These procedures

may

not adequately simulate thetemperature and oxygen ex-

posure conditions typically encountered in the extrusion

process. That is, in most extruders, thepolymer melt con-

tinuously flowsthrough the system, limiting the residence

time in the heated zones. This contrastswith staticproce-

dures, in which the polymer may be exposed to the equiva-

lent temperature,

but

for

an

effectively

longer period of

time,

thus resulting In an

exaggerated thermal exposure.

In

a similar way, the concern

over oxygen

in

the indus-

trial extrusion processis minimized as theextruder screw

design

forces entrapped air back along the barrel during

the initial compression and melting process. The air exits

the system via the hopper; consequently, hot polymer is

only

briefly In contact with

oxygen when

it

Is extruded

throughthe die. Again, this

is in contrast to statictesting,

in which hot polymer may be exposed to air for extended

periods

of time.

In view

of these concerns,

the accuracy

of data obtained from

these

procedures may

be limited

when used

to predict emissions generated by polypropy-

lene processors.

As an alternative to

small-scale static technology,

a

better approach Is to measure emissions directly from the

extrusion

process.

Since the type and

quantity of emis-

sions areoften influenced

by

operational

param-

eters,

the

Ideal situation

Is

to

study

each

process

under the specificoperat-

ing

conditions

of con-

cern. Parameters that can

alter

the

nature

of

the

emissions

include

ex-

truder size andtype, melt

temperature and rate, the

air-exposed

surface

to

volume

ratio

of

the

extrudate,

the

cooling

rate of the extrudate, and

the shear effect from the

extruder

screw.

Other

variables

related

to

the

material(s)

being

ex-

truded can also influence

emissions. These include

resin

type,

age

of

the

resin,

additive package,

and

any

additional

materials

added to

the resin prior to extrusion.

If a processor uses

recycled materials,

the thermal history

is also an impor-

tant factor.

In view of these variables, a considerable

task would

be to

devise and conduct emission measurement studies

for all

major extrusion applications. Therefore,

SPI’s oh-

jectiveIn this work was to develop baseline emission fac-

tors

for polypropylene processing under conditions

that

would

provide

reasonable

reference

data for

processors

involved in similar extrusion operations.

The fiveresin

types evaluated were a reactor grade ho-

mopolymer, acontrolled rheology homopolymer with and

without antistat,

a random

copolymer,

and a reactor im-

pact copolyrner.

The samples used were

mixtures of com-

mercial

resins from

the sponsoring companies.

The test

matrix

used was designed to provide emissions data as a

function oftheir resin type andtypical melt temperature(s).

This information is provided in Table 1, together with the

average additive content

of the resin

mixtures. These

are

typical additives normally found in polypropylene.

A small commercial extruder was equippedwith a 1.5-

in. screw and fitted with

an eight-strand die.

The emis-

sions

were measured over a 30-mm.

period and were re-

lated to theweight of resin extruded. The emission factor

for each substance measured is reported as pounds evolved

to

the atmosphere

per million pounds

of polymer pro-

cessed IppmIwt/wtl

~.

Processors using similar equip-

ment

can

use

these

emission

factors

as

reference

points

to

assist

in

estimating emissions for their spe-

cific process.

laDle L

Polypropylene

emission lesI

suns; resin characteristics

additive concantralion and net

temperature.

Run No.

Sequence

ResIn lype

Melt Flow

Rate

(gtIO

mIst

s

230

‘C)

Number ci Resins

In

composite

Melt Temp (F)

Average AddItIve

Concentratlsi (ppm)

1

2

3

ConttolledRheology

Homopolymer

Non Antislat

30—35

6

400

Sill

605

Mtloxi~it

1,700

PA

1,000

4

Controlted

Aheology

tionopolynier

will Antislat

30—35

6

490

Mlioxidanl

1.71))

AS”

3.400

PA’

2.500

S

6

Reactor Grade

Homopolyrner

3—7

7

490

570

Antloxldait

1./00

PA* ~

7

Reactor Impact

Copolymer

15-20w1

EPR

3—10

4

505

Antioxidant

2,500

PA’

1.500

8

Random

Copolyiner

3—Set

Ethylene

3-7

3

510

Antioxidant 2,~

PA

2,200

Slip/AS

3.000

‘Process

aid

Antistat

50

Jou’nal

off,.

Air

&

Waste

Managrant

Associafion

V~jrne

49

Jenuery

1099

---

Adams eta!.

Thesubstances targeted for monitoring included par-

ticulate matter, VOCs, light hydrocarbons(ethane, ethyl-

ene, and propylene),

aldehydes (formaldehyde,

acrolein,

acetaldehyde,

and

propionaldehyde),

ketones

(acetone

and methyl ethyl ketone), and organic acids (formic, ace-

tic, and acrylic

acid). These

are the analytes

of interest,

either because they areon

the MAPs

list,

as stated earlier,

or

they are the expected thermal

and thermo-oxidative

breakdown products of the polymers tested.

EXPERIMENTAL

In the following section, briefdescriptions ofthe extruder,

the entrainment zone, and

sampling manifolds

are pro-

vided.

Details of the sampling methods, procedures, and

analytical instrumentation are provided elsewhere.2-’2

Experimental

Process Conditions

An 11PM Corporation

15-hp unvented extruder was used

to

process the

polypropylene test

sample mixtures

at

Battelle.

The extruder was equipped

with a 1.5-in,

single

screw

(LID

ratio of

30:1) and fitted

with

an

eight-strand

die (Figures 1 and 2). Extruded resin strands were allowed

to flow Into a stainless steel drum located directly under

the die head (Figure

2). Processing conditions, shown

in

Table 2, were selected to be representative of commercial

processing applications. The order of the polypropylene

emissions test

runs Is

listed in Table

1,

Capture and Collection of Emissions

Emissions

released

at the die

head

were

separately col-

lected

for

30

mm.

during the extrusIon runs

(Table

3).

Emissions from the hopper

were excluded

from analysis

since previous emission studies

showed their contribution to be

Insignificant

(less

than

2

of

the total).2

Table

3

shows

the

sampling

strategy

and overall

analytical schemeemployed for

the polypropyLene test runs.

Die Head Emissions

Emissions

released

at

the die

head

during

extrusion

were

captured at thepoint of release

in

a

continuous

flow of

clean

air.

A

portion of

this

air flow

was

subsequently

sampled

downstream is described in the

following

paragraphs.

The

emissions

were

Initially cap-

tured in a

stainless steel enclo-

sure surrounding

the die

head

(Figure

3).

The

air stream

was

Flours t

Extruler sfrand do

head used in

polypropylene en~ssions

testing program.

immediately drawn

through

a divergent

nozzle entrain-

ment cone, which provided a sheath of clean air between

the die

head emission

flow

and

the walls

of the carrier

duct. This minimized interactionof the hot exhaust with

the cooler duct walls.

The

total

air flow

employed for

capturing

die

head

emissions was

set at

700

Llmin.

This

was

composed

of

the die head entrainment flow at 525

L/mln, thesheath

flow at Llrnin,

and

75 Llmln of residual air flow that

was

madeup from room air drawing into the open bottom of

thestainless steeldie head enclosure. This residual air flow

was

used to

facilitate effective capture

of emissions

from

Figure

t

View

of the extnjder system and the vahous sampling

locations.

S

C

Olese

luesg

ofDl.n.nt

NonM

LPM

titruder

Bwel

(Ktln~ Zones

1,2.63)

Extiudats

CenteIns,

Purge

(20

LPN

to Vein)

Vokirne

49

Jenuary 1990

Joun’wJ

ofthe Air

&

Waste Management

AssocIation

51

---

Adams

et

at

TebI. 2.

REIn

throughput

and

key

110w

parameters during Ihe polypropylene extrusion runs.

TestRanNe.

1

2

3

4

5

6

1

8

Extnjler Condllletis

Restn

Tyop

Controlled

Controlled

Controlled

Controlled

Reactor

Reaclor

Reactor

Random

rheolo9y

rheology

Theology

rheology

grade

grade

impact

copolynier

homopolymer

hontpolyme

homopolymer

homopolyrcer

homopolymer

horr~polymer

copolyrner

(3—6 wt

UT)

(withailislat)

(15—20w1EPRI

Melt

Flow

Rate

MFR 30-35

MFR

30-35

MFR

30-35

MFR

30-35

MFR

3-7

MFR

3-7

MFR 3-10

MFR-3-7

Avera~

Die

Head

400

510

605

490

490

570

505

510

Mell Ten~

(F)

Zone

3

Temp (°F)

428

489

568

471

497

643

496

497

Zone

2

Temp (‘F)

403

430

469

320

369

436

369

369

Zone

1

Temp (‘F)

382

318

315

308

312

313

300

308

Pressure(psig)

50

50

50

50

750

250

400

2CC

Resin

Throughput

12.1/

9.29/

9.23/

7.58/

53.8/

41.9/

39.5/

23.6/

l0M,r)/(~iVmin)

91.6

70.3

69.8

57.4

407

317

299

179

RolorS~ed(rpm)

98

98

98

98

83

68

83

83

RunDur~Ion(rn~’)

30

30

30

30

30

3)

30

30

Flows

Tolal Manilold Flow (11mm)

700

700

700

700

700

7(X)

700

7(0

Flow Rate

Into

Sheath

100

100

100

100

100

100

1(0

100

(L/n~n)

Flow Rate

Into

Entrainment

525

525

525

525

525

525

525

525

(ljtnin)

FlowRateThrough

10

10

10

10

10

10

10

10

Hopper (ljmin)

FlowThraighrubestor

0.5

0.5

0.5

0.5

0.5

0.5

0.5

0.5

Camonyls

ft/mm)

Flowihroughlubestor

5

5

5

5

5

5

5

5

Adds

Organic

(1mm)

Flow Into

Canisters (iJminl

0.16

0.16

0.16

0.15

0.t6

0.16

0.16

0,16

FlowThroogh4O2THC

1

1

1

1

1

1

1

1

Anntyzer (L/min)

Flow Through

Filter

Holder

(lJmIn)15

IS

15

15

15

15

15

15

Table

3-Analytical scheme

for

polypropylene

test

runs.

Substances

Monilored

Organic Acids

Aldehydes/

Particulate

VOCs

Ketones

Heaty

Hydrocarbon

Light Hydrocarbon

Collection Media

OH

Impregnated Filter

IThPH

Tube

Glass Fiber Filter

SUMMA Canister

Analytlcil

Method

Desotplion

With

Dilute

l1,S0~

and

Analysis

by

Desorption With

Acetonitrile

an

Gravimetric

Modified T0-14

ton Exclusion

ChromotographyflJv

Analysis by HPLC

HP-i

Fused Silica

Capillary

Al,0fl~a2SO4

Column

Capillary

Column

GC/MS

j

GC/FID

GC/FID

Sampling Location

Manifold

Melt

Temp

(‘F)

Run

No.

Number

ot

Samples

Analyzed

400

1

2

2

1

1

2

1

510

2

2

2

1

1

2

1

605

3

2

2

1

1

2

i

490

4

2

2

1

1

2

1

490

5

2

2

1

1

2

1

570

6

2

2

1

1

2

1

505

7

2

2

1

1

2

1

52

aims

0/The Ar

& ~

Management

AssociatIon

Vokmie 49

Jefl4Jfl

1999

---

Adams et

al.

the

polymer. These

flows are depicted

in Fig-

ures

2 and

3. An orifice plateand control valve

connected to a magnahetic

gauge were

used to

set the flow at each

location. A calibrated

mass

flow meter was used before and after the test

runs to verify the settings. The flow setpoints

were within +/-3

of thestated values.

Die head emissions

were transported by

the 700-L/min air flow

to a sampling point

10

ft downstream

of the die head

using 4-inch-

diameter glass

tubing.

The location for this

sampling point

(Figure

2) was based on pre-

vious studies performed at Battelle that in-

volved

design,

engineering,

implementa-

tion,

and proof-of-principle

stages for the

pilot plant

system.2’2

Two separate

sampling manifolds

were

med at the sampling

location;

one for collect-

ing gases and

vapors

and theother for collect-

ing particulates

(Figure

4). For gases and va-

pors, a

10-Llmin substream was diverted from

the main emission entrainment stream using

a

0.5-inch stainless steel tube (0.425

inch

i.d.) wrapped

with

heating tape

and maintained

at 50°C.VOCs

and

oxygenates were sampled

fromthis manifold. Similarly,

particulates were sampled isokinetically

from a separate

15-L/min substream using a 0.25-inch stainless unheated

steel probe (0.1375 in. id.)

Two different methods were used

to measure VOC

emissions. One was the Beckman 402 Hydrocarbon Ana-

lyzer,

which continually analyzed the air emission stream

throughout the run

and provided a

direct reading of all

VOC substances responding to the flame ionization de-

tector. The other method

used an evacuated canisterfor

sample collection and gas

chromatography for analysis.

With this

method, total VOCs were determined by sum-

ming up the heavy hydrocarbon (containing a carbon

number ranging from

C3 through

C14) and light hydro-

carbon (containing

a carbon number

ranging from

C2

through C3) results.

The

total VOCs determined with

the 402

Analyzer

are in general agreement with the VOC values obtained

by

summing up the light and heavy hydrocarbons spe-

cies from the two GC methods. The 402

Analyzerresults

areconsistently

higher. The data obtained with

the GC

speciatlon

method

more closely resembles the TO-12

method, which

Is frequently used to measure source emis-

sions of VOCs. Information on

theTO-12 method and

the GC speciation method (TO-14) can be obtained from

the literature.9

This

study did not

include any measurements

of

emissions from the drum collection area,

as all

commer-

cial extrusion processes quench the molten

resin shortly

after

it exits the die.

Emissions from the extrudate in the

collection

drum were prevented

from entering the die

head entrainment area by drawing air from the drum at

20 L/min and venting to the exhaust duct. Several back-

ground

sampleswere taken, and smoke tubes were em-

ployed to confirm that the discharge from the entrain-

ment

area was not contributing

material

to

the sam-

pling

manifold.

VALIDATION OF THE

ANALYTICAL METHOD

The purpose of the manifold

spiking experiments wasto

determine thecollection and recovery efficiencies of the

canister,

acid,

and carbonyl collection methods. During

the first spiking experiment,

all three collection methods

were evaluated.2 During thesecond spiking experiment,

collection/recovery efficiencies were determined only

for

the canister sampling method. Theresults from the

two

spiking

experiments are summarized

in

Table

4. The

analytes measured by the spiking experiments are listed

in

column

one. Column

two shows

the method used.

Column three shows the calculated concentrations of the

spiked compounds

In the air

stream of the manifold. The

concentrations found from duplicate sampling and analy-

ses, corrected

for background levels,

are

shown in the next

two columns.

Finally, the average percent recovered

is

given in the last column.

The results

from thefirst experiment

are summarized

in Table 4 to show recoveries

of themanifold spiked com-

pounds. The threeorganic acids were spiked at a nominal

air concentration of about 0.6

to 0.8

pm/L.

Recoveries

using the 1(01-1-coated filters

ranged

from

107 to

122.

Ta~erI..v

70°

LPM

Figure

3.

View of emission entrairynent area.

V~umo

49

Jenua.y

1999

Journol

of

The

Afr

&

Waste Management Associahon

53

---

Adams et al.

Formaldehyde (1.63

jzmlL)

served

as the surrogate for the

aidehyde/ketone species, and the DNPH

cartridge method

showed a

recovery of

130.

Deuterated benzene (0.092

pmIL)

served

as the representative compound

for the can-

istercollection

method. The amount recovered was 9596.

During thesecond

experiment, additional

recovery

Table

5. This shows

the average die head

melt tempera-

ture

for each run

and provides emission values

in pg/g

for the target species

in the following categories: particu-

late matter, VOCs, and oxygenated species—aidehydes,

ketones,

and

organic acids. The concentrations are directly

translatable to pounds

of material generated per million

pounds of resin processed at that extrusion temperature.

Figure

5

shows a

bar

graph of the just-mentioned emis-

sion categories by test run.

Emissions plotted include par-

ticulate matter, VOCs

as measured

by the Beckman

402

Analyzer, VOCs as measured

by the gas chromatographic

speciation

methods

(e.g.,

light and heavy hydrocarbon

methods), and,

finally, the sum of the oxygenate species—

aldehydes,

ketones, and organicacids.

Examination of the five different resin mixtures ex-

trudedat a similar temperature (500 ‘F),

that is, Test Runs

2, 4, 5,

7, and 8 show the controlled rheologyhomopoly-

mer sampies

(2 and

4)

generate the highest concentra-

tion of particulates and

VOCs. Figure 5

clearly demon-

strates the effect of melt temperature (400

to 600 ‘F) on

emissions from

a single resin type. Test Runs

1,

2,

and 3

show,

as expected,

that emissions of all

species increase

with

increasingextrusion temperature; Test Runs

S and 6

show

similar behavior,

but to a lesser extent. Note that

these data may not

be extrapolated

to

the higher tem-

peratures used

for the meltspinning

process.

Individual organic acid

emissions ranged

from

less

than the detection level to

6.6

pg/g). Formic and acetic

acid concentration varied by factors of 20 and

15, respec-

tively, over the eight runs, but therelative levels of for-

mic and acetic acid were similar (within a factor of 2) from

test

run to

test

run.

Acrylic acid emissions, if any, were

below the detection limits of the equipment. Test Runs 3

data was obtained for the canister method

using an

expanded

list

of compounds. The

additional

compounds

included deuterated

benzene for comparison with

the first ex-

periment as

well as benzene, methyl

acry-

late, deuterated

methyl acrylate,

and vinyl

acetate.

The expected spike level of these

fivespecieswas nominally 0.24 pm/L. Mass

ions from the mass spectrometric detector

that were specific for each compound were

used in

calculating

recovery efficiencies,

since the fivespecies were

not well-resolved

with

theanalytical column

(i.e.,

the two

methyl acrylates

were

seen

as one

peak

when

monitoring the flame ionization detector).

POLYPROPYLENE EMISSION FACFOR

RESULTS

The extrusion

test

run results from theeight

polypropyleneresin mixtures

are shown in

Table 4.

Results iron spiking

experiments.

Anatyte

Method

Spike

Rscov

Level

(pgk)

Sell

sty (pgil.)

S.i2

Average

Recovered’

Formic

Acid

OH

tillers

First Expertmenf

0.71

0.987

0.733

122 ±18

Acetic

Acid

KOH

litters

0.77

1.023

0.640

121±12

Acrylic

Acid

KOH tillers

0.59

0.687

0.567

107±11

Formaldehyde

DNPH

Caxlddge

1.63

2.20

2.03

130

±

S

Oenzene-d5

Canister

0.092

0.~8

0,6

95

±

2

Oenzene-d,

Canister

Berurerue

Canisier

Second Experiment’

0.24

0.27

0.22

0.22

0,25

0.22

108±4

100

Methyl

Acr’yiale-d,

Canisler

Methyl

Acrylale

Canister

0.25

0.26

0.25

0.25

0.24

0.23

100

*

4

95

±

4

VinylActeate

Canister

0.24

0.26

0.25

110±6

‘Relative error Is

the relative per~nt

difference: the

absolute

difference in the two

samples

multiplied by

100

and

hen

divided

by

heir average.

‘4-

-‘It——

FIgure

4.

Sarnplng

manifolds

for emissions generated in

die head.

54

Journal

of

The

Air &

Waste Management Associatb,

Voknre 49

January1999

---

Adams eta!.

and

4

showed the

highest

levels of organic acids. The to.

tai

organic acid emission values for these runs

were 10.6

and 10.9 pg/g, respectively.

Figure 5 graphically shows

the total

oxygenates detected.

Even

at the highest

melt

temperatures employed in this study, the oxygenates con-

tributed less than

11

of the total VOCs emitted.

The individual

carbonyl species ranged in

emission

values from less than the detection level to 26.9 pg/g. All

eight species were resolved. Acetone was the most pre-

dominant component,

followed by formaldehyde and ac-

etaldehyde. Test Runs 3,

4, and

6 showed thehighest

level

of total

carbonyl species. The total carbonyl content from

these

runs

were

73.8,

14.9, and 21.8 pg/g, respectively.

Note that the

EPA

Is proposing to revise its definition

of VOCs for purposes of

preparing state implementation

plans (SIPs) to attainthe nationalambient air quality stan-

dards

(NAAQS) for ozone under Title

I of the CAAA9O

and for the federal implementation plan

for the Chicago

ozone nonattainment area. The proposed revision would

add acetone to the list of compounds

excluded fromthe

T6b1

LSumnary ol

polypropylene extrusion emissions br generic resin grades (mg/g).

definition of VOC on the basis that these compounds have

negligible contribution to

troposphericozone formation.’0

The significance

of this data becomes apparent when

placed in

the context of the 1990

CAAA9O definition of

“major” source for VOC emissions. Categorization of an

emission

source as a majorsource

subjects it to more strin-

gent permitting requirements.

The definition of a major

source varies with the severity of the ozone nonattainment

situation of the area where the source is located. Thecur-

rent VOC emission limits are

10 tons/yr for an emission

source within an extreme ozone nonattainment classifi-

cation, 25

tons/yr

for a source in the severeclassification,

and

50 tons/yr

for

a source

in the serious classification.

Currently, the only extreme nonattainment area in the

United

States is the Los Angeles area.

The

utility of this data can be

illustrated

in the fol-

lowing examples.

Based on

theemissions data developed

in this effort, aprocessor with equipment similar to that

used

in this study can

extrude

annually up to 24.4 mil-

lion

pounds

of controlled rheology polypropylene at a

TestRunl4o.

1

2

3

4

5

a

i

a

Extrieder ConditIons

Resin

lype

Controlled

Controlled

Controlled

Controlled

Reactor

Reactor

Reactor

Random

rheology

rheology

rheology

rheology

grade

grade

impact

copolynier

homopolymer

homopolynier

homopolynier

homopolymer

(wlIh

anlistat)

homopobynier

homopolynoer

copolymer

(15—20w1

EPR)

(3-8 v4

El)

Melt

Average

Die

400

510

605

490

400

570

505

510

Melt Temp

(‘F)

Particulate Matter

30.3

68.4

653

150

17.3

218

34.5

27.9

VOCs

Beclerian

402-

THC4

104

177

819

191

33.4

202

80.3

59.4

Hea~

Hydrocarbons

79.1

175

587

104

24.6

127

65.1

29.8

Light Hydrocarbons

En~ne

0.90

1.39

4,55

0.78

0.07

0.37

0.02

0.08

Ethylene

0.38

1.44

1.36

0.50

0.03

0.05

0.02

0.05

Propylene

0.21

0.80

13.9

0.70

0.12

2.24

0.06

0.26

Aldehydes

Formaldehyde1

0.74

1.38

19.1

1.30

0.17

7.05

0.18

0.09

Acrolelp’

0.01

0.05

0.81

0.14

0.01

0.10

0.01

0.01

Acetaldehydee

0.46

0.54

15.8

0.53

0.09

5.63

0.20

0.08

Proplonaldehyd?

0.05

0.07

1.60

3.31

0.02

0.97

0.95

0.02

Bu~raidehyde

0.78

1.05

3.32

0.92

0.04

0.38

0.08

0.01

Benzalthhyde

0.12

014

5.21

0.51

0.08

0.88

0.02

0.08

Ketonez

Acetone

9.66

12.6

26.9

9.36

0.15

2.82

0.31

0,18

Methyl

EThyl Kelone1

0.19

0.24

9.62

0.26

0.07

5.23

0.04

0.04

Organic

Acids

Fo,micAcid

0.69

1,43

3.98

5.98

0.2

1.19

0.2

0.31

Acetic Acid

1.10

1.25

6.60

4.90

0.2

2.64

0.25

0.52

AcrylicAcid

0.08

0.08

0.08

0.08

0.08

0.08

0.08

0.08

‘THC

—

Total hydrocarbons

(methane

is not included).tHazardot~air pollutants (FlAPs).

Note:

The emission

vaiuesare averages Iron duplicate

runs. In

general, the

ditterences werec+/-15.

Volume 49 January1999

Journalof the Air &

Waste

Management Assocaltbn

66

---

Adams

et

al.

Ft9UTO 5.

Bar

graph

showing the particulabes, VOCs

obtained with

the 402 ~Analyzer,

VOCs obtained

by

GC

speciatien

and oxygenated

orgent

species

(tactors

tt

jg/g).

melt temperature of 600 °For

1,156 millIon

pounds of

reactor grade homo polypropyleneat a melt temperature

of 500 ‘F without

exceeding the

10-ton/yr limit for

an

extremeozone nonattainment area.

CONCLUSIONS

Based upon the results

of this

study, the following six

conclusions are made:

(1)

For the resins studied, themajor emission com-

ponents were particulate matterand VOCs. Much

tower amounts were found of the oxygenated

species—aidehydes,

ketones, and organic acids.

(2)

Emission

rates are directly correiatable with

the

melt temperature.

(3)

Although the

collection and MS

speclatton of

VOCs most closely follows the

EPA procedures

(TO-12and TO-14) for

measuring

VOCs,

the more

conservative approach

using the Beckman 402

Analyzer,

which yields higher

VOCs values,

should

be employed.

(4)

The data providespolypropylene processorswith

a baseline for estimating the VOCs generated by

the resins they handleon

a daily basis under pro-

cessing conditions similar to

those used

in this

study and at the maximum meit temperatures re-

ported.

Thefollowing weightsof each

resin can

be

processed without exceeding the 10-ton limit

of an “extreme” ozone nonattainment area:

24.4

million pounds of controlled rheologypolypropy-

lene at 600 ‘F, 99.0 million

pounds of reactor grade

homopotymer at 570

‘F, 249.1

million pounds of re-

actor impact copolymer at 505 ‘F,

and 336.7 ml-

lion pounds of random copolymer at 510 ‘F.

(5)

In some cases, theemission

factors determined

in this

study may overestimate”

or under esti-

mate emissions from

a particular process. Profes-

sional judgement and conservative measures

must be exercised as necessary when using the

data for estimating

emission quantities.

(6)

This study was not designed to

meet the needs

of industrial

hygienists.

However,

this type of

apparatus can be used at different extrusion con-

ditions to gather dataon other types of extrudates

such

as fiber, film,

or sheet.

RLFERENCES

1.

Sherman, L.M.

‘clean-air

rules challenge

processors,’

Plasiicr Technol.

1995,41,2.83-86.

2.

Barlow.

A.;

conoos,

D.;

Hoidren,

M.; Ganison, P.;

Harris, L;janke, B.

‘Developmentof emission

factors for polyethylene processing.’

1.

Air

&

Waste Manage.

Assx.

1996, 46,

569-580.

3,

Battelle Final

Report

to the Society of the

Plastics Industry, ‘Sam-

pling

and

analysis ofemissions evolved

during thermal processing

0!

ethylene-vinyl

acetate

and

ethylene

methyl acrylate resin

cornpos-

Ites,’ March 1995.

4.

Hoff, A.; jacobson, 5. “Thermal oxidation

of

polypropylene close

to

industrial

processing

condleions,’J.

AppliedIbl5en. Sd.

1982,27,2,539.

5.

Patet,

s.H.;

Xanthos, M. ‘Volatile

emissions

during thermopaastics

processtng—A Review,’

Advances in Poi~an.T&,nof.

5995,

14. 67.

6.

Hoff, A.;jacobsson,

S.

‘Thermo-ox,idative degradation

of

tow density

polyethylene close to Industrial processing condition,,’

1.

Applied

Poi~ss.

Sc!. 1981, 26, 3,409-3,423.

7.

Hughes,

lw.;

bland, R.F.; Runaldi, G.M.

Source

Assessment:

Plastic

Processing.

State of the

Art. Monsanto

Research Corporation.

March

1978, EPA-600/2.78/004C, 27.28.

B.

“Air

facility subsystem

source classification

codes

and

emission fac-

tor

listing for

criteria

air potlutants,”

EPA. March

1990,

EPA

450/4’

90/003.

9.

compendium ofMethods

For The Determination

of Toxic Organic

Compounds in

Ambient Air,

1.1.5. Environmental Protection

Agency,

june

1988.

Available from NTIS

as PB90-127374.

10.

Fed.Reglst.

1994, 59, 189, 49,877.

11.

Forrest, Mj.;jolly,

AM.;

Holding,

SR.; Richards, SJ.

‘Emissionshorn

processIng

therinoplastics,’Annals

ofOccupationall~bne.

1995,39,

1, 35-53.

12.

Battelle Final Report

to the society of rhe Plastics Industry, ‘Sam-

pling and

analysis of

emissions during thermal

processing of

polypro

-

pylene

resin mixtures.’

April 1995.

f2

,,‘...,..~.

a’~A~’.’5

a’—c,~,’

~

~LM

AbOUt

ths Authors

Anthony

Barlow, Ph.D.,

is

a

product steward for Quantum

ChemIcal

Company.

John

Bankston

is supervisor

of

product

regtiatien at

Aristech Chemical Corporation.

Michael Holdren

Is

a

senior research

soientlst

at

Battelle

Memorial

Institute.

Vince

Marchesani, Ph.D.. Is dIrector

of

health and environ-

mental affairs for Montell North America. JeifTey Meyer,

Ph.D..

is manager

of polymer physics

and

testing at Amoco Poly-

mers,

Inc.

Ken

Adams

(corresponding

author) is assistant

technical directorfor the Society of the

Plastics Industry, Inc..

1801

K

St., NW,

SuIte 600K, WashIngton, DC 20006.

55

Jo,xn&

0/

The

Air & ~

Management Associalion

VoIsra4g

Jenuy

1999

---

TECHNICAL PAPER

lSSN

1047.3289/.

Air &

W

cooyvighl

2002 Al & Waste Mar

Development of Emission

Factors for Polycarbonate Processing

Verne 1. Rhodes

Product Regulatory Services, ma, Cull Breeze, Florida

George Krlek and

Nelson Lazear

Bayer Corporation,

Pittsburgh, Pennsylvania

Jean Kasakevjch

The

Dow Chemical Company,

Midland, Michigan

Marie Martinko

The

Society of the Plastics Industry

Inc., Washington, DC

R.P. Heggs,

M.W.

Holdren, A.S. Wisbith,

GW.

Keigley, J.D. Williams, J.C. Chuang,

and J.R.

Satola

Battelle,

Columbus, Ohio

ABSTRA~

Emission factors for selected volatile organic compounds

(VOC5) and particulate emissions

were developed

while

processing eight commercial grades ofpolycarbonate(PC)

and

one grade of

a

PC/acrylonitrile-butadiene-styrene

(ABS) blend.

A small commercial-type extruder was used,

andtheextrusion temperaturewas held constant at 304 ‘C.

An

emission

factor

was

calculated

for each substance

measured and is reported as pounds released to the atmo-

sphere/million

pounds of

poiyrner resin processed

ppm

(wt/wt).

Scaled to

production volumes, these

emission

factors

can

be

used

by

processors to

estimate emission

quantities from similar PC processing operations.

INTRODUCTION

The

Clean

Air Act Amendments

of

1990

(CAAA)

man-

dated thereduction of various pollutants

released to the

atmosphere.

Asa result, companies arefaced with the task

of establishing

an “emissions

inventory”

for the chemi-

calsgenerated and released by their production processes.

The chemicals targeted are those considered volatile or-

ganic compounds (VOC5) and those that are on the U.S.

Environmental

Protection

Agency’s

(EPA)

current

list

of

188 hazardous

air pollutants.

Titie

V

of the

CAAA

establishes

a

permit

program

for

emission

sources

to

ensure

an eventual

reduction in

these

chemical emis-

sions. When applying for a state operating permit, pro-

cessing companies

are required to

establish

a baseline

of their potentiai

emissions.’

in response to the needs of the plastics industry, the

Society of thePlastics Industry, Inc. (SN) organized astudy

to determine the emission factors

for extruding polycar-

bonate

(PC)

homopolymers,

copoiymers,

and blends.

Sponsored

by

two major resin producers, the

study was

performed

at

Battelle.

This

work

follows

previous SPl/

Battelte

studies

on

the

emissions

from

acrylonitrile-

butadiene-styrene

(ABS),3

polyethylene,3

ethylene-vinyl

acryiate

and ethylene-methyl

acrylate

copolymers,

poiypropylene,5and polyamide.6

Thereare limited literature references about emissions

from PC,

but most of these use static, small-scale proce-

dures and were intended to predictemissions from either

a

fire scenario or worker exposure.” These procedures do

not accurately simulate thetemperature profile and oxy-

gen exposure conditions typical of extrusion processing.

Static testing

usuaiiy exposes the resin

to

temperatures

outside

(both greater than and

less than)

typical

extru-

sion

temperature ranges and to

atmospheric oxygen

for

extended periods of time. During commercial processing,

the resin

is

molten

for

a few

minutes

at most, and the

equipment is designed to force air out of contact with the

melt

in

the barrel.

Hot resin

is

in

contact with

oxygen

IMPliCATIONS

This study provides quantitat’r/e

emission data

collected

while

processing nine types

of

PC-based

resins.

These

data are directlyrelated to production throughput arid can

be

used as reference points to estimate

emIssIons from

simiiar PC resins

processed

on similar equipment.

Vckwrie

52

Jiiy

2002

~ta-nafotfheAk

& Waste Menagama,t Association

751

---

Rhodes et aL

only briefly as it

exits the die. In light of these differences,

the data obtained from static tests

are of limited

use in

predicting emissions from commercial processing.

Greater accuracy would, of course,

be possible by

measuring

emissions

from

actual

production

equip-

ment.

Because operating parameters can influence the

type and quantity

of

emissions,

the greatest

accuracy

can

be achieved by

studying

each

process.

Parameters

that can influence emissions include extruder/injection

molder size and type, melt

temperature, processing rate,

the ratio

of air-exposed

surface

to

the

volume

of the

product, and sheareffects caused by screwdesign. Vari-

ables associated with

the material being processed that

can also affect emissions include

resin type, age

of the

resin,

additive

packages,

and

heat

history

of any

re-

cycled resin, It would be a daunting

task to design and

Implement emission

studies

for

all

combinations

of

processing variables.

To strikeabalance between the Inapplicability of static

tests

and thecomplexityofmeasuring each

process,

SPI and

majorPCproducers initiated work todevelop baseline emis-

sion factors for PC processing under conditions that would

provide reasonable reference datafor similar processing op-

erations. Extrusion was chosen as the preferredprocess be-

cause of Its

continuous

nature and

the ability

to

reach

steady-state conditions for accurate measurement Extrusion

is

also

believed

to

have higher emission

rates than

other

processes, such asInjection molding operations,9 and, there-

fore, should lead to more conservative extrapoiations.

For the current study, threecomposites and six single

resins were evaluated

(see Table 1). The composites were

abiend of Bayer Makrolon and Dow Calibre Intended for

food

contact,

compact

discs,

and

UV-stabiiized product

markets.

Bayer then

tested three grades of Makrolon In-

tended for radiation-stabilized, impact-modified,

and ig-

nition-resistant markets. Dow testeda radiation-stabitized

grade,

a branched PC, and a PC/ABS blend.

‘ThbIe 1. Test

runs

for

PC

resins program.

Sampling and

analytical measurements were con-

ducted to determine emission factors

for the following:

•

total

particulate matter;

•

total VOCs;

•

eight

targeted

VOCs:

methyimethacrylate,

monochlorobenzene,

carbon

tetrachloride,

me-

thylene

chloride, p/rn-xylene,

styrene,

o-xylene,

and toluene; and

•

four targeted semi-volatile

organic compounds

(SVOC5): diphenylcarbonate, bisphenol

A, phe-

nol, and p-cumyi phenol.

The targetedorganicspecieswere chosenbased on their

known or expected presence as thermal andthermal oxida-

tive breakdown products of the polymers selected for study.

EXPERIMENTAL

Resin Blending Procedure

For runs

1—3, equal

portions of each contributed resin were

homogeneously mixed

in

10-gal

metal

cans

to

form

a

composite

blend immediately

before the test

run.

Each

container

was

filled

to

approximately two-thirds

of ca-

pacity

and then

thoroughly

blended

by rotation on

an

automated

can-rolling

device.

Each resin

(runs

4—9) or

resin

mixture (runs

1—3) was placed

in

a drying hopper

and dried at 126.7°Cfor 6 hr to a dew point of —28.9 °C.

Extruder Operating

Procedures

The HPM Corp.

1.5-in., singLe-screw,

30:1 L/D (length-to-

diameter

ratio),

15-hp

plastic extruder was thoroughly

cleaned before the

PC

experiments.

The extruder is ca-

pable of -27.2 kg/hrthroughput and 426.7°C(maximum)

barrel temperatures

for the three heat zones.

A specially

constructed

screw used

on

a previous polyamide studr

was

used

and

is

shown in

Figure

1.

An eight-strand die

head used in previous SPI-sponsored emission studies was

used for this study and is shown in FIgure 2. Thedie head

was cleaned and

Inspected, the holes were

reamed

to

a

3/16-in, diameter, andthe surface

was

polished

before

the start

of

experimental work.

Each PC resin or mixture was

initially extruded for 10—20

mm

before the actual test run to en-

sure

stable

process

conditions.

During this time, the total

VOCs

were

monitored

by

online

instrumentation

to

indicate

equilibration of the exhaust ef-

fluent.

A check of operating pa-

rameters was recorded

initiaHy

and

at

5-mm

Intervals

during

each

20-mm

test

run.

These

parameters

included

lan

Na.

Resin

Sample lescrlptlsn

Appllcatluss

layer

MMROtON

law

CAUUE

Eitudlng

TmuIp.tattlre

1

Compcsile’

Food contact

3108

201

304°C

2

Composite’

Compact

discs

MAS-140 and CD2005

XLI 73109,OIL

304°C

3

Composite’

uv

stabilized

3103

302

304°C

4

Single

Radiation

stabilized

RX-2530

304°C

5

Single

Impact rn~dtfled

T-7855

304°C

B

Single

Flame

retarded

6465

304°C

7

Single

Radiation

stabilized

2081

304°C

B

Single

Branched

603-3

304°C

9

Single

PC/ABS bleid

Putse

830

304°C

‘Equal

weights of

resins

dry blended.

782

~iont

of

the Air &

Waste Management

Association

~,lume

52 July

2002

---

Rhodes

etaL

Plgurs

1.

Screw profie

fl-1PM Corporation).

Plgure t

Extrude,

strand

die head

used

in

poeyamide

emissIons

testing program.

•

check that the temperature at the die head had

reached target and was stable;

•

check that the RPM setting was at 60

(60

RPM);

•

check of the extruder cooling water flow (inand

out);

•

check of manIfold airflow rates; and

•

check of the

flow

settings

for

all

sampling

equipment.

For

each

test

run,

a

second

repetitive run

was

carried

out

immediately aftercompletion of the first runusing thesame

operating conditions. Duplicateruns were conducted to al-

low better assessment of sampiing andanalytical precision.

Die Head

Emission Collection

The stainless-steel emission-sampling manifold is shown

in Figure

3. EmIssions were entrained

In pre-condmtloned

air

(i.e., purified

through

a

charcoal filter).

Incoming fil-

tered air was preset at a flow of

400 L/mmn using thevari-

able flow blower and were maintained at this rate for all

test

runs. Thisflow was directed through the laminar flow

head

assembly and

across

the extrusion

die

head.

The

variable flow blower on

the receiving side

of the mani-

foid

system

was adjusted to

match

the 400-L/min inlet

flow.

Additional

flow

from the sampling equipment

re-

sulted In

—10

greater flow into the receiving end of the

sampling

manifold. Smoke

tubes were

used during the

test runs to confirm efficient transfer of the emissions.

The manifold was equipped with

multiple

ports

for

connecting the various sampling devices. Each port was

0.25-in. o.d. andprotruded

1

in.

into the airstream. The

*19t—OI—t059L

SCREW PROFILE

CUSTOMER

BATTELLE

MEMORIAl.

INSTITUTE

cowu

BUS.

OR

SHANK

FEED

TRANSITION

PUMP

TORPEDO

120

slzr

1.5

L/b

30:1

MATERIAL

TO

BE

PROCESSED

NYtON

6/6

c/.,

06—Dolt

MO~tS:43401IR

w/cooAOwov

86

fLIGHTS

ORDER

ND_.M~i~J.

CHROME

PLATED

FULL

LENGTh

COOLING

HOLE

~ume52

July2002

~rna/

oftheA~

&

Waste

ManagementAssOciation

783

---

Rhodes et al.

sembly and an in-line stainless steel probe (0.25-In. o.d.)

connected to a 47-mm filter pack.

Sampling

and Analysis

Methods

Themethodsemployed for chaiacterlzingthe emissions from

the resin

extrusion

process are summarized in

Table

2.

Detailed information is provided in the following sections.

Target VOfl.

The collection and analysis of target VOCs

followed EPA Method TO-14A guIdelines. Evacuated and

polished

SUMMA

6-L canisters

(100 mtorr)

were used

to

collect whole air samples.

The 6-L canisters were ini-

tially cleaned by

placing them in a 50°Coven and using

a

five-step sequence of

evacuating to

less than

1 torr

(1 mm of mercury vacuum) and filling to

—4 psig (lb/in.2

gauge) using humidified

ultra-zero air. A

final canIster

vacuum

of

100

mtorr was

achieved with

an

oil-free

mechanical pump. Each canister was connected to the

sampling manifold, and a 20-mm

integratedsample was

obtained during the collection period. Aftercollection,

the canister pressure wasrecorded, andthe canister was

filled to 5.0

psig with ultra-zero air to facilitate repeated

analyses of air from the canister.

Table

2.

Sanple collection and

analysis

methods

for polycarbonale lest wirs.

Substancus

CellectIes

MedIa

Matytlcil MeUied

Moaltered

Total

VOCs

Real-time

monitoring

Continuous

FID

Target SVOCs

XAD-2 adsorbent

CC/MS

Particulate matter

Glass fiber tiller

Gravinelric

weighing

Target VOCs

SUMMA canister

GC/panllel FID

and

MSD

A Fisons MD 800

gas chromatographic (GC)

system

equipped

with

parallel flame

Ionization

detectors

(FID)

and mass spectrometricdetectors (MSD) was used to ana-

lyze the target VOCs present in the canistersamples.

The

GC

contained a cryogenic

preconcentration

trap.

The

trap was a 1/8-

x

8-in, coiled

stainless

steel tube packed

wIth 60/80 mesh glass beads. The

trap was maintained

at —185 °Cduring sample collection and at 150°Cdur-

Ing sample desorption.

A six-port valve was used to con-

trol

sample

collection

and

Injection.

Analytes

were

chromatographically

resolved

on

a

Restek

Rtx-1,

60

m

x 0.5

mm

id.

fused silica capillary

column

(1

gm film

thickness). Optimal

analytical results

were

achieved by

temperature-programming

the GC oven from —50 to 220

°Cat 8 °C/mm. The column exit flow was spilt

to direct

one-third of theflow to the MSDand the remaining flow

to the FID.

The mass spectrometer (MS) was operated

in

thetotal ionization mode so that all masses were scanned

between 30 and 300

amu at

a

rate of 1 scan/0.4 sec. Iden-

tification of VOCs was performed by matching the mass

spectra acquired

from the samples to

the mass spectral

library from the NationalInstitute of Standards andTech-

nology

(141ST). The sample volume was 60cm’. With this

sample

volume,

the

FID detection level was 1.0

ppb.

Detector calibration was

based on

instrument response

to

known concentrations

of dilute

calibration gas con-

taining

the target VOCs

(traceable

to

NIST

calibration

cylinders).

The

calibration

range extended

from

0.1

to

1000 Rg/L.

Target SVOCs.

XAD-2adsorbent tubeswere used to collect

SVOC emissions. Analyseswere carriedout using a GC/MS

system. The adsorbent cleaning, sampling, and analytical

procedures

are described In the next paragraphs.

FIgure 3.

Emission enclosure

apparatus.

manifold was also equipped with a 4-In, filter holder as-

764

Jou-nalof theAir &

Waste ManagementAssociation

~h,me

52

July 2002

---

Rhodes et

al.

Thesampling module consisted of an inletjet equipped

with

a

quartz

fiber

filter

(Pallflex)

and a

glass

cartridge

packed

with precleaned XAD-2 (Supelco).

Thefilters were

purged in an oven (450 °C)overnight before

use.

The

XAD-2

cartridge assembly was sealed

at both ends, wrapped with

aluminum foil, and labeled with a sample

code.

Single XAD cartridge sampling was conductedover a

20-mm

collection

period using nominal

flow rates of

4

L/mln.

An

51CC

sampling

pump

was used

to

draw the

sample into

the cartridge assembly.

A

mass

flow

meter

(0—5 L/min) was used during the sampling period to mea-

sure actual flow rate. Aftersampling, the XAD-2 assembly

wascapped and stored in

a refrigesator. For runs

1A, 2A,

and SB, aknown amount of bisphenol-A (deuterated, d6)

wasspiked onto the XAD-2 cartridge lust beforesampling.

The

filter/XAD-2 samples

from

each

run

were

ex-

tracted

separately

with

dichloromethane for

16

hr. The

extracts were concentrated by evaporation with a ICuderna-

Danish

(K-D) apparatus

to a final volume

of

10 mL. The

concentrated extracts were analyzed by GC/MS to deter-

mine

SVOC concentrations.

A Hewlett Packard Model 5973 GCIMS, operated in

theelectron impact mode, was used. Sample extracts were

analyzed by

GC/MS in the full mass scan mode to deter-

mine

SVOC levels. A fused silicacapillary

DB-5 column,

60 m

x 0.32mm i.d.

(0.25

sm film thickness), was used

for analyte resolution. The initial CC oven temperature

was 70°C.After2 mm, thetemperature was programmed

Table 3.

Total

manifold exhaust flow

and resin

throughput

rates for generic PC resin

grades.

to 150°Cat

15 °C/minand then to 290°Cat 6 °C/min.

Helium

was used

as the carrier gas.

The

MS was

set to

scan from m/z

35 to

500 amu at

3 scans/sec. Identifica-

tion of the target analytewas based on a comparison of

mass

spectra

and

retention

times

relative

to

the

cor-

responding

Internal

standards

(naphthaiene-d,

and

phenanthrene-d10,). Tentativeidentification of nontarget

compounds was accomplished by manual interpretation

of background-corrected spectra together with an online

library search.

Total Particulate Material.

The concentration of particu-

lateemissions was determined by passing a sample of the

exhaust effluent through a

pre-weighed filter and then

conducting a gravimetric analysis

of the sampled filter.

The pre-weighed

filter (8

x

10

in.)

and holder were in-

serted into

the exhaust port of the sampling

manifold.

The sample volume was determined from a calibrated orI-

fice and Magnehelic gauge located on

the sample mani-

fold blower. A flow rate of 200 L/mmn wasused during the

20-mm

test runs. Gravimetric analyses of the filter before

and after sampling

were carried out in a controlled envI-

ronmental

facility (temperature

21 ±

1

°C,relative hu-

midity

50 ±

5).

The filters were preconditioned to

the

controlled

environment

for 24

hr and then weighed.

Total

VOCs.

A

VIG Industries

Model

20 total

hydrocar-

bon analyzer equippedwith a hydrogen flame Ionization

Tnt

Reala

OrifIce

hewer

Slower @

Total

XMl-i

CalMer

Total

Real.

Rn

Ifle

(lathes at

@140°F

75°F

or

VOC

Sampler

Sampler

Macillold

T)vnghput

NI.

water)

or

SO °C

(tJmln)

24°C

liMbs)

Aeialper

((1.1.)

(I/wIn)

((1mm)

Flow

(liSa)

(/mln)

1A

Food

contacl

4

417

393

2

4.0

0.2

399,2

354

lB

4

417

393

2

4.0

0.2

399.2

333

2A

Corn~actdiss

4

417

393

2

4.0

0.2

399.2

370

26

4

417

393

2

4,0

0.2

399.2

368

3A

liv stabilized

4

417

393

2

3.9

0.2

399.1

341

3B

4

417

393

2

3.9

0.2

399.1

322

4A

Radiation stabilized

4

417

393

2

4.0

0.2

399.2

356

48

4

417

393

2

3.9

0.2

399.1

359

5A

Impact

modified

4

417

393

2

3.9

0.2

399.1

309

58

4

417

393

2

3.9

0.2

399.1

310

6A

Ignition

resistant

4

417

393

2

3.9

0.2

399.1

344

68

4

417

393

2

3.9

0.2

399.1

351

7A

Radiation

stabilized

4

417

393

2

4.0

0.2

399.2

348

76

4

417

393

2

4.0

0.2

399.2

346

8A

Branched

4

417

393

2

4.0

0.2

399.2

325

88

4

417

393

2

4.0

0.2

399.2

323

9A

PC/ABS blend

4

417

393

2

4.0

0.2

399.2

285

96

4

417

393

2

4.0

0.2

399.2

287

Vokjme52

July2002

Journal of the Afr &

Waste ManagementAssocIation

788

---

Rhodes et at

detector

(HFID) wasused to continuously

I

monitor the VOC content of theexhaust

effluent. A heated sample line (149 ‘C) was

I

connected to

the extruder sample mani-

fold, andthe sample flow wasmaintained

at 2 L/mln. Theanalyzer was calibrated at

the

beginning

of

each test

day against a

•

NIST-traceable reference cylinder contain-

I

;

ing amixture of propane in 42-pxg/L ultra-

I

zero

air

(minimal

total

hydrocarbons,

5

water, C02, CO. or other impurities). Un-

J

~

earity

was demonstrated

by

challenging

the analyzer calibration standards of 3,46,280,

and 4480 mxg/L of methane.

Total

Manifold How

4

The total

manifold exhaust flow for the

Individual test runs

was

needed

for the

eventual calculation of emission factors.Table

3

lists the total

flows for each test

run. The orifice

a?

value is theobserved

reading

for

each run.

From the experi-

k

4

mentally

derived

regression equation,

—

flow

=

74223(AP)

÷

119.77 (R2=

0.9943),

I

~

a flow rate (typically expressed as L/min)

through the blower can be determined

using

this

a?

value.

However, the flow

I

across

the orifice

was

originally call-

j

I

brated at

75 °F(218

°C).The Rankine

temperature (°R)Is commonly employed

—

(‘R

=

‘P

+

45967). To correct the flow to

the

manifold operatIng temperature of

140°F(60‘C), thefollowing flow orifice

equation was used:

1/2

Q2

=

QI

(1)

where

Q,

was the

flow rate during

test

runs,

Q2

was the flow

rate at

75

‘F (535

°R),

T,

was the temperature

of

the ex-

j

haust air

eR),

and

T2

was

the tempera-

:

ture

at calibration (535 °R).

A temperature correction

factor of

0544

was

applied

to the flow

rate

dur-

~

ing the test runs

to

determIne the

flow

~

rate

at 75

‘F. In

addition,

the flow

rates

.~

from the individual

sampling

compo-

nents

were

needed

to

obtain

a

total

!‘

manifold

flow. The total

manifold

flow

is shown in the last row of Table 3. For

all test runs, the total manifold flow was

~‘

balancedat thepreset Incomingflow rate

~

of 400

L/min.

9’

01

01

~

tO

N-

01

0)

0)N-tflN~~~flr-Jc-J

•

.~‘N

.

.R~Ln~~1

~o,°0

0

0

0

0

0

0’’~

0

0

0

0

0

0

0

0

0

0

a

a

a

o

0

0

0

d

0

0

0

0

0

0

0

0

‘-a

a

~-

(N

(fl~

——

—

—

(N

‘-0000

ii

jfl

S

1.!

1••

—fl

785

Journal of the Air&

Waste ManagementAssociation

\k~&,iie52July2002

---

Rhodes et at

C

H

H

I

a

I

a

Emission

Factors

Amounts

of

the target chemicals detected

in the manifold exhaust flow are shown

in Table 4 (jsg/L). Emission factors for the

amount of target chemicals detected for

each resin tested (~ig/G)

were calculated

from the

measured emission

levels

in

Table 4 using this

formula:

E=(Cxfl/O

(2)

where

E

was

iig emisslons/g processed

resin,

C was the measured

concentration

of emissions in

~sgIL,

F

wasthetotal mani-

fold flow rate In L/min, and 0 wastheresin

throughput

in

g/min.

Emission

factors

(pg/G)are summarized in TableS. Dimen-

sional analysis shows that these emission

factors can also be read as lb emissions/

million

lb resin processed.

Significance

of Emission Factors

This

study

provides

emission

data col-

lected during extrusion of various PCres-

ins

under specific

operating

conditions.

The

calculated emission

factors

can

be

used by processors to determine their ex-

pected annual emissions, whIch

areused

to

categorize

industrial

sites

under

the

1990

CAAA. The most

stringent current

limitation is

10 tlyear

of

VOC

emissions

within

an extreme 03 management area.

A

processor with

equipment

similar

to

that

used

in

this

study could

extrude 100-

800

million

lb/year

of PC,

depending

upon the product mix, before achieving

maximum

permit

levels.

In less restricted

areas,

where

the

VOC

emissions

can

be

up

to

50

t/year, the processor could po-

tentially process5 times this amount.

RESULTS

The

primary

results

of the

study

are

shown

In

Table

5.

Some specific obser-

vations are as follows:

(1)

Overall

emissions

were

low.

Manygrades Indicated less than

100

lb

emlssionsfmillion

lb

PC

processed. Processingconditions

differed from resin to resin, most

notably by temperature, soemis-

sion

data

from

different resins

were not directly comparable.

Ii

I

~~coo2o~ooo

000000

0—00

0———’-(N

0

00000~~

—

9’cO

00J

e~)

C~W

0)~~(N

•-

•—

to

Q)tO

In

—

015

-

•.

cn,-,~

•

C

to

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

—

a

a

~

c-,

‘0

à

—

e-i

.-

o

a

a

a

~

0

01

(0

0)

0)

01

0)

0)

•_

eq

fl

01

01

•

•~N~

~

~R

a

~

S

g

•~

~

E

a

a

~

Li-

~

~

c

II

11

‘3’

In

H

I

S

U

lu

I

a

2

2

a

S

I

VaLre

52

July

2002

Jour-ti

0,

the A~-

& Waste ManagementAssociation

707

---

Rhodes et a!.

(2)

The

PC/ABS

blend produced the highest

emis-

sions. This

was predicted

by

the

previous

SPI.

sponsored ABS study.

(3)

Impact-modified

PC was the next highest emit-

ter.

Again,

this was expected because this blend

contained a toughener component.

Table S shows that verygood precision was observed for

the nine duplicate

runs

across all four measurement

tech-

niques. Calculated precision was 8

for particulatemat-

ter,

6

for

VOCs,

14

for

targeted VOCs, and

15

for

SVOCs.

Several

of

the

targeted

VOCs

were

either

nondetectable or present at extremely low

levels in

all

resins,

particularly carbon tetrachloride, methylene chlo-

ride,

o-xylene, and toluene.

Others,

such

as p,m-xylene

and styrene, were only present in the PC/ABS blend.

CONCLUSIONS

The data collected in this study provide processors with

a

baseline for estimating emissions generated by PC resins

processed under similar conditions. Discrepancies

between

total

VOCs

(as measured

by the total

hydrocarbon

ana-

lyzer) and total SVOCs (as measured by gas chromatogra-

phy) are

a

resultof differencesin Instrument calibrations.

The larger value of the two should be used to ensure con-

servative

estimates.

The

emission factors

reported

here

may not

represent

those for other

PC types or for other

methods

of

processing. Professional judgment and con-

servative

measures must be

exercised as

necessary when

using these data for estimating emission quantities.

REFERENCES

1.

sherman,

LM. Clean-Air

Rules Challenge

Processors;

Piasttcs Technot,

1995, 41(2). 83-86.

2.

Contos, D.A.; Hoidren,

Mw.;

Smith, aL.

Brooke,

ltc.;

Rhodes, V.L.;

Rainey, ML.