Date:	 September 15, 2011

Subject: 	Technical Approach for Mineral Wool MACT Floor Calculations

	EPA Contract No. EP-D-06-118; EPA Work Assignment No. 4-10

	RTI Project No. 0210426.004.010

From: 		Cindy Hancy

		Dave Reeves

To: 		Susan Fairchild, EPA/OAQPS/SPPD/MMG

Introduction

Section 112 of the Clean Air Act (CAA) requires that the U.S.
Environmental Protection Agency (EPA) to establish National Emission
Standards for Hazardous Air Pollutants (NESHAP) for the control of the
hazardous air pollutants (HAP) emitted from both new and existing major
sources in a source category.  These standards must reflect the maximum
degree of reduction in the HAP emissions that is achievable.  The
minimum level of control is referred to as the “Maximum Achievable
Control Technology (MACT) floor.”  The method for determining the MACT
floor for a NESHAP is defined for both new and existing sources in CAA
section 112(d)(3).  For new sources, the MACT floor cannot be less
stringent than the emission control that is achieved in practice by the
best-controlled similar source.  For existing sources, the MACT floor
cannot be less stringent than the average emission limitation achieved
by the best-performing 12% of existing sources for source categories
with 30 or more sources, or the best-performing five sources for source
categories with fewer than 30 sources.

The purpose of this memorandum is to present the data, the methodology,
and the results of the MACT floor analysis for the Mineral Wool
Production NESHAP source category. This analysis is part of EPA's
obligation under CAA section 112(f)(2) and 112(d)(6) to conduct a
residual risk and technology review.  The MACT floor analysis is also
being performed in response to a petition for rulemaking by the Natural
Resources Defense Council and Sierra Club that states EPA failed to set
emission limits for COS and other HAPs emitted by mineral wool
facilities and challenges HAP surrogacies claimed in the current Mineral
Wool NESHAP. This MACT floor analysis uses data collected from a
nationwide voluntary Information Collection Request (ICR) of mineral
wool manufacturers conducted by EPA in 2010.  Data on process
operations, emission controls, and air emissions data reported by
respondents to the ICR were compiled into a Microsoft Access data base
that serves as the data set used for this MACT floor analysis (referred
to in this memorandum as the “ICR data set”).

Background Information

The current Mineral Wool Production NESHAP was promulgated on June 1,
1999 and applies to existing, new, and reconstructed cupolas and curing
ovens at mineral wool production facilities that are major sources of
HAP emissions. The 1999 Mineral Wool Production NESHAP sets particulate
matter (PM) emission limits for new and existing cupolas and carbon
monoxide (CO) limits for new cupolas.  The NESHAP also established
emission limits for formaldehyde from new and existing curing ovens.
Based on the 1999 NESHAP, these regulated pollutants currently serve as
surrogates for other HAPs as follows: PM serves as a surrogate for HAP
metals, CO for carbonyl sulfide (COS), and formaldehyde for phenol and
methanol.  

The mineral wool production source category currently consists of 7
facilities in the US; combined, these 7 facilities operate 11 cupolas
and 3 curing ovens. The three facilities that operate curing ovens
produce bonded mineral wool products.  In addition to curing ovens,
sources of HAP emissions associated with bonded product manufacturing
include collection operations.  There were no MACT emission limits
established for collection operations in the 1999 Mineral Wool NESHAP.
Bonded lines use HAP containing binders that result in formaldehyde,
phenol, and methanol emissions from curing ovens and collection sources.
The stack configurations of curing and collection operations vary at the
three facilities operating bonded lines. The proposed amendments to the
rule as a result of this review will include formaldehyde, methanol, and
phenol limits for collection operations; however, due to the differences
in processes, associated exhaust streams, and stack configurations at
the three bonded lines, EPA is proposing pollutant-specific emission
limits for both curing ovens and collection operations combined.  

The specific chemicals, compounds, or groups of compounds designated as
HAP are listed in CAA section 112(b). From this list, the following
additional HAPs were identified as being emitted from cupolas and would
be regulated for new and existing sources in the proposed rule: carbonyl
sulfide (COS), hydrogen fluoride (HF), and hydrogen chloride (HCl).  

MACT Floor Analysis - Subcategories

Under CAA section 112(d)(1), EPA has the discretion to “...distinguish
among classes, types, and sizes of sources within a category or
subcategory in establishing...” standards.  When separate
subcategories are established, a MACT floor is determined separately for
each subcategory.  To determine whether the mineral wool production
facilities warrant subcategorization for the MACT floor analysis, EPA
reviewed unit and process designs, operating information, and air
emissions data compiled in the ICR data set and other information
collected by the Agency for development of the NESHAP for this source
category.  Based on this review, EPA concluded that there are
significant design and operational differences in the collection
operations at each of the three facilities that operate a bonded line. 

These three collection subcategories are listed below. 

Vertical Collection:  In this design, the molten rock/slag mixture is
poured from the cupola spout onto a group of stainless steel drums
spinning in opposite directions.  The spinning drums form fine fibers
of the mineral mixture.  High air volume directs the fibers off the
fiberization spinners, toward a fast-moving porous vertical conveyor
belt. A strong vacuum is drawn on the opposite side of the belt, causing
the fibers to lie against the vertical belt as it moves upward.  At the
top of the conveyance, the belt travels around a curve, the vacuum is
released, and the fibers are removed on a second belt that conveys the
layer of binder-sprayed mineral wool fibers into the curing oven. 

The vertical conveyer belt requires high volumes of air to be drawn
through the belt and results in the collection of a thin fiber layer
upon the belt.  In this design, ‘shot’, BB-sized black granules
that are high in iron (a result of using slag from the iron and steel
industry), fall out of the fiber layer.  The vertical design is used to
produce a specific type of mineral wool that is low in ‘shot’ and
may be used in the hydroponic gardening market, as well as, in a
specialized market of insulation products in which shot is undesirable. 

Horizontal Collection: This process is similar to vertical collection,
except that the conveyor belt works with gravitational forces and
results in a thicker layer of mineral wool collected upon the horizontal
belt.  This method also requires lower air volume drawn through the
fiber layer than in the vertical design, and the ‘shot’ is not
selectively removed. The air stream is conducive to thermal oxidation at
the hottest part of the cupola exhaust stack or at the existing thermal
oxidizer on the curing oven. 

Drum Collection:  In this design, fibers are drawn using a very high
volume air flow into the center of a rotating drum.  The sides of the
rotating drum have small holes through which the air flow may exit but
the fibers are caught.  The angle of the drum, the vacuum, and
centrifugal forces pull the fibers against the inside wall of the drum
and out the end. The entire drum is enclosed and the air flow may be
vented to the hottest part of the cupola exhaust stack or to an existing
thermal oxidizer on the curing oven. 

MACT Floor Analysis Methodology 

Existing Sources 

A MACT floor analysis was completed for each proposed regulated
pollutant as summarized in Table 1. 

Table 1 – Summary of Proposed Limits

	Existing Sources	New Sources

Cupolas

PM	Current PM limit will remain the same. No MACT floor analysis
performed.

CO	EPA is proposing to remove the current CO limit since it was/is a
surrogate for COS in the 1999 NESHAP

COS	X	X

HF	X	X

HCl	X	X

Curing & Collection (combined)

Formaldehyde	X	X

Phenol	X	X

Methanol	X	X



A MACT floor analysis was not performed for PM since EPA is not
proposing any changes to the current PM limits. The current formaldehyde
limit only applies to curing ovens. As previously mentioned, EPA is now
developing new MACT emission limits for collection operations.
Therefore, a MACT floor analysis that includes both curing and
collection was performed for new and existing sources.  No current
emission limits exist for COS, HF, HCl, phenol, or methanol; in order to
propose limits for each pollutant, a MACT floor analysis was performed
for each of these pollutants for new and existing sources.  

The first step in the MACT floor analysis for each regulated source and
HAP was to rank each unit (for which emissions data was provided) by
emission level (lowest to highest) for each pollutant.  From this
ranking, a MACT floor pool of sources was identified for determining the
minimum control level allowed for the MACT floor, consistent with the
criteria defined for new and existing sources by CAA section 112(d)(3). 
For the new source MACT floors, the best-controlled source was
identified for which there were individual source test run data in the
ICR data set.  For the existing source MACT floors, selection of the
MACT floor pool size (i.e., number of emission units to be included in
the determination of the average emission limitation value) was
determined on an individual unit category basis as described below. 

Cupolas.   This category includes fewer than 30 sources.  Therefore, the
existing source MACT floors for COS, HF, and HCl emissions were based on
the top 5 best performing cupolas. For new sources, the MACT floor was
based on the best performing cupola for each pollutant. 

Curing & Vertical Collection.  Currently, only one facility operates
this type of collection.  Formaldehyde, phenol, and methanol MACT floors
for new and existing sources were based on emission test runs for
combined curing and collection operations from this facility. 

Curing & Horizontal Collection.  Currently, only one facility operates
this type of collection.  Formaldehyde, phenol, and methanol MACT floors
for new and existing sources were based on emission test runs for
combined curing and collection operations from this facility.

Curing & Drum Collection.  Currently, only one facility operates this
type of collection.  Formaldehyde, phenol, and methanol MACT floors for
new and existing sources were based on emission test runs for combined
curing and collection operations from this facility.

The next step in the MACT floor analysis was to account for data
variability in the calculations of the applicable MACT floor limits for
the subcategories using the data’s 99% upper prediction limit (UPL). 
Specifically, the MACT floor limit was determined as the UPL calculated
with the Student’s t-test using the “TINV” function in Microsoft
Excel software.  The UPL approach has also been used in other EPA
rulemakings (e.g., NESHAP for Portland Cement, NSPS for
Hospital/Medical/Infectious Waste Incinerators, NESHAP for Industrial,
Commercial, Institutional Boilers and Process Heaters, and NESHAP for
Electric Generating Units) to account for variability in emissions data
for a specified level of confidence.  The level of confidence represents
the level of protection afforded to facilities whose emissions are in
line with the best performers.  For example, a 99% level of confidence
means that a facility whose emissions are consistent with the best
performers has one chance in 100 of exceeding the floor limit.  A
prediction interval for a single future observation (or an average of
several test observations) is an interval that will, with a specified
degree of confidence, contain the next (or the average of some other
pre-specified number) of randomly selected observation(s) from a
population.  In other words, the upper prediction limit estimates what
the upper bound of future values will be, based upon present or past
background samples taken.  The UPL consequently represents the value at
which we can expect the mean of future observations for the HAP
emissions to fall within a specified level of confidence, based upon the
results of an independent sample from the same population.  This method
accounts for the point-to-point variability in the data.  

The form of the UPL equation differs somewhat depending upon the number
of data points and data distribution to which it is applied. Attachment
A includes a flow diagram that summarizes the UPL approaches used. To
this end, the data sets were evaluated for each HAP to ascertain whether
the data were normally distributed, or fit another type of distribution
(e.g., log normal distribution).  According to the Central Limit Theorem
(Durrett, 1996), when a data set includes 15 or more sources, the UPL is
based on the assumption that the data fit a normal distribution.  The
Central Limit Theorem states that regardless of the shape of the
original distribution, if the distribution has a finite mean (μ) and
variance (σ²), the sampling distribution of the mean approaches a
normal distribution with a mean of (μ) and a variance of σ²/N as N,
the sample size, increases (Durrett, 1996). 

The mineral wool data sets used to calculate MACT floors for each
pollutant all contained less than 15 sources. When the sample size is
smaller than 15 and the distribution of the data is unknown, the Central
Limit Theorem cannot be used to support the normality assumption. 
Statistical test of the kurtosis, skewness, and goodness of fit test are
then used to evaluate the normality assumption.  The skewness statistic
(S) characterizes the degree of asymmetry of a given data distribution. 
Normally distributed data have an S value of 0.  An S value that is
greater (less) than 0 indicates that the data are asymmetrically
distributed with a right (left) tail extending towards positive
(negative) values.  The standard error of the skewness statistic (SES)
was also used in determining the normality of the data distribution. 
The kurtosis statistic (K) characterizes the degree of peakedness or
flatness of a given data distribution in comparison to a normal
distribution.  Normally distributed data have a K value of 0.  A K value
that is greater (less) than 0 indicates a relatively peaked (flat)
distribution.  The standard error of the kurtosis statistic (SEK) was
also used in determining the normality of the data distribution.

For each data set to which the UPL was applied (i.e., the separate data
sets for each HAP applicable to a source), the S and K values were
calculated using the reported test values.  If both kurtosis and
skewness tests indicate the data is normally distributed, the UPL was
calculated using the UPL pooled variance Equation 1.  

   		Equation 1

where:

  	=  the average (mean) of the best performing existing sources;

	t(p,df) 	=  the t statistic for a confidence level p, and df degrees of
freedom;

	s2	=  the pooled variance;

	n	=  the total number of test runs (all sources) used in the analysis;
and

	m	=  the number of (future) compliance test runs [for run-by-run data,
m=3].

Degrees of freedom calculated by:

 							    Equation 1a

Mean calculated by: 

                                                                        
               Equation 1b

Pooled variance calculated by:

                                                      		    Equation 1c

If the kurtosis or skewness tests indicate the data was not normally
distributed, the data were log-transformed. Once the logs of all the
test runs were calculated, new kurtosis and skewness tests were
performed on the log-transformed data. If both kurtosis and skewness
tests indicated the log-transformed data were normally distributed, the
UPL can be calculated using the Equation 2.  

                     Equation 2

where:

	μ	=  the average of the best performing existing sources;

  	=  the z statistic for a lognormal distribution at 99 percent;

 	=  the variance;

	n	=  the total number of test runs (all sources) used in the analysis;
and

	m	=  the number of (future) compliance test runs [for run-by-run data
m=3].				        	

Mean is calculated by:

                                                                        
                        Equation 2a

Variance is calculated by:

                                                                        
           Equation 2b

If the raw data and the log-transformed data were not normally
distributed and n ≥13, the UPL was calculated using the UPL pooled
variance with skewness adjustment in Equation 3.  Note this adjustment
cannot be used if the number of individual runs is less than 13.  

                                                                        
    Equation 3

where:

  	=  the average of the best performing existing sources;

	t(p,df) 	=  the t statistic for a confidence level p, and df degrees of
freedom;

	s2	=  the pooled variance;

	n	=  the total number of test runs (all sources) used in the analysis;

	m	=  the number of (future) compliance test runs [for run-by-run data
m=3]; and

	Skew	=  the skewness of the test runs used in the analysis.

	

If the raw data and the log-transformed data were not normally
distributed and n was less than 13 the UPL was calculated using the UPL
pooled variance in Equation 1, but instead of using the Excel TINV
formula to find the t statistic t(p,df), a trial and error method was
used to find the t static that gives a 99 percent confidence level by
correcting the probability values  (p) using Equation 4

λ3(t) – Kurtosis x Pλ4(t) + λ32xP λ3/2(t)       Equation 4

Adjustment for Below Detection Level Emissions Data

Prior to calculating the UPLs, test runs that were below detection limit
(BDL) were identified.  If a dataset contained any BDL values, the UPL
was calculated using BDL values as reported in the test report. After
the UPL was completed for a particular source and pollutant, a
comparison of the UPL and the representative method detection level
(RDL) was performed.  The first step in the comparison was to, for a
particular emissions unit category, find the mean of the emissions data
set, which was calculated during the UPL calculations. The second step
was to identify the largest BDL value of the dataset that was no greater
than the mean of the entire dataset. This value was the RDL for the
pollutant data set of the emission unit category. The RDL was then
multiplied by 3 and the product is compared to the UPL.  If the 3x RDL
value was less than or equal to the calculated UPL value, the UPL value
was used as the proposed MACT Floor Limit. If the 3x RDL value was
greater than the calculated UPL value, the 3x RDL value was used as the
proposed MACT Floor Limit. If no BDL values were included in the
pollutant data set for a particular unit type, these steps were not
performed. 

MACT Floor Analysis Results

-



 

!

 h

h

Y

h

hç

- h¢ 

- h¢ 

h¢ 

h¢ 

	”

Æ

hd

hd

hd

摧ᥤÔ

.Production source category.  

Table 2. Summary of Proposed MACT floors for the Mineral Wool Source
Category

	Total Sources in Industry	Mean of Sources Used in UPL Analysis	99% UPL
Caclulated for Existing Sources	99% UPL Caclulated for New Sources

Cupolas	11

Carbonyl Sulfide

1.07	3.30	0.017

Hydrogen Flouride

0.0089	0.014	0.014

Hydrogen Chloride

0.0066	0.0096	0.0096

Curing & Drum Collection	1

Formaldehyde

0.045	0.067	0.067

Phenol

0.00039	0.0023	0.0023

Methanol

0.00011	0.00077	0.00077

Curing & Vertical Collection	1

Formaldehyde

0.45	0.46	0.46

Phenol

0.50	0.52	0.52

Methanol

0.62	0.63	0.63

Curing & Horizontal Collection	1

Formaldehyde

0.054	0.054	0.054

Phenol

0.013	0.15	0.15

Methanol

0.0081	0.022	0.022



Attachment A

 

DRAFT 3-1-11

  PAGE   \* MERGEFORMAT  10 

