ILLINOIS POLLUTION CONTROL BOARD

September

20,

1979

IN THE MATTER OF:

EMISSIONS

OF CARBON MONOXIDE

)

R78-1

FROM STEEL MILLS

PROPOSED OPINION OF THE BOARD

(by Mr. Dumelle):

This proceeding was initiated on January 18, 1978 when

Interlake,

Inc.; United States Steel Corporation; WSC Corporation;

and Granite City Steel, Division of National Steel Corporation

(Petitioners)

filed a Petition for amendment of Rule 206(d)

of the Air Pollution Control Regulations.

The Petition was

published

in Environmental Register #165 on February 15,

1978.

Hearings were held on May 16, 1978

in Chicago; June

1,

1978 in Edwardsville; and September

8,

1978,

October 24,

1978, and March

7,

1979 in Chicago.

On April

12, 1979 the

Board denied a request by Granite City Steel that the proposed

amendment be submitted as

a tentative revision to the Illinois

State

Implementation Plan.

On June 18,

1979 the Institute

of Natural Resources submitted a study to the Board entitled

Economic Impact of Changing the Carbon Monoxide Emission

Limitations for Steel Mills,

R78—l

(INR Document No.

79/14).

Hearings on the study were held on July

23,

1979

in Chicago

and August

1,

1979 in Granite City.

On August 23, 1979 the

Board proposed that Rule 206(d) be deleted.

This Proposed

Opinion supports the Board’s Proposed Order.

Rule 206(d)

governs emissions

of carbon monoxide from

sintering plants, blast furnaces, and basic oxygen furnaces.

Petitioners operate steel mills

in Illinois which are affected

by Rule 206(d).

Republic Steel Company, although not a

petitioner, operates

a blast furnace and a basic oxygen

furnace in Chicago which are also affected.

None of the operating sintering plants, blast furnaces,

or basic oxygen furnaces at Petitioners’

facilities comply

with the 200 ppm corrected to 50

excess air standard of

Rule 206(d).

Throughout this proceeding various technologies

were evaluated which could possibly result in compliance.

Absorption in copper anmoniurn salts solution was discounted

because of the high oxygen content

in the exhaust gas streams

from Petitioners’

facilities.

This oxygen would have to be

absorbed along with the carbon monoxide which would result in

an operation of enormous size.

(Ex.W.,p.25)

Emissions from basic oxygen furnaces can be captured

35—421

---

through installation of a “closed hood”

or “off gas”

system.

These

waste

gases can then be

flared or reused elsewhere.

Interlake investigated

this

alternative and reported its

findings in Exhibit

H.

Interlake dismissed this technology

because the

cost to capture, store and reuse the gas was

prohibitive~

If the

gas

is simply burned instead of stored

for

reuse, the

resulting concentration of carbon monoxide

would still be

4,000 ppm,

far in excess of the 200 ppm limit

in Rule

206~d),

Republic Steel employs an

“off gas” collection system on

~ts

basic oxygen furnace in Chicago.

Since

no one from

Renublic

testified in this proceeding,

the performance of

this

system was difficult to gauge.

Petitioners felt that

since

the gases were being

flared, their estimate of 4,000

ppm of CO should be applied to Republic.

In a comment,

the

~gency

concluded that Republic might be complying

with Rule

206(i),

but that these emissions could not be measured.

The Board is not convinced that “off gas” systems on

basic oxygen furnaces constitute a feasible method of complying

with Rule

206(d) unless

the gases are captured and stored

for

reuse.

The

combustion which occurs directly above the

furnaces

employing an open hood system

(such as Petitioners’)

may already be destroying more carbon monoxide than closed

hood flaring.

If

capturing these exhaust gases

for reuse

becomes economically attractive in the future,

Petitioners

will not need encouragement from the Board to do so.

Catalytic oxidation was studied as an alternative

control

technique for all affected sources.

The authors of

the economic impact study concluded that poisoning of catalysts

from metal

oxide particulates might make catalytic oxidation

unworkable

in

the context of these sources.

(Ex.W.,pp.xii,

24

R.513)

The only known application of this technology to

steel making failed

to comply with the present Rule 206(d).

(Ex.~’.,p.5)

rp~~

Board was advised that this application has

recently been abandoned,

(R.118)

The Agency also stated

that catalytic

oxidation was not feasible.

(R.658)

Consequently

the Board

concludes that catalytic oxidation

is

not an

available alternative

for compliance with Rule 206(d).

Direct

flame incineration was the only technology

which,

at least in theory, might comply with Rule 206(d).

The

study authors were unable to find any equipment vendor

who

“,

,

,

could cite an installation where carbon monoxide

destruction

of

the

magnitude required by Rule 206(d)

actually

occurred,”

(Ex,W. ,p.4,

R,512)

Each

of the

Petitioners prepared an estimate of costs

to install and operate afterburners which might comply with

Rule

206(d).

(Ex.B,C,D,I,P)

The economic study authors

compared these estimates with their own calculations and

35

—4

22

---

—3—

included calculations based on maximum use of heat recovery.

(Ex.W.,pp.xii,36-37)

Assuming that heat recovery would be

used because of significant savings and excluding the costs

calculated from Republic because of the uncertainty associated

with Republic’s emissions, application of direct flame

incineration would cost Petitioners approximately $25 million

per year.

EFFECT ON AIR QUALITY

In

order to determine if the emission levels of CO

which would be allowed under the rule change

would lead to

violations of the ambient air quality standard for CO,

two

air quality simulation modeling studies were performed.

One

study,

(Ex..BB)

“Modeling of Carbon Monoxide Dispersion

from

Four Illinois Steel Production Facilities,” was done by

Environmental Quality Research, Inc.

(EQR)

for the steel

mills

in

the Chicago area,

The other study,

(Ex.L)

“A

Program

of

Ambient Air Quality Analysis for Granite City

Steel,

Granite City,

Illinois,” was done by Air Resources,

Inc.

(ARI).

Both studies determined what meteorological

conditions would

lead to worst—case

(highest) ground

level

concentrations of CO and then calculated those concentrations.

In both

studies,

the maximum concentration occurred when the

wind

direction was such that the plumes from individual

point sources

(stacks) of

CO were aligned.

The maximum

concentrations were due to a combination of all sources.

The emission inventory data for the Chicago area study

was supplied to

EQR by Technical Center,

Interlake,

Inc.

Much of this

data

was based on actual stack measurements

during typical

operating conditions.

(R.162)

However,

since

WSC Corp.

is not presently operating its

sinter plant,

emission levels used in the study for it were extrapolated

to the levels that would be expected if the sinter plant were

operating.

(R,13)

Early versions of the modeling study

(Ex.J,Q,R)

found

that the U.S.

Steel

blast furnace contributed significantly

more to the maximum ground level concentration than any

other source.

Upon further investigation,

it was found that

the

CO emission rate for USS’s blast furnace was the highest

reported CO concentration for that source.

The concentration

that should have been used in the study was the average of

several readings which were taken at three minute intervals.

(R.549)

Such an average would more closely approximate a

one hour average than the highest instantaneous sample.

Since

the air quality simulation

model calculates one hour

averages,

it

would be more appropriate to use an emission

rate that approximates

a one hour average than it would be

to use the highest instantaneous value.

The final EQR

report

(Ex.BB)

used the average

CO reading for U.S. Steel’s

blast

furnace.

35—423

---

—4--

EQR used U.S.

EPA’s UNAMAP model PTMAX to determine the

location of regions where maximum ground—level concentrations

would occur from each source.

These areas were then super—

positioned

to “producej

a pattern of candidate regions for

the maximum ground—level concentration due to the combined

sources.”

This information was used to plan the locations

of the receptors for which concentrations would be calculated

using U.S. EPA’s UNAMAP model

PTMTP.

Since the report

authors

found

‘that their version of PTMAX occasionally

produced erroneous concentrations, they recalculated all of

these concentrations using “Workbook of Atmospheric Dispersion

Estimates”

(Turner, D.B,

1969).

Next EQR calculated one—hour ground

level concentrations

using

PTMTP,

The calculations were performed for

36 combinations

of wind speed and stability classes which correspond to the

STAR meteorological data classes

for Chicago during the last

five years.

A 300 meter mixing depth was used for all

calculations

(R,392).

The expected 8—hour average calculations

were derived from

1—hour averages by using the method described

in Turner’s Workbook

(“Workbook of Atmospheric Dispersion

Estimates,” Turner, p.37,

1970,)

This

information

was used

to determine the location and magnitude of the maximum

1—hour and 8-hour ground—level

concentrations

for each class

of

meteorological conditions

(Ex..BB,pp,4-5).

The maxini~m

calculated one—hour average ~as

about 4000 to 5000 ug/m

(the standard is 40,000 ug/m

~.

The maximum eight—hour

average was 27~0to 3360 ug/m

(the corresponding standard

is 10,000

ug/ni

).

If the lake breeze effect

is taken into

account,

the calculated concentrations would be approximately

double

(Ex,BB, p.29).

This method of calculation was questioned

at hearing

since

it used dispersion coefficients which are more appropriate

for dispersion simulations

in rural

areas than

it is

for

urban areas.

In general, heat sources, structures,

etc.

found in urban

areas tend to decrease atmospheric stability.

Some dispersion modelers feel

that the UNAMAP model

RAM

would be more appropriate since it

uses dispersion coefficients

developed

from studies of urban areas

(R.649).

This may be

true,

However,

a certain amount of conservatism was built

into the study that EQR did.

According to their calculations,

class

A stability

(the most unstable) caused the highest

ground—level CO concentrations from the point sources.

EQR

used these concentrations

for determining the ground—level

impact,

even though there

is some debate

in the literature

that class A stability may not exist

(R.398,421).

(The next

more stable condition is class

B for w~ichthe maximum

calc4lated concentration was 3634

ug/rn

as compared to 4795

ug/m~for class A.)

ft.

Bruce Turner in

“Atmospheric Dispersion Modeling:

A

Critical Review”

(official notice taken of Journal

of

the Air Pollution Control Association,

Vol.

29, No.

5,

pp.

35—424

---

—5—

502—519, May 1979)

states that there are several sources of

error in

dispersion modeling.

Some

of these are due to

oversimplifications

such as the use of

“a single stability

condition with its related dispersion rate”,

and “the assumptions

of complete eddy reflection at the ground and the top of the

mixing height.”

(Turner, p.512)

These oversimplifications

which were used

in this study would

lead to an overcalculation

of ground-level concentrations.

The Board

finds that although

EQR could have used the more sophisticated and possibly more

accurate version of the urban RAM model, the model that was

used probably overestimates ground—level CO concentrations

and

for

purposes of this proceeding is adequate.

The

ground—level

CO concentrations from the point

sources were overlaid on the CO concentrations due to mobile

sources

in

the area.

EQR used Illinois Department of Transportation

Environmental Policy Processes-Air Quality Manual

(IDOT

Manual)

to estimate the mobile source contribution.

The

composite emission factor used was a

1978 emission factor.

The

percentiles of vehicle type and speed used

in the

modeling study were extracted from data in the IDOT manual.

Cook County’s 1975 traffic map was used to obtain information

on average daily traffic

(ADT)

(Ex.BB,p.5).

However,

since

the modeling study over estimated the average traffic speed

during the worst 8—hours, the emission factors used are

probably less than they should be.

Consequently, the calculated

CO concentrations

from mobile sources may be under

estiffiates.

The roadway maximums were determined by calculating CO con—

centrations at receptors located 50

feet from the roadways with

the highest ADT volumes

in the area.

A wind-roadway angle of

22.5 degrees and class

F stability were used since the constitute

the most adverse meteorology for dispersal of roadway pollution.

In reality,

there is a very small percentage of occurrence

of

class

F stability which coincides with maximum traffic

conditions.

One or two percent of the total

class

F stability

occurring in any period of time could reasonably coincide

with high traffic volumes

(i.e. the eight—hour maximum

volume),

Therefore, the assumption of class

F stability for

a

‘worst probable’ condition for computational purposes

presents

an ultra—conservative approach to prediction of

pollution concentrations due to motor vehicles.

(Official

notice taken of the Illinois Department

of Transportation

Air Quality Manual,

Sept,

1978, prepared by Leonard

F.

Vik

and Miles

E.

Byers, pp.2—9 to 2—10).

The general ambient background was assumed to be 625 ug/m3

(0.5 ppm).

This value does not include any significant point,

area,

or roadway sources that might be

in the area

(R.648).

The worst—case maximum concentrations were obtained by

adding the contributions from the stationary sources, roadways

35—425

---

—6—

and background.

The maximum along the roadway was the sum

of the concentration due to

the stationary sources at that

location with class

A stability (worst—case meteorology),

roadway contributions at 50 feet from the roadway, and

background.

The modeling study did predict that violations of the

8--hour

standard would occur along

some of the roadways.

The

report states that “the

highest expected 8—hour CO concentration

within the ~sig~ificant

area’, which

is defined by the area

having

625

119/rn

3or more stationary source cSntributions,

is

about 21525 ug/m

It consis9

of 1156 ug/m

(5)

from

stationary s9urces,

19744 ug/m’

(92)

from mobile sources,

and 625 ug/rn~ (3)

of background”

(Ex.BB,p.29).

The meteorological

conditions that would lead to this situation are expected to

occur

less than once per year.

The highest possible CO

conc?~ntration

in the area with the highest ADT is 54503

uq/m~for

the next 8—hour average.

“The contributions from

the

stationary sources on this area is negligible”

(Ex.BB,p.30).

Thus

it

appe~rsthat exceedences of the eight—hour standard

(10,000

ug/m

)

would occur even if the steel mills ceased

emitting CO altogether and that the amount of CO which the

steel mills contribute to

the violation is very small

in

comparison to the amount contributed by mobile sources.

The

study does not predict that either the 1-hour or

the more stringent 8-hour standard will be exceeded in the

areas where the steel mills would have their maximum impact

even when the lake shore effect is

considered.

These areas

are

located

far enough from the large mobile sources that

the

CO attributable to these mobile sources at these locations

is negligible.

In summary, although

the

EQR

modeling

study

has

some

shortcomings,

it

has been a useful tool to illustrate that

:Ln

the areas of the steel

mills, mobile sources contribute

the

most

CO to ground—level concentrations.

The air quality simulation study for the Granite City

Steel

facility

(Ex.L)

was done using the ARI Air Quality

Analysis Model

(AQAM),

This model

is based on Turner’s

Workbook and uses Pasquill—Gifford Gaussian dispersion

equations, modified Holland or Briggs plume rise equations

and dispersion parameters generalized by Slade.

In this

particular study,

the

)3riggs plume rise formula, incorporating

stack aerodynamic downwash effects was used.

The CO emission

data for the

Granite City Steel facility was supplied by

that facility.

Since the meteorological condition that could produce

the maximum short—term ground—level concentration was not

observed in that area during the five years between January

1970 and December

1974,

ARI also calculated the maximum

35—426

---

—7—

one—hour ground—level concentration for the most adverse

observed meteorological condition.

The theoretical short—term

maximum

concentration due to the point sources alone was

found

to he

5.7 ppm,

while the maximum value calculated for

observed meteorology was

2,4 ppm.

Both these values are

well below the one—hour standard of

35 ppm.

The other

concentrations calculated

for longer averaging times were

also well below the respective standards.

The ARI study also found

that

“vehicular

emissions

from

major streets and highways in this region are the major

sources of CO.

The highest ground concentrations from this

source

usually occur under stable conditions and low wind

speeds.”

These are not the same conditions which would

lead

to maximum ground—level concentrations due to elevated point

sources.

The highest CO concentration attributable to point

sources was calculated to

be 0,307 ppm

(Ex.L,p.v-1).

In conclusion, the Board believes that allowing the

Granite City Steel Company to continue emitting CO at its

current rate will not cause

a violation of any CO standards

and will not create

a human health problem.

In the

Illinois State Implementation Plan for CO

(Ex,DD), the Illinois EPA has determined that in the two

areas of the state with CO air quality violations

(the

Chicago six county area and the Peoria two county area),

83

of the CO emissions is from on—highway mobile sources.

These sources make the most significant contributions to

violations because, in addition to producing most of the CO,

their emissions occur near ground—level and near heavily

populated areas.

Carbon monoxide from stationary sources,

on the other hand,

is generally emitted from stacks elevated

above ground

level

so that the CO concentration

is greatly

diluted by the time

it reaches ground—level.

Hence,

“

.

the CO

SIP is

based on control of emissions from mobile

sources,

with emissions from stationary and area sources

considered in the analyses as secondary effects,”

ECONOMIC

IMPACT

The Illinois Institute of Natural Resources submitted

to the Board

a document entitled Economic Impact of Changing

the Carbon Monoxide Emission Limitations

for Steel Mills,

R78—1

(IINR

Doc.

No,

79/14)

(Ex.W, hereafter “Study”).

Section

2 of the Study described the affected processes

(and

related emissions)

and discussed control technology.

Estimated

capital costs were compared

(Ex,W.,

Tables

2—5

through 2—8).

Section

3

analyzed three control scenarios in terms of

prices,

output, and employment.

Section 3.1, Economic Characteristics of the Iron and

Steel Industry, while not determinative in this proceeding,

---

—8—

does provide

a useful description,

Section 3.2 and 3.3

make

it clear that compliance with the existing Rule would

have severe economic consequences for the Illinois steel

industry.

Section

3,4 summarizes those consequences

(Ex.W.,

Table

3—22).

Section 4,0,

Health and Environmental Consequences,

examines the differences between compliance with the existing

Rule 206(d)

and the proposal

of P78—i

(Ex.W.,p.90).

The

effects are analyzed both qualitatively and quantitatively.

The qualitative discussion encompassed effects on human

health,

fauna and flora, and energy demand.

The quantitative

analysis

attempted to monetize the

carbon

monoxide damage which is attributable to the proposal

of P78—i.

~fl,io

methods were used: one used a damage factor

based

on emission rates and the other used a damage factor

based on

ambient levels.

Both approaches yield results

which

are of limited

usefulness.

The author of this section

of the Study attributed

the greatest uncertainty to •the

estimates of

health effects at low ambient concentrations

(Ex,W,

1.p,li9)

The Agency raised objections to the methodologies used

to

assess CO damages and concluded that

“.

.

.

the quantitative

information

.

.

.

must be regarded with

a degree of

skepticism

that would negate its value

for decision—making purposes”

(Ex.CC,p.5).

The Board determines that the proposed regulation may

have

a slight adverse economic impact on the people of the

State of

Illinois due to the slightly higher ambient levels

of

CO that

will

be permissible.

However, a positive economic

impact also accrues

to the people of the State of Illinois

through the abolition of the existing rule.

DELETION OF RULE 206(d)

At the last hearing in

‘this matter,

Petitioners endorsed

the Agency’s proposed language for Rule 206(d)

in Exhibit V.

The Agency’s proposal was similar to Rule 206(h) which

limits emissions based on the fuel value of the waste gas

stream.

The proposal also would have required

a

fall back

to the present

Rule

206(d)

in the event that any affected

source contributed

to

a violation of ambient air quality

standards by more than specified increments.

The Board has chosen to reject the proposed language

in

Exhibit V for the following reasons,

First,

since none of the

affected sources has

a fuel value which approaches

20

of that required for flame incineration of the waste gas

stream

at 1460°F (Ex.B,D,P;

P.112), the proposal would

not have required any controls.

Second, the Board can see no

35—428

---

—9—

reason to require a fallback to 200 ppm when that standard

is either impossible to achieve or unreasonably expensive.

Third, the specified increments are borrowed from Federal

regulations

(Ex.T,U) with no record to support them here,

Fourth, the language of Exhibit V does not specify how these

violations or contributions to violations are to be calculated.

Fifth, and most important,

the record in this proceeding

demonstrates that controlling carbon monoxide emissions from

these sources, to any level, will have no significant effect

on ambient air quality.

Mr.

Werner concurs.

I,

Christan L.

Moffett, Clerk of the Illinois Pollution

Control Board, hereby certify the above Pro os d 0 inion was

adopted on the

__________________

day of

_________________,

1979

by a vote of

~

istan L.

Moffet!t

erk

Illinois

Pollution

rol Board

35—429