INFORMAL PUBLIC HEARINGS FOR THE PROPOSED RULE

ON OCCUPATIONAL EXPOSURE TO

RESPIRABLE CRYSTALLINE SILICA

+ + +

UNITED STATES DEPARTMENT OF LABOR

OCCUPATIONAL SAFETY & HEALTH ADMINISTRATION 

+ + +

March 26, 2014

9:30 a.m.

Frances Perkins Building Auditorium

200 Constitution Avenue, N.W.

Washington, D.C. 20210

	

BEFORE: 	DANIEL F. SOLOMON

	   	Administrative Law Judge

DEPARTMENT OF LABOR (DOL):

ANNE RYDER

Attorney, Office of the Solicitor

KRISTEN LINDBERG

Attorney, Office of the Solicitor

OCCUPATIONAL SAFETY AND HEALTH ADMINISTRATION (OSHA):

WILLIAM PERRY

Acting Director, Directorate of Standards and Guidance 

JESSICA SCHIFANO

Office of the Director 

ANNETTE IANNUCCI

Health Scientist, Office of Chemical Hazards 

- Non-Metals

JOSEPH COBLE, Sc.D., CIH

Director, Office of Technological Feasibility 

TOM MOCKLER 

Acting Director, Office of Regulatory Analysis 

- Safety

JANET CARTER

Office of Chemical Hazards - Metals

STEPHEN SCHAYER

Office of Physical Hazards

TIFFANY DeFOE

Office of Chemical Hazards - Metals

B.J. ALBRECHT

Chemist, Salt Lake Technical Center

DANIEL JOHANSEN

Chemist, Salt Lake Technical Center 

ROBERT BURT

Acting Deputy Director, Directorate of Standards and Guidance	

OCCUPATIONAL SAFETY AND HEALTH ADMINISTRATION (OSHA) (cont.):

ROBERT STONE

Director, Office of Regulatory Analysis - Health 

INTERNATIONAL UNION, UAW:

ANDREW COMAI

Coordinator, UAW Health and Safety Department

ANDREW MERCER

Recording Secretary, UAW Local 8

JEFF P'POOLE

President, UAW Local 523

STAN BURKEEN

Trustee, UAW Local 523

RODNEY GRAVES

UAW Local 2317

GREG ESSEX

President/Bargaining Chair, UAW Local 226

SHAWN RAGLE

Health and Safety Chairman, UAW Local 974

ROBERT HITCHCOCK

Bargaining Chair, UAW Local 211

RICHARD BOECKER

Safety Rep, UAW Local 211

MATTHEW WAFFORD

Steward, UAW Local 2339

DARIUS D. SIVIN, Ph.D.

Health and Safety Department, United Auto Workers

AMERICAN CHEMISTRY COUNCIL (ACC) 

CRYSTALLINE SILICA PANEL:

JACKSON MORRILL

Director

NEIL KING

Attorney

SAM BOXERMAN

Attorney, Sidley Austin 

PETER MORFELD, Ph.D.

Head of Institute for Occupational Epidemiology and Risk Assessment of
Evonik Industries

PAUL K. SCOTT

Cardno ChemRisk, LLC

KELLY BAILEY, CIH

Chairman

Director of Safety, Health and Environment

Vulcan Materials Company 

RICHARD LEE, Ph.D.

President/CEO, RG Lee Group 

JACK WAGGENER

Senior Principal/Engineering Manager, URS

STUART SESSIONS

President, Environomics, Inc.

American Chemistry Council's Crystalline Silica Panel

Construction Industry Safety Coalition 

OTHER PARTICIPANTS:

PEG SEMINARIO

Safety and Health Director, AFL-CIO

JAMES FREDERICK

Assistant Director, Health, Safety and Environment  United Steelworkers
Union

CHRIS TRAHAN

Building and Construction Trades Department, AFL-CIO

REBECCA REINDEL

AFL-CIO

FRANK HEARL

Chief of Staff, NIOSH

DR. ROBERT PARK

Risk Evaluation Branch, NIOSH

INDEX

										PAGE

		

INTRODUCTION			

Judge Daniel F. Solomon						  PAGEREF a1Solomon \h  1828 		  			

INTERNATIONAL UNION, UAW

	Andrew Comai							  PAGEREF a2Comai \h  1829 

	Andrew Mercer							  PAGEREF a3Mercer \h  1832 

	Jeff P'Poole							  PAGEREF a4PPoole \h  1837 

	Stan Burkeen							  PAGEREF a5Burkeen \h  1841 

	Rodney Graves							  PAGEREF a6Graves \h  1843 

	

	Greg Essex							  PAGEREF a7Essex \h  1846 

	Shawn Ragle							  PAGEREF a8Ragle \h  1854 

	Robert Hitchcock						  PAGEREF a9Hitchcock \h  1858 

	Richard Boecker						  PAGEREF a10Boecker \h  1865 

	Matthew Wafford						  PAGEREF a11Wafford \h  1870 

	Questions								  PAGEREF a12Q \h  1876 

INDEX

										PAGE

AMERICAN CHEMISTRY COUNCIL (ACC)

CRYSTALLINE SILICA PANEL	

	Jackson Morrill						  PAGEREF b1morrill \h  1928 

	Neil King and Sam Boxerman				  PAGEREF b2King \h  1933 

	Peter Morfeld, Ph.D.					  PAGEREF b5Morfeld \h  1944 

	Paul K. Scott							  PAGEREF b6Scott \h  1963 

	Kelly Bailey, CIH and Richard Lee, Ph.D. 	  PAGEREF b7Bailey \h  1987 

	Jack Waggener							  PAGEREF b10Waggener \h  1998 

	Stuart Sessions						  PAGEREF b11Sessions \h  2022 

	Questions								  PAGEREF b12Q \h  2043 

	

ADJOURNMENT								  PAGEREF cAdjourn \h  2136 						

EXHIBITS

EXHIBITS		DESCRIPTION				 	PAGE 

Exhibit 53	UAW PowerPoint					  PAGEREF e54mr \h  1926 

Exhibit 54	UAW videos					  PAGEREF e54mr \h  1926 

Exhibit 56 	Mr. King/Mr. Boxerman 

			presentation					  PAGEREF e56_64mr \h  1928 

Exhibit 57	Dr. Morfeld's presentation		  PAGEREF e56_64mr \h  1928 

Exhibit 58	Dr. Morfeld's testimony			  PAGEREF e56_64mr \h  1928 

Exhibit 59	Mr. Scott's presentation			  PAGEREF e56_64mr \h  1928 

Exhibit 60	Mr. Scott's testimony			  PAGEREF e56_64mr \h  1928 

Exhibit 61	Mr. Bailey's presentation		  PAGEREF e56_64mr \h  1928 

Exhibit 62	Mr. Bailey/Mr. Lee testimony		  PAGEREF e56_64mr \h  1928 

Exhibit 63	Mr. Waggener's presentation		  PAGEREF e56_64mr \h  1928 

Exhibit 64 	Mr. Sessions' presentation		  PAGEREF e56_64mr \h  1928 P R
O C E E D I N G S

(9:34 a.m.)

		JUDGE SOLOMON:	 We're on the record.  I'm Daniel Solomon.  I'm an
administrative law judge with the United States Department of labor.  My
address is 800 K Street, Northwest, 4th Floor, Washington, D.C., and zip
code is 20001-8002.

		This hearing is taking place at the Department of Labor.  The first
group is the International Union, UAW.  But, Ms. Ryder, would you enter
your appearance for the day, please?

		MS. RYDER:  Yes.  My name is Anne Ryder from the Solicitor's Office.

		JUDGE SOLOMON:  Okay.  Is there any business -- anything I should
know about before we proceed with the panel?

		MS. RYDER:  No.

		JUDGE SOLOMON:  So the instructions for today, I guess, will be that
the panel will have 90 minutes.  And then there'll be some questions
from the public, and then OSHA will have a questioning.

		And, hypothetically, we'll be able to break for lunch at 12:00.  

		Out of necessity, sometimes we've had to change that a bit.  And we
have a long panel this afternoon, so we'll see how things go.  So who's
going to do the introduction for the Union?

		MR. COMAI:  I'll do that.

		JUDGE SOLOMON:  Okay.  Would you state your name, please?

		MR. COMAI:  My name's Andrew Comai.

		JUDGE SOLOMON:  And what is your position?

		MR. COMAI:  I'm the Coordinator of the UAW Health and Safety
Department out Detroit.

		JUDGE SOLOMON:  Okay.  Do you want to introduce everybody first?

		MR. COMAI:  We could do that.  So we have a group of workers, experts
I call them, from around the country.  These are people that work with
silica on a daily basis, work in industries.

		We have Andrew Mercer, sitting to my left, Recording Secretary, UAW
Local 8, Sparta, Michigan.  He's going to describe hazards and small
foundries.  

		Jeff P'Poole from UAW Local 523, Calvert City, Kentucky.  He's going
to describe the method for making silicon metal out of granite with an
electric arc furnace reduction process.

		Stan Burkeen, also from UAW Local 523, is going to talk about health
effects experienced by workers in that industry, his own personal
experiences, as well as a case that we would like to describe.

		Rodney Graves comes out of UAW Local 2317, Lafayette, Indiana.  He'll
be describing hazards faced by skill tradesmen, who 56 hours a week are
cutting brick, rehabbing furnaces, using refractory mineral fiber as
insulation.

		Greg Essex, UAW Local 226 out of Indianapolis, Indiana.  He's the
bargaining chair.  He's going to discuss the health impacts of working
around foundry silica and how workers believe that this is contributing
to adverse health effects and early death for many retirees.

		Shawn Ragle comes to us out of Local 974 Mapleton.  He spent 20 years
maintaining ventilation systems in a major foundry in Mapleton.  He's
going to describe the needs that he sees in terms of improving
ventilation systems.

		Robert Hitchcock, UAW Local 211 out of Defiance, Ohio.  They're going
to describe what we consider a model program for industrial hygiene, for
air sampling, for involvement of workers in solving the problems around
silica exposure.

		Richard Boecker is the full-time safety rep out of Local 211 in
Defiance, Ohio.  He's going to describe that industrial hygiene
sampling, his role as the full-time safety rep, and what they do to
reduce exposures.

		Finally, we've got Matthew Wafford.  He's a steward out of Local 2339,
Rushville, Indiana.  He's going to describe sort of the exposure in a
major foundry and his need for better air monitoring, better education,
and certainly better ventilation.

		Dr. Darius Sivin, sitting to my left, will also be talking.  We're
going to roll through, in the order that read those people.  We're going
to roll through the slide show and give you an efficient 90-minute
presentation.  Then we'll be free to ask questions, I think, or answer
questions.

		JUDGE SOLOMON:  You're going to go first, Mr. Comai?

		MR. COMAI:  Yes, I'm just going to run the PowerPoint, and I'll
introduce Andy Mercer here.

		JUDGE SOLOMON:  Okay, go ahead.

		MR. COMAI:  So, again, we've got three workers who are going to talk
directly about exposures faced by workers.  Andrew Mercer, again
Recording Secretary for Local 8.  And then, we're going to do a musical
chairs, and Jeff P'Poole will come up, and then Rodney Graves.  And Stan
Burkeen is also in there, too.

		So Hazards Faced in Small Foundries.  This is a plant that has a
casting operation.  We've got a couple of short videos that people can
see.  We'll try to step through this, run it once.  

		So this is a casting operation in Sparta, Michigan.

		MR. MERCER:  Good morning.  My name's Andy Mercer.  I work at Local 8
at a small foundry in West Michigan.  We represent about 180 workers,
and this here is one of them. 

		He's working in a shakeout operation, removing hot sand from the
freshly cast iron.  What you're going to see here is this machine's
shaking.  Now he's going to pound on the cast iron screw, which is --
not only the sand is leaving, but he's stirring up more dust that way. 
And then he's going to transfer it to a blast machine.  

		JUDGE SOLOMON:  Let me just say.  The problem with video is that we're
making a written transcript here.  So the issue then becomes how does
one describe what is in the video in order to reduce it to writing.

		MS. RYDER:  This will be made available as part of the record and
available for viewing in the docket office.

		JUDGE SOLOMON:  We haven't admitted the -- whatever --

		MS. RYDER:  I think they submitted the video.  I believe we have a
copy of it and --

		DR. SIVIN:  If it hasn't been officially entered into the docket yet,
it's our plan to do that today.

		JUDGE SOLOMON:  Okay.  Okay.  Could we write a written description and
enter that later?

		MS. RYDER:  I think the video is fine.  That'll be a part of the
record.  It'll be made part of the record for anyone to look at.

		JUDGE SOLOMON:  Sorry to bother you.  Go ahead.

		MR. MERCER:  The operator on this particular machine, as you can see
with the dust coming up and being stirred up by all the commotion, and
also, an air intake is going to blow air down to try to cool these guys.
 

		The operation requires two hands for safety purposes.  That prevents
him from pushing one button and reaching through the safety guard.  So
he's obligated to stand in that exposure area and not be able to walk
away from it and just let it run on its own.

		And you'll see sand spill out, freshly shaken out, coming off.

		MR. COMAI:  And that's essentially a hole in the dumper where the --

		MR. MERCER:  Yeah.  This here is -- this is actual casting.  It's
freshly removed from the sand.  Still hot, still has loose sand attached
to it.  He's starting the next process to run while he finishes up the
first one there.  

		So he's going to hand-strip these piston rings off this metal screw,
stirring up dust right in his own personal space as he does it.

		And then he's going to put this scrap screw onto a shaker that's going
to vibrate its way down to a separate area that's going to sort it out
and store it for reuse.  There's going to be various tumbling and
different operations, all putting particles into the air.

		MS. RYDER:  I just wanted to say that video was 4049.  Maybe just say
the title of each video before you start describing it.

		MR. MERCER:  So where are we at?  This is -- these are both the same
equipment.  This is originally set up without ventilation, this area
right here. 

		All that dust you see in the video accumulates.  This is where the
operator was working controls.  And then you've got several of these
lines all the way down.  This is two different perspectives.

		This is what was added in after the fact.  It's actually pulling some
of the ventilation away from there, the dust particles away from there.

		MS. RYDER:  What's the name of that piece of equipment?

		MR. MERCER:  That's a -- the first part of the operation, where he
removed the model, is called a shakeout.  It's a standard foundry
operation. 

		This over here is a shot blast machine that's going to use fine steel
shot.  And it's going to cycle around and clean it off.  That all has to
try to be captured.  Some of it's reused, some of it's collected
outside.

		MR. COMAI:  If I could just add, there's a particular issue with the
direct contact.  If I can get my video to run again.  So he's actually
touching the rings.  He's physical stripping those off the screw.  So
where you have the foundry sand is in direct contact with the molten
metal at that surface, at that location where the metal hits the sand is
where you have the generation of more crystal and silica.  

		As you're stripping that off, consider the fact that is the business
end of the silica.  That's where the silica dust is of most concern.  

		That screw then goes into a shaker conveyor.  There's no real
ventilation there.  So, again, you get a generation from there.  From
the shaker conveyor, it goes into a large tumbler.  Again, an
unventilated tumbler that generates dust as the screws and the waste
product are cleaned, as well.  

		MR. MERCER:  The one thing that we tend to notice is not so much the
feeling you get to walk into a room and be overexposed.  A lot of people
notice with time away on vacation and whatnot, a clearing of the lungs,
a less shortness of breath that's kind of in the day-to-day routine of
it all.  

		So in our opinion, there's definitely an effect to this from long-term
exposure.  I've been in it 19 years.  I don't feel the same as I did. 
And you can attribute some to age and weight gain or whatnot, but you
can't help but wonder if it's from all the years of that exposure right
there.

		I'm here because our local chairman was originally slated to come
down.  He was just recently diagnosed with a brain tumor, so he's out
dealing with that and chemotherapy and radiation and all that stuff.

		My grandpa retired.  He worked very near that line right there.  His
two brothers, they all worked the same period of time.  My grandpa
worked in that operation.  The two brothers went into machining and
mechanics, different parts of the building.

		My grandfather passed away from cancer.  The other two: long, healthy
lives after retirement.  Definitely things that keep you thinking about
what you're being exposed to while you're at work.

		MR. COMAI:  So next up, we're again going to shift seats here.  Jeff
P'Poole, President of Local 523, Calvert City, Kentucky.  He's going to
walk us through a process, smelting silicone.

		MR. P'POOLE:  Hello.  As Andy said, I'm Jeff P'Poole, Local 523,
Calvert City, Kentucky.  We run one of the largest -- three electric arc
submerged furnace operation in America.

		What we do, we get silicone rock in.  It comes in on barges or
railcars.  That's some of our exposure there, the silicone rock.  

		We take that rock and had to handle it two or three times before it's
finally dumped into the furnaces.  

		What you're seeing there is the furnace.  It drops from the top down
into there, which you could have some dust come from that.  Then, we tap
it out into ladles and pour it.  When we're doing this, there's a lot of
dust that's up front.  

		We've got another slide on there that'll show the dust that's up at
the top of the building that is probably left behind.  

		Well, the granite rock, we get rock that's mined in, I think it comes
out of Georgia, most of it.  When it comes in, if it's dusty, our loader
operators, when we're unloading the rock cars and stuff, will dump water
on it to try to keep the dust down to protect our workers.  

		We've got a belt.  When it goes in to be fed to the furnace, there's
sprinklers on the belts to keep the dust down.  There's a reason for
that.  It's to keep the silica dust off of us.

		When it's processed, we make a product called 75 high purity.  At
times, they mix in sand and lime in it.  

		Once we process it, and it comes like molten metal, like a volcano
exploding is what -- how I look at it.  They skim those impurities off,
so those impurities come a lot from that sand and lime.

		We take it over to the ship inside and crush it down, which produces
more dust in our air, dust particles in our air.  And one of the main
things that, I guess, the reason that I'm here or we're here is we
supply most of these fellows behind me with our product, and what they
take our product and make, I think they get exposed too.  

		We start it.  Maybe the mining of the rock starts, and then we're just
part of the process.  

		This is upstairs from our building, like our walks that we walk up. 
And you can see like the beam right there, and he's got the green light
on the beams.  It's just the dust that's in the air that our process
makes.  That's ash and dust.

		Okay, and that's more right there he didn't -- you can just see dust
particles in the air without a flash on the camera right there.  What we
do, too, with our metal, after the whole process is done, the stuff
that's on those beams gets taken out to our baghouse.

		Now, as you can see, it doesn't take all of it out.  But we've got men
in the plant that we sell the silicone fume.  The fume is sold and put
in railcars.  I think they ship it out, and they put it in concrete and
other type of things.  But we do sell that.  And we do also sell, when I
said the impurities, when we skim the impurities off the metal, it's
silicone dross.  And we also sell the dross.  That has part of the
crystalline silica in it, also.

		Okay.  This picture here is when we don't take the stairs, the
building, you know, we've probably got -- I probably couldn't -- how
many stories, Stan?  It's like a six-story building that you go to the
top of.  And you can take the stairs or you can take these man lifts. 
And you take these man lifts here.

		And when you get off at certain levels, you can see the dust that's
left in our building and what we work in there.  You've -- we've heard
people come out with like raccoon eyes and stuff like that.  When you're
in there, and you're doing your job, you'll come out and you'll look
like a coal miner at times when people are getting cleaned up to go
home.  

		I just mentioned whenever you're off for a week and things, your nose
is full of black stuff.  Once you come off of vacation, you come back
and the whole process starts all over again.  

		If there's one thing that I've told Andy and stuff, that I hope that
we could get out of it is more testing of the air.  I asked our health
and safety girl at the plant to give me the readings from our last air
test.  And they were from 2011.

		And I would just hope that we could probably test the air more often. 
Thank you.

		MR. COMAI:  We're going to shift seats again.  We're going to call up
Stan Burkeen, Trustee.  He's a former bargaining chair at Local 523 in
Calvert City.  He's going to talk about his health issues and health
impact on workers at the plant.

		MR. BURKEEN:  Yes, thank you.  My name is Stanley Burkeen.  I am a
Trustee at UAW 523, former president, prior to Jeff.

		I've worked at CCMA in Calvert City 523 for over 20 years.  My father
and several of his brothers before me worked there since the plant
opened in 1950.  My son works there now.

		There's been a lot of improvements over the years, through the
pollution, through work between the union and management to try to make
the air cleaner.  But it is a dirty and dusty place.  As Jeff said, I've
left many a day and looked like I'd come out of a coal mine.

		What I'm here to talk about, there was a young man that I hired in
with 25 years ago.  Tom was diagnosed with silicosis.  All the years
I've been there, that's the first I'd ever heard of it.  I know nothing
about it.

		Tom, his breathing got worse.  He took off work.  He was diagnosed at
Vanderbilt Hospital in Nashville with silicosis.  He filed a workers'
comp claim.  He was denied.

		The next step was to hire a lawyer.  He did that.  He went to the
University of Kentucky Medical Center, where they said he was
misdiagnosed.  Tom didn't live very long after that.

		We don't know if he died from silicosis, if he died from some other
lung disease because of the, I guess, inability to have the proper
testing to see what he had or the testing of the air in the plant.  We
feel like, if nothing else, if we had stronger standards, better
education of what causes it, had a test for it, things like that
wouldn't slip through the cracks, I hope.

		We have a very young group of people in that plant now.  We've had a
big turnaround the last few years.  I would hope that we take better
care of them.  My son's one of them.  I would hope that we'd look out
for them and take better care of them than we took care of our retirees
before them.  I guess that's about all I've got to say.  Thank you.

		MR. COMAI:  Thanks, Stan.  Just to point out, there's a number of
health studies on the silicon smelting industry that we'd also like to
enter into the record.  Just to cite them here, and we'll get you those
PDFs.

		DR. SIVIN:  We'll enter them in post-hearing comments.

		MR. COMAI:  Next up is Rodney Graves.  He's out of Local 2317,
Lafayette, Indiana.  He's a skilled trade bricklayer, works 50, 60, 80
hours a week cutting brick and rebricking ovens.

		MR. GRAVES:	Hello.  My name is Rodney Graves.  We run one of the
largest heat-treat facilities in North America under one roof in our
facility.  We have five gas car furnaces, which a small one would be
roughly the size of this room.  Then, we have, as you can see in this
photo, this is a pit furnace that long shafts go down into.  On the
left, that's a liner that's inserted.  

		As you can see on this picture, this part right here, that's the brick
area that we repair and stuff.  This, towards the top, they get so hot
that this starts to pretty much become dust.  Even when the operators
and stuff open these lids to change the parts out, there's a -- this is
about three foot in diameter.

		Down below, there's a fan in there that stands about a foot tall and
about two and a half foot in diameter.  It blows all this stuff up
through and out, so all that dust they're breathing every time they
unload and load these.  

		We do a lot of work inside.  This is actually an open one.  The gas
carves all cases and annealers, furnaces.  Them are all enclosed. 
There's where we run into the biggest hazards because we check the air
quality to clear the machine before we go in, to make sure the CO2
levels are right and everything.  

		Then, once we go in, there's no ventilation because they're so tight. 
Generally, when you're in there laying brick, you're sitting down.  The
wall thicknesses and roof thicknesses and bottoms are the minimum two
foot.  You're in there, curled up because you've got heating elements
above your head and stuff.

		You're sitting there, while you're repairing walls that might have
ripped out half the wall.  You take a handsaw, and you cut these brick. 
Well, it gets so dusty in there, you have to take off your glasses and
stuff so you can see.  It just fills up.

		You have this stuff all over you.  I think he's got some more
pictures.  The picture on your left is just a place where we make bungs
and stuff that go around the heating tubes.  That's all in the open.

		On the right, that's our saw.  As you can see there, it has water in
it.  That's good.  We can have water in it when we cut the hard tile. 
But when we cut the dry brick, we cannot use water because too much
moisture is absorbed in the brick.  When we lay them in the wall, we
fire the furnace it, it'll cause an explosion from all the water in it. 
So we can't do that.  But the hard brick won't absorb the water. 
There's just another picture of the saw.

		MR. COMAI:  So the controls for this would be enclosure, a good
exhaust ventilation.  There's some good radial saw ventilation diagrams
in the ACGIH manual that we refer to, so we can reduce that.  

		MR. GRAVES:  Yeah, currently, just to speed up the process, and we
don't have the manpower we used to have.  They've been bringing these
saws up into the actual production area right next to the machine and
cutting large quantities of dry brick, which just throws that dust
throughout the whole shop.

		And then we don't have very good ventilation as it is.  The
ventilation we do have just recirculates it all over heat-treat.

		MR. COMAI:  Next, we've got Greg Essex, bargaining chair out of Local
226, Indianapolis, Indiana, Navistar foundry.

		MR. ESSEX:  Morning, everybody.  My name's Greg Essex.  I'm from
Indianapolis, Indiana.  I'm the President and Bargaining Chairman of
Local 226.

		I'm here to testify basically or speak on behalf of the workers at our
foundry in Indianapolis.  We are primarily a gray iron casting foundry. 
It's been in existence since World War II.

		We've done a lot of work.  Most of you might know the name
International Harvester, back in the ag days.  We produced blocks and
castings for them.

		Since then, we've done stuff for the military, and Ford was a big
customer.  Over the years, with my military history, working for the
railroad the last, and now this place, for 14 years, I've seen a lot of
exposure to all kinds of elements.  I have really no doubt in my mind,
in my personal opinion, that these exposures are contributed to health
concerns.

		I have a list here that I owe to my, I guess, elders and/or deceased. 
To mention their names today because they've all been subjected to these
exposures over the past 40, 50 years.

		Jack Pollard, cancer of the lung; Jim Chamberland, Burkitt lymphoma;
Mike Driver, cancer; Jim Galladay, cancer; William Canler, cancer; Jim
Pascul, cancer; James Nobel, cancer; Tom Gloid, cancer, lung transplant;
John Pugh, cancer; Russell Bebe, cancer; Bill Farmer, cancer; Ray
Kraft, cancer; Phil Ray, cancer; Carol Ramsey, cancer; Carl Stamps,
cancer; Walby Shuck, cancer; Doc Mitchell, cancer; Rick Hale, cancer;
Randy Dant, cancer; Charlie Rogers, cancer; Ray Turner, cancer; Gene
Jones, cancer; Les Everhardt, cancer, Dick Stanford, cancer; and Dick
Holmes, cancer; Jess Jennings, cancer; Doug McBane, cancer (all phonetic
spellings).

		We have six individuals who are suffering from cancer today that are
still surviving, obviously going through some terrible pain and medical
conditions. 

		MR. COMAI:  This is a 2006 video.  It's gotten better.  Describe
what's going on in the video.

		MR. ESSEX:  This is our shaker system.  The shaker system is about the
length of a football field.  It winds like a snake.  This is where the
castings are dropped out of the molds.  This is where they cool.  The
shaker is just like a vibrating system.

		The castings move on down the line.  You see these tubes here and
there.  This is what we refer to as the covered wagon.  This actually
covers the shaker in this particular area.  The ventilation tubes pull
out the sand and dust particles, and the dust. 

		That is, like I said, the length of a football field as these
particles accumulate.  Then, it's not always been like that.  Some of
these guys I know personally, like Bob Truex (ph.), he died.  I seen him
dwindled away from a 5'6" man of stature, about 180 pounds, to 88
pounds.  He was diagnosed in November of 2005.  Cancer took him within
six months; he died in April of 2006.  

		Like I said, he told me stories.  You couldn't see from here to the
back wall in that foundry 40 years ago.  There's been significant
improvement in at least the last 20 years.  I've been there 14, so I've
seen significant improvement in the last 14 years.

		This is also part of the covered wagon.  You see the dust.  You see
the glare off the light.  That's all dust.  Here's piles of sand and
dust.

		MR. COMAI:  So we're trying to get a shot of the backside, failure to
maintain the system, failure to maintain the ventilation.  We've got
major sand leaks on the ground here.  

		MR. ESSEX:  What the previous guy had mentioned, walking out of there
looking like a coal miner, I've seen that in my last 14 years I've been
there.

		MS. RYDER:  Just for the record, I want to know the first video you
showed, was that 4042?  

		MR. COMAI:  Yes.

		MS. RYDER:  Then you showed another video.  Did that have a number?

		MR. COMAI:  Not on the screen, no, it didn't. 

		MS. RYDER:  Was it titled anything?

		MR. ESSEX:  These shaker systems have sections.  There might be a
section from here to that wall.  They'd number them, so they know when
they have to go change a spring or an arm, what section they need to go
to. 

		Like I said, it's 300 yards long.  These things can be anywhere in
length from 12 to 30 feet in one stretch.  They number them
specifically.  You'll hear a bunch of different terms in here.  We all
pretty much work in the same environment.  Even though we have the same
processes, same type of equipment, we all in different areas of the
country have named them nicknames.

		MR. COMAI:  We submitted these videos, 4042, this one is 4043.

		JUDGE SOLOMON:  42 has headings on some of it.  Just for the record,
can you state what the headings are?

		MR. COMAI:  So video 4042 is the shaker conveyor.

		JUDGE SOLOMON:  After that, the next one that you showed.

		MR. COMAI:  This is the same system, covered wagon.

		JUDGE SOLOMON:  And after that?

		MR. COMAI:  The number on the covered wagon video is 4043, the ones
that we submitted.  Then, this is 4044, exterior sand spill.

		JUDGE SOLOMON:  And that says casting sand spills.

		MR. COMAI:  Casting sand spills, and we're up to 4044.

		MS. RYDER:  Whenever you start a new video, do you want to just say
what number it's going to be?

		MR. COMAI:  Sure.

		MR. ESSEX:  This is where the sand comes out of the shaker, and you
have the conveyors that take that sand away and take it to a place where
it is usually put in a dump truck, tarped over, and taken out.

		This is the 13-belt that goes outside.  It's part of the process. 
That's at the tail end of the shaker.  The 13-belt takes it up, out into
the yard, and makes it out and ends up in that dump truck.  That's where
you see the big pile first, right here.

		MR. COMAI:  So this is ground level.  This is a 2006 video
demonstrating large piles that are falling through leaks in the sand
conveyor system on the roof.  When this happens, workers have to get up
there with a shovel and a wheelbarrow. 

		Can you describe any respiratory protection they have when they do
that?

		MR. ESSEX:  Respiratory protection, I think it's an educational thing.
 Some of these older guys weren't educated on, and I don't know that
they were aware of, the complications of some of these particles could
have caused.  Otherwise, I might not have had to mention their names
today.  I thank them for paving the way.

		JUDGE SOLOMON:  Again, this is video 4045.

		MR. ESSEX:  This is slag in the furnace.  That iron in that furnace is
about 2600 degrees.  You can see the smoke.

		Like I said, this foundry is about 70 to 80 years old.  You can
imagine the dust particles that have collected throughout the I-beams in
the structure over the years.  You can see the smoke coming off of it,
obviously.

		Like he said, this video is from 2006.  We've since implemented
procedures where we wear face masks to protect our face.  That looked
like a good operation there where there wasn't a significant amount of
sparks that fly out of that furnace.

		Since then, we put face masks, hard hats, and they've made significant
improvements in recent years.

		DR. SIVIN:  Dr. Selikoff, whose work established the hazards of
asbestos, once said that "Statistics are human beings with the tears
wiped away."  We've heard a lot of the human stories behind the
statistics. 

		I just wanted also to add that Stan Burkeen's story illustrates what
has been talked about a lot in these hearings about misdiagnosis and
under-diagnosis of silicosis.  

		JUDGE SOLOMON:  I don't want to butt in, but just for the court
reporter's benefit, you are Dr. Sivin.

		DR. SIVIN:  I'm Dr. Darius Sivin with the UAW Health and Safety
Department.  In any case, OSHA, in its Federal Register, asked
commenters to comment on whether 50 µg was the most appropriate
permissible exposure unit.

		OSHA's own risk assessment indicates 43 excess deaths from
nonmalignant respiratory disease, 18 to 26 excess deaths from lung
cancer, and 32 excess deaths from kidney disease per 1,000 workers at
that level, which is a total of 93 to 101 deaths per 1,000 workers
exposed, approximately 10 percent.

		We believe that this is unacceptably high.  Since the OSH Act itself
says to the extent feasible, no worker shall suffer material impairment,
we believe that in those industries, where it is feasible, we strongly
support OSHA's regulatory alternative of 25 µg/m3.

		Obviously, there are some industries where that may not be feasible. 
For those industries, we strongly support OSHA's as proposed.

		MR. COMAI:  I'm going to introduce Shawn Ragle.  He is, again, a
skill trades ventilation repair and maintenance.  He's been working for
20 years, fixing ventilation systems that exist in the Mapleton foundry.
 I've got a couple of other people that will be talking about the
feasibility of ventilation and the feasibility of reducing to that 25
µg level.

		MR. RAGLE:  Hi.  My name's Shawn Ragle.  I'm the Caterpillar Mapleton
foundry Health and Safety Chairman for the UAW Local 974.

		We represent about 482 UAW employees.  There's also about 50 contract
employees in the building that are potentially exposed to silica. 
Workers are exposed to silica contained in airborne dust produced by
transporting on flat belts, bucket elevators, leaks in sand
transporters.

		The dust collects and falls from overhead light fixtures, building
columns and overheard walkways.  We've seen in the video all the dust
that can collect on the beams is what I'm talking about.  In some of the
areas, it gets so dusty that the dust, as you walk, gets kicked up, if
there's that much dust. 

		What I do there is a repairman, as Andy said.  Some of my work
involves EPA-compliance inspections, maintain dust collectors and
ventilation fans.  When they put in new ventilation or dust collection
in our plant, I would appreciate the opportunity to have some input into
it, at least kick around some ideas.  This never happens.

		When dust collectors go down in an area that it draws from, typically
the work doesn't stop.  The work will only stop if the employee raises
his hand and says there's something wrong.  Then, we investigate it. 
Meanwhile, we're still running.

		In addition to that, we've got jobs that really put a worker at a
higher potential to have an inhalation of this dust.  For instance, like
a core making specialist, that's where they make the cores that they sit
inside the molds.  When the core box opens, they'll blow it off of
there, and you'll have a potential for dust there.  Sand mix specialist,
that's where they mix the sand that's being transported.  We've got a
lot of potential there.

		Maintenance is another area.  Maintenance don't do a lot of standard
work.  It's mostly nonstandard.  They're going in places where typically
most people don't go.  That's stairwells that have all this dust on it
working in the rafters.

		 Another thing is back up to the EPA compliance.  Caterpillar seems to
be really good about being concerned about what they put out their
stacks into the environment.  When it comes to inside the four walls of
that plant, sometimes I think we're a little lax on what we see there. 
I would like to see some improvement there.

		We've got like a mold line specialist.  What that guy would typically
do, these shaker cables that we're talking about, where they transfer
the castings, and they want them to ride on a layer of sand.  You'll
have your castings.  A lot of them are not covered.  The dust will rise
up off of there.  It just collects.

		You can get this dust on you.  It could be totally accidental.  You'd
just be standing in the wrong spot.  Somebody does something, knocks it
off a rafter on you.  There's just a lot of dust in a foundry is what it
amounts to.

		I had an employee that raised his hand.  He was in the pit.  He runs
heavy equipment wheel-loader.  It's got an air-conditioned cab.  There
was an area of the pit where these castings are conveyed down a shaker
table.  They take the casting itself and separate it onto another
conveyor.  The tramp iron, the screw, the gating, the stuff that is not
a product, they allow it to fall into a shoot.  Then the sand also
follows another path down into the pit where it falls.

		Now, this stuff falls a good 30 feet into a pit.  When it's falling,
they have very limited dust collection on that.  They've got a dust
collection at each end, but the center's open.  When it hits the floor,
there's nothing there.  It just plumes up, and it goes everywhere.

		We filed a safety complaint on that.  The company chose to look at the
operator, put the monitoring on the operator instead of an area sample
at the time.  Well, the operator raised his hand and said the complaint
is over here on the floor, not in the cab of my tractor.  

		Obviously, the readings that we got back, the results we got back,
reflected that it was pretty clean down there.  That really isn't the
case.

		MR. COMAI:  Yeah, the air sample pump was placed on the driver of a
vehicle who was in an air-conditioned cab for the remainder of the
shift.

		MR. RAGLE:  That's all I got.

		MR. COMAI:  Next up, we have Robert Hitchcock.  He's the bargaining
chair out of Local 211, Defiance, Ohio, another foundry.  He'll be
talking about the fact that UAW at General Motors has negotiated a lower
standard, a lower exposure limit.  We protect workers to that limit.

		MR. HITCHCOCK:  Good morning.  Yes, I'm Robert Hitchcock.  I'm the
chairman at Local 211 UAW for General Motors in Defiance, Ohio.  I tell
you, just in the last 24 hours, I got to meet these fellows sitting
beside me.  Their stories are just amazing.  I just can't believe some
of the things that they're put through.  Something that we, here in
Defiance, take as a norm is something that they are obviously striving
for.

		If I wasn't appreciative of where I work, if I hadn't been, I
certainly am today.  I truly do believe, though, that the UAW and
General Motors at our facility does a fantastic job ensuring the safety
of the workers out there.  We do that in various ways.

		More importantly, we do have a very structured and committed safety
plan at our plant, both union and management.  We have, over the years,
developed a very good relations, that we can sit down, we can identify a
problem, and we can work through it. 

		Not every problem is cheap to work through.  There is expense to it. 
At the same time, when you have two parties striving for the same goal,
it does tend to be easier to accomplish the goal and make changes for
the welfare of the people, and you know really for our namesake. 
Whether it's UAW or whether it's General Motors, we are proud to be who
we are.  I'm very fortunate for that.

		Now, even though I've described our plant as being a very
safety-conscious plant, I will say that it's not always been like that. 
I am the third generation that has worked at this plant.  My
grandfather, back in the '50s and '60s, died at a very young age of lung
disease.  I believe he even had some heart issues.

		I also had my father, back in 1994, die at the age of 49 from cancer. 
Very difficult time.  In fact, during some of his most painful times, I
was in the U.S. military, never thinking I was ever going to come back
to Defiance, Ohio, but I did.  I guess fate brought me back.

		I ended up getting a job at General Motors.  I started working there. 
I have been very fortunate to be part of the UAW.  I can think back, and
I remember as a child, meaning a teenager, where my father had attained
the permission to take me on a small tour of the plant back in the late
'70s, early '80s.  

		I remember walking down the main aisleway there at the plant, right by
the cafeteria.  You could look down and you could see your footsteps. 
That's how dirty and dusty, with the silica sand that we use that was
just infested all over our plant.

		Now, over the years, and correct me if I'm wrong up here, but I
believe in the early '90s, somewhere around '91 or '92, the UAW
negotiated, along with General Motors, to exceed the standards set by
OSHA, their standard of being 100, down to 50.  Since then, the UAW has
been very aggressive in taking advantage of going out, identifying areas
that were over that limit, and working collectively to correct those
areas.  That is exactly what we have done.

		Now, through this commitment and bargaining, basically what it did was
it provided a commitment that both parties were committed to the folks
working out there.  They were committed to their health and safety, not
only for air quality, but for other various health and safety issues.

		We made tremendous improvement on ventilation, better exhaust.  You
have heard some of the folks talk about dust collectors.  We have them
all over our plant.  

		We have also instituted a very effective preventive maintenance
program, so that we can identify up front that, hey, something is
starting to go bad, so that we can keep an eye on it, so that we can fix
it prior to it going down and affecting the people.

		We also have areas that are on that edge of 50.  We have to keep close
eye on that.  We have got an industrial hygienist at the plant full-time
that continuously checks and monitors these areas.  Some of the things
that we've had to revert to in some cases was even an air-cooled
respirator.  We've taken that extra step because the folks are working
out in these high temperatures, the environment is not very good.  To do
something that would make the employee more comfortable, more willing to
want to do that job, we've done things like that.  

		Just recently within the last year, since I've been the chairman here
in the last three years, we actually are doing some experimental tests
with bringing in a unit that they actually use in NASCAR in the
racecars.  What it does is there is a unit that actually conditions the
air that is provided, that goes up through the hoses into the helmets
for the workers that makes it cooler.  It actually provides more airflow
for them.

		We don't have near the issues that we may have once had from heat
stress or heat stroke.  At the same time, by having the conditioned air,
they are not breathing the silica dust that surrounds them.

		Even though our plant isn't perfect, we continue to try to make things
the best that we can.  I think we've been very successful with that,
from the commitment from the UAW and management, that you can't just
settle on something as a norm.  If there's something out there that's an
approved, better way of doing something, we are certainly going to look
at it.  We're going to investigate it.

		We're going to bring something into the plant.  We're going to try it.
 We may try it for a month.  We may try it for two months.  Based on the
results, based on the people's reaction to it, we may decide to keep it
or we may decide to start looking somewhere else.  We continue to go
down that path.  I'm very fortunate for that.

		Today -- I described what it was years ago, where you could see your
footsteps.  Today, you can walk down in our main aisle.  You can walk
out into our plant.  In some cases, you wouldn't even think you're in a
foundry, to be quite honest with you.

		Now, I'm not shortchanging nothing.  Like I said before, we still do
have a few areas that we struggle with, but we keep it closely
monitored.  We continue to seek improvements.  We will continue to do
so, at least as long as I'm there.

		There is a very good commitment.  I greatly appreciate the UAW
bargaining that lower standard and the commitment that General Motors
has made to it.  Is there anything else you'd like me to talk on?

		Recent ventilation improvements, I'd say within the last five years,
we made some pretty significant investments out to our plant.  We
actually identified through testing, through velocity checks and what
have you, where we've had some ductwork that's been plugged.

		What happens is that as you melt iron and go through a process with
the hot sand, we actually get a creosote buildup in our ductwork.  You
wouldn't necessarily notice it because you can't see it within the
ductwork.  What you do notice is the restriction of airflow.

		Through the testing, we have identified that this buildup is in there.
 Creosote is very difficult to remove.  What we've had to do in a couple
of areas is we've just gone in and actually just cut out the entire
ductwork and ran new ductwork.  We've also had areas where we've had to
take dust collectors and make adjustments to the fan blades, change the
sheaves on them to increase the airflow.

		It's just a continuous project, is what I would call it.  Something
like I mentioned earlier, that we monitor closely.  And when we start to
see the readings go down, well, then we're reacting.  We've got some
really good people in place to address that.  

		I have along with me here today our Health and Safety Rep at Defiance,
Ohio.  His name is Rick Boecker.  Maybe he can shed some more light on
what we do at Defiance.  Thank you.

		MR. BOECKER:  Good morning.  As Rob said, my name's Richard Boecker
with Local 211, Defiance, Ohio.  I am the UAW Safety Rep and electrician
by trade.  I have worked on the floor, so I'm familiar with some of the
hazards from being a skilled tradesman, as well as a safety rep.

		Some of my comments may be a little redundant with what Rob said, so I
apologize for that.  As Rob said, in Defiance, we are kind of a model
for these gentlemen behind us.  There is sampling we do, and what we do,
as far as how do we protect our people, that comes from years of
originally tough negotiations.  

		Now, what I see with union management relations at Defiance, it's more
of a joint effort to take care of our greatest assets, which our people
are.  I'll try to paint a picture here a little bit of some of the stuff
that we do jointly that was originally negotiated.  

		As a Health and Safety Rep, I'm part of our local joint Health Safety
Committee, which I have a counterpart on management that is also present
on that committee.  We have industrial hygiene, who is full-time.  He
has a management counterpart, as well.  We also have joint ergonomics
rep.

		With these resources in places on a full-time basis, we have the
ability to address concerns and complaints immediately.  Some of these
guys don't function full-time.  We've got an unfair advantage where we
can address it on the spot.

		As I said, our full-time industrial hygiene rep, that is really
important.  I wish all these guys had that opportunity.  Each year, our
joint industrial hygiene representatives from management and UAW submit
a written proposal of an air sampling plan that we're going to do
throughout the plant.

		That's approved by the International parties.  We do deviate from that
on a regular basis.  If we got complaints that come forward, we address
them quickly.

		Another thing that we just started -- as Rob said, he is our shop
chair.  He's been instrumental in bringing some of this stuff forward. 
He helped us start regular joint meetings with key players in the plant,
with our local joint Health and Safety Committee.  

		We asked maintenance folks to be present, salary and hourly.  People
that have direct impact on how our equipment is operational, say like
bag changes, cartridge changes, that type of stuff.  The reason the
industrial hygiene is so important for air sampling, like the one
gentleman probably knows back here, on your dust collection system, they
do velocity checks.

		They'll do differential pressure checks across the filters, as well. 
If there's no differential pressure change, a lot of times, your bags
are blown or your filters have got a hole in them.  We can't always rely
on mechanical gauges.  That's why we need to do our air sampling.

		I guess to put a personal note on it as well, from myself, my dad was
shop chair when I was a kid.  I also went through that plant and seen
how dirty and filthy it was.  It really was a terrible place to work, to
be honest with you, very dirty.

		My father was very instrumental in getting some of these things
negotiated at the plant.  Like Rob's father, he passed away at an
earlier age, too, than he should have.  We just buried him about seven
weeks ago from multi-organ shutdown.  I assume some of them health
issues come from working at that foundry.  It's my goal to make sure my
children or your children can work in that environment and be healthy. 
Thank you.

		DR. SIVIN:  I'd just like to point out a couple of things.  Last week,
we saw some data presented by Dr. Frank Mirer, which suggested that in
the '70s and '80s, the Detroit Three foundries may have been worse than
some of the other foundries.  

		Today, we just heard the sequel to that story, whereby once we got
contract language, legally enforceable, we got improvements at the
General Motors foundries.  Similarly, if we get a legal enforceable OSHA
standard, we would expect to see similar improvements at all these other
workplaces you've been hearing from today.  

		I also wanted to mention that Dr. Mirer referred to the industrial
hygiene tech program.  You just heard a full description of it.  

		We just heard a little bit about the medical surveillance program at
General Motors.  I just wanted to add a little more from our testimony,
including we believe that medical surveillance should be triggered at
the action level.  Medical examinations should be provided in response
to employee reports of signs and symptoms related to silica exposure, as
well as to the action level. 

		It should provide that a physician or other licensed health care
provider can determine that it is medically necessary for a worker to be
removed.  In order to protect confidentially, it should be the worker
who decides where that information is conveyed to the employer.  We also
believe the standards should provide for multiple physician review.

		MR. COMAI:  Our last presenter is going to be Matt Wafford.  He comes
out of Local 2339 in Rushville, Indiana, another foundry worker.  He's a
shop steward in the molding department.  

		We went through an OSHA investigation together.  The company was
eventually cited for HazCom.  This is the way it went.  There was a
fatality in the plant.  We went through 15 OSHA interviews.  The OSHA
inspector was a good one.  He was a good inspector from Indiana.  I'll
go on record saying that.

		He asked us the following questions, every single worker, skilled
trades, lots of experience:  Do you work in a foundry?  Yes.  Do you
work with hazardous chemicals?  No.  Do you work in a dusty environment?
 Yes.  Do you know what's in the dust?  Well, it's foundry dust.  Has
anyone ever mentioned silica?  Yeah, everybody knows about silica.  

		He asked that series a questions.  That just really drove home to us
this idea that we need more training, especially in that plant, on
general safety, lack of -- confined space, but the dust was an issue, as
well.

		We want more training.  We think that's important.  Regulated areas,
signage, those basic levels of protection where workers can get
notification and take action to protect themselves.  I will turn it over
to Matt.  He's going to describe some of the work that he's done in and
around hot sand.

		MR. WAFFORD:  Hi.  I'm Matt Wafford from 2339 Rushville.  If you look
right there, there's files.  This right here are turrets that spin.  The
sand comes in from the belt, return sand.  That cools it to go back into
the other.  That gets buildup on it.  It gets buildup on the walls in
there.

		We've got to climb in this door right here and get in there and scrape
that off each night so it can be kicked out and get discharged. 

		JUDGE SOLOMON:  Just for the record, the title of this is "Sand Cooler
Cleanout."  

		MR. WAFFORD:  You've got this blower here, if you see that blows cool
air in to blow that.  It just builds up.  You're in there scraping it. 
Once you scrape, it poofs up, and you breathe that in your respiratory. 
You're coughing and stuff.

		Here's a picture of it.  See how it's all built up there and there and
on the walls?  This right here adds the water into it.  Then, when you
put the water in it, it's more likely to stick to you if it's wet when
you get in there.

		Here's a picture of it where you have the double turrets.  This one
spins this way, and this one spins this way.  It discharges it through a
door.  There's a door about right here.  There's another level where you
climb in to get into there.

		See how it's real thick?  You've got to get in there and scrape that. 
You're going to be breathing that when you scrape that.  There ain't no
ventilation.  Suck it out while you're in there scraping it in the air.

		Here's where the access doors used to be before the accident happened.
 If it comes on, you're there.  Now, they put these interlocks where
they're locked.  If the power's not off, that don't come off to get in,
which helps.

		JUDGE SOLOMON:  This slide is entitled "Before and After."

		MR. WAFFORD:  They've made a lot of improvements on access to stuff
since that happened.  You've got this guy down here and this table right
now.  This slide is named "Desprue."  This table right here shakes all
day long with sand.  

		You see the two guys there?  There's nothing on their heads.  They've
got ventilation right there, but it's not accurate.  There's dust always
flying.  You come out looking like they said before, like raccoons or
worse.  There's sometimes that table's full up to there with sand on it
when something gets plugged from ventilation.  

		No dust masks.  There ain't no hood or nothing.  There's ventilation
right there.  You see right there, there's ventilation right there that
sucks it.  It supposed to suck it right there, but it's just when it
goes down there.  All this other is exposed when it shakes.  It flies up
in the air.  It gets on you.  You breathe it in.

		They gave us new respirators that fit over your head.  It's got a
tube, and you hook it to an air hose.  There's like 10 of them.  None of
us use them because it's just too big and bulky.  It constricts you. 
You can't hardly move.  It limits limitation on your job.  It's actually
more a hindrance than anything.  It probably weighs three pounds.  You
get it on your head, trying to move around and do your job, and it's
just more trouble than anything.  What they need is something better
that will suck the ventilation, more training on what we're exposed to,
a better system.

		The last time we did an air sample, the paper I got was two years ago.
 They done it on two employees that were in desprue, but it was right
after they worked on the dust collectors, not everyday use.  

		We've got some employees there that have been in desprue for 15 years.
 I know he's been there, and him and another one, there's two of them,
one's been nine years.  They have the COPD.  They're hacking.  They're
coughing.  They're breathing.

		We've got an employee that works at the other end of the plant.  This
part is where they come off the machine, where they're already done. 
The lady that works where they make the molds, she's had stuff removed
from her throat because she's had buildup from -- I can't remember what
she called it.  She's had them removed, and they're back again.  Polyps,
yeah, that's it, from breathing the dust.  It's always dusty, and they
need better ventilation for the employees to breathe.

		Generally, 98 percent of us work 7 days a week.  On Mondays, we're
okay, but on Friday, we're all wore out.  We're tired.  We feel like we
can't go anymore.  To just keep breathing it, it just wears you down
constantly. 

		MR. COMAI:  We brought a hood along.  It's up in the coat rack, we
think.  I don't know if we can --

		DR. SIVIN:  We've already submitted for the record evidence that shows
that 25 µg is feasible in dental equipment manufacturing.  It didn't
get into the slide show, but it has been submitted for the record.

		JUDGE SOLOMON:  Okay, that was Dr. Sivin again.  Mr. Comai, do you
have anything else?

		MR. COMAI:  No, we're done.  Thank you.

		JUDGE SOLOMON:  Okay.  It's 10:46.  Do one of you want to go up and
get that hood?  Okay.  And so, you're just showing that to us?

		MR. WAFFORD:  Yeah, you've got to --

		JUDGE SOLOMON:  Do you have a picture of that or something that we
could use?

		DR. SIVIN:  We'll submit a picture for the record.

		MR. WAFFORD:  You've got to put it on your head, and then you've got
air to airline.  You're constricted in the way you move your head
because you've got this tube going down your back.  

		It steams up.  You can't see, so most of us just take it off and put
our regular hard hat back on.  These are just put in our lockers.  I
don't know that anyone uses them that they've gave them to.

		DR. SIVIN:  What does it look like when you put it on?

		MR. WAFFORD:  You want me to put it on?  This goes on your back.  

		MR. COMAI:  So we'll let the record show that he's steaming up
already.

		JUDGE SOLOMON:  Okay.  So could I get an idea of how many people
wanted to ask questions?

		Ms. Seminario, you're in the first row with the people who also want
to ask questions.  They all seem to be to my right, to your left.  

		Would you line up behind her?  After you're done, then OSHA will have
some questions.  Apparently, we will be able to break at noon for lunch.
 

		Okay, so state your name and spell your last name.

		MS. SEMINARIO:  I'm Peg Seminario, 

S-e-m-i-n-a-r-i-o.  I'm the Safety and Health Director for the AFL-CIO
here in Washington, D.C.  Good morning.

		Thank you very much for coming and for your testimony.  Just a few
questions about conditions in the different facilities.  Before I start
that line of questioning, I just want to clarify.  In what year was the
50 µg/m3 silica level negotiated between the UAW and General Motors? 
Someone mentioned 1991; is that correct?

		MR. HITCHCOCK:  I believe it was either between '90 and '92.  I'd have
to get the exact date on that.

		MS. SEMINARIO:  If you could provide that, that would just be useful
as a point of historical reference from the changes that took place.

		JUDGE SOLOMON:  Let me remind everybody, you have to identify yourself
for the record before you speak.

		MR. HITCHCOCK:  I'm sorry.  Rob Hitchcock.  I believe it was somewhere
around 1990 to 1992.  We could get clarification on that.

		MS. SEMINARIO:  Thank you.  There have been some suggestions or
recommendations made by some of the industry representatives in these
proceedings, that to control silica in a final standard, that OSHA
should move away from the traditional approach of controlling hazards,
which is applying a hierarchy of controls of engineering controls first,
such as ventilation, the enclosed cabs that a number of you spoke about,
work practices to limit exposure, and then finally with respiratory
protection.

		They have suggested that should be abandoned, and in its place the
standards should allow any kind of method that an employer might choose,
including allowing personal protective equipment, including the use of
powered air-purifying respirators, as was just demonstrated.

		In your view, would allowing such a change in the hierarchy, and
relying primarily on respiratory protection as the primary means,
provide adequate protection to workers against silica?

		MR. RAGLE:  Shawn Ragle.  No, we've got to have some sort of controls.
 For instance, on the dust collection, the production workers down below
filed a complaint that it was very dusty down there.  What we ended up
doing was putting an Andon system in.  Basically, that's a light telling
the operator that everything's working fine.

		What really happens is the electrical engineer attached the light,
triggers the light by looking at a motor starter and not the actual dust
collector or whether it's working or not.  The motor can be spinning. 
It will give you the green light, the go light, to go ahead and run. 
You may not have any belts on your collector.  

		You could have any number of problems.  I think what we should be
looking at is our differential pressure, maybe velocity.  That will give
us a true representation of how our dust collection is working.

		MS. SEMINARIO:  Okay.  Thank you.  Dr. Frank Mirer, the former health
and safety director of the UAW, was here last week and testified with a
lot of focus on foundries.  One of the issues that he talked about and
presented some data on and his experience on is that, in the foundry
environment, there are several jobs where silica dust is generated.  

		There's exposures that occur, cross-contamination to other workers
working in the facility who may not be working directly in that
location.  Is your experience at cross-contamination of other workers
who aren't directly involved in that process, where they may be the
highest exposures, also end up being exposed to silica dust,
particularly when you don't have it controlled at the source?

		MR. P'POOLE:  Jeff P'Poole, Local 523.  We have a process in our
facility where they spray ladles with -- I have to ask Stan exactly
what it is.  The spray on those ladles contains silica in it.  If you're
on the casting floor or, just say, doing another job in our furnace
room, the over-spray from that spray, you can be contaminated with it
that way.

		MS. SEMINARIO:  If you control the dust at the source, it would
protect not only the people working right there, but other people in the
facility.  Is that correct?

		MR. P'POOLE:	Exactly.  And like I was mentioning in my presentation,
whenever we make the high purity, there's an exhaust fan on that stir
station that does that, but it doesn't collect everything.  There's men
that are working on the floor all the time that will get the fumes off
the sand and lime that goes into the pot.

		MS. SEMINARIO:  Another issue that's been talked about quite a bit by
some of the witnesses, particularly some of those from industry, is that
if you apply this standard with certain control measures, that it is
going to somehow cause productivity in a facility to decline.  

		The question that I have, do you think the application of engineering
controls, which would control the dust at its source, as has been done,
I guess, in the GM facility in Defiance, if those kinds of controls were
applied in these other facilities, where there's still a problem in
lowering dust, would that have a negative impact on productivity, or
would it have a positive impact on productivity, given that workers
would no longer be exposed to these high levels of dust?

		MR. HITCHCOCK:  Rob Hitchcock, Local 211.  I don't think it would
hinder anything.  In fact, as an example, one of the things I spoke on
earlier about us upgrading some ductwork, basically, what we were able
to do is not only improve the air quality for the people working in the
area, but we also prevented a fire hazard.  In fact, the ductwork had
actually caught on fire a couple of times prior to that.

		I think if nothing else, I think it provides a sense of security, a
sense of safety-ness for the people working in the area.  For
management, it provides a sense of ability to run.

		They know that their equipment is good running condition.  They know
that the people are safe.  They can go to a meeting or go to wherever
they may go, knowing that they don't have to worry about someone getting
hurt or someone falling out, whether it is respiratory issues or heat
stress or whatever it may be.

		MS. SEMINARIO:  Thank you very much.  

		MR. COMAI:  Can we backtrack to a question, so other people can answer
it as well?

		JUDGE SOLOMON:  Well, there's no question on the floor right now. 
What question are you referring to?

		MR. COMAI:  It was just that several jobs where dust is generated,
people who are nearby that operation are exposed.

		Andy Mercer from Local 8.

		MR. MERCER:  My name's Andy Mercer from Local 8.  I would like to add
to that.  The PPE, if it was the ventilated hood for whatever, the guys
in our areas in the high concentration, I'm not sure of our overall
building size, but we basically only have one interior wall.  All the
operations, I might be performing in an area of high silica, and at the
end of this right here, 10 feet away, somebody else, a totally different
job, open floor plan.

		It's going to have exposure indirectly just from being in the
building.  Their job isn't going to require that because they're not
labeled as a high-concentration area.  Overall collection of it and
getting out of there for the whole facility is the way to go.

		MR. COMAI:  Again, when we toured that facility, there was a worker
whose main job was doing quality control and going around sampling
silica.  He had a high cristobalite reading.  

		We have people that are doing ladle repairs.  If we put those people
in respirators, and sure they're able to repair the ladle and not be
exposed, there's some really simple methods to ventilate that ladle
repair using a snorkel and some local exhaust ventilation, where the
operator can be respiratory-protection free.  He's got the local exhaust
ventilation, and we're protecting all those other people around that
area who are nearby.

		 A couple solid examples where engineering controls are necessary.  If
we don't do that, then we spread that contamination and the exposure to
a broader work floor.

		We had some videos, as well, where we have people on the pour deck. 
They're skimming dross.  Why do they have this high silica level?  For
that operation, they're not near the sand molds.  Those sand molds are
50, 60 feet away and kind of curl around that whole melt deck.  People
are exposed to that sand drifting down on them.

		MS. SEMINARIO:  Thank you.

		MR. FREDERICK:  Good morning.  My name is Jim Frederick with the
United Steelworkers Union, and it's F-r-e-d-e-r-i-c-k, for the record. 
I have a couple of questions.  

		First, thank you to the UAW and your panel for coming in today.  I
know you all had to take some time out and travel to Washington.  It's
much appreciated to add your stories into the record.

		The first question is for Greg Essex.  It's a question about health
effects of silica and silica exposure.  Earlier in the hearings, there
was a person that was a former medical director from Navistar.  They
stated that they were unaware of any cases of silica-related disease in
anyone hired by the company since the 1960s.  Are you aware of any
silica cases, disease-related, among your members and co-workers in the
foundry?

		MR. ESSEX:  I'm not a medical professional, so I cannot attest to that
any of these were related.  All of the names that I had mentioned
earlier, they all died of cancer.  It's kind of hard for me to conceive
and comprehend how a medical professional from Navistar could make that
statement.

		I could give you an example.  I filed a claim two years ago.  That
claim was dismissed without merit.  I wasn't sent out.  I was treated on
site.  No further testing; it was just dismissed.  I'm on thyroid
medicine today and have been for the last two or three years.

		MR. FREDERICK:  A couple of questions again for the panel, anyone or a
couple of people if they want to respond to this, about industrial
hygiene and air monitoring in the facility.  Several of you touched on
this in your testimony.  I just wanted to get a little bit better sense
of how frequently employers are currently monitoring for silica exposure
in your workplaces?

		MR. RAGLE:  Shawn Ragle, Local 974.  The only time that I've witnessed
any silica monitoring is when it's requested.  I'm not even aware, even
being the safety chair there at the plant, I'm not even aware of what
their practice is as far as their plan to monitor.

		In fact, we've got an active safety complaint right now where I went
down and took air velocity readings on three ducts.  Two of the ducts
didn't meet the minimum airflow.  The other duct just barely met the
minimum.

		I asked, what are we going to do about it?  Well, we'll see what we
can do.  It's just an ongoing process of trying to get them to work with
us.  

		As far as an active monitoring, I haven't seen it.

		MR. FREDERICK:  Okay.  And before anybody else answers that -- sorry,
you all have very good testimony.  It's running together in my head
slightly.  You identified that you do maintenance on ventilation system.
 It's kind of your day job, correct?

		MR. RAGLE:  Yeah, that's correct.  I'm a repairman there.  I've been
there for about 20 years a repairman.  I've been a full-time, every day
working on the ventilation for the last 10 of those 20 years.

		MR. FREDERICK:  Again, in earlier testimony in the hearings, it's been
stated, and I think we certainly from the USW agree that ventilation is
a significant investment in the workplace, and it's an important
investment.

		In your workplace, is there a preventative maintenance program so that
you go around and make certain that equipment is working properly in
exhaust ventilation?

		MR. RAGLE:  We do have preventive maintenance that we do perform. 
It's more for the well-being of the machinery, not so much as the dust
collector or the ventilation actually doing what we want it to do.  It's
more for how can we make it last longer.

		That's why I believe we need to start looking at controls.  I'd like
to see static pressure readings or differential pressure readings. 
Let's put on some photohelics or magnehelics, so that we can actually
see what's going on.  

		A lot of people walk up to a dust collector or a fan, and it's
running, but that's all they know.  It can easily be running, the fan
can be turning, everything looks good.  But it may not be doing you any
good because it may be plugged up.

		MR. FREDERICK:  And just to follow up on that, and then we'll bounce
over to another, but to follow up on that, does the equipment operate
more effectively after you've done maintenance on it?

		MR. RAGLE:  Oh, most definitely.  Actually, the equipment runs better.
 Also, the people inside, it makes a world of difference when you do a
little maintenance to it.  It's just the blowdowns.  

		The blowdowns is a part of the machinery that keeps your filters
clean.  If those aren't working, you're basically running on a plugged
up vacuum sweeper is what you've got.		

		MR. BOECKER:  Richard Boecker, Defiance, Local 211.  At our plant, we
have a system called Maximo, where it tracks preventative maintenance,
where you do the velocity checks.  Check the louvers, check the docks. 
I think it's on like a monthly basis.

		Then, that is incorporated into like a maintenance plan for the
weekend schedule, that type thing.  As another example, like Monday
inside the plant, we were getting complaints of excessive dust on one of
the mold lines.  

		We didn't even need to do an air sampling.  It was basically called to
the skilled trades supervisor and the manager of the area to coordinate
the filter change for the next down time.  I guess the point is there is
easily preventative maintenance measures that can be put in place.

		MR. COMAI:  Just to clarify, Maximo again is a computerized program. 
The fact that the ductwork or the dust-collecting system plugged up, you
go back into that computer system.  You go back into that computer
system and say, well, instead of every month, we're going to have to
change that to every two weeks.

		It's a computerized system where there's constant feedback.  If we get
a complaint from the floor, then we go back and we change that
preventative maintenance schedule.  A red flag comes up, maintenance
supervisor hands out tickets saying, okay, here's the preventative
maintenance that needs to occur during the weekend shutdown. 

		MR. RAGLE:  Shawn Ragle, Local 974.  When you talk about preventative
maintenance, now, like I said in my testimony, I do EPA compliance, and
we do have it for that.  That's really if you want to say, do we do PMs
to see if the collectors are operating properly?  It's only because of
EPA compliance.  That would be the only reason.

		MR. FREDERICK:  If I can jump back to the industrial hygiene question
from earlier, and switch gears over to the example from the UAW GM
facility that discussed in the testimony about the training for local
union representatives as industrial hygiene technicians.

		Can you describe that program a little bit more in kind of the
availability of training for the UAW members across the GM chain to go
to that, receive the training?  Then, a little bit more detail of you
described a plan for the year of air sampling.  At your facility, is
silica included in the air sampling on a regular and periodic basis?

		MR. HITCHCOCK:  Rob Hitchcock, Local 211.  I'll try to answer that the
best I can.  There's a few questions in that.  

		For starters, our location, and I believe just about every location
for General Motors UAW, has someone identified for an industrial hygiene
representative, a member from management and a member from union.

		There is specific training that is provided for them through the
parties in Detroit.  We have a Center for Human Resources in Detroit
that specializes in training for several different things.  

		We also have, when we talk about a plan, and I believe Rick Boecker
spoke on this earlier, each year, our plant submits a plan of testings
that they would like to do throughout the year.  Now, it is not one,
two, three, four, five.  I mean, it's hundreds of them.

		And what they do is they scatter that out throughout a 12-month
period.  When a question is asked, how often do we do it, the air
sampling, we could pretty much say on a daily basis, based on the
average of how many we do per year, with the exception of a situation
that arises, an out-of-standard situation where a piece of equipment may
break down or what have you.

		We have the ability to go out there right then and there, run the
test, and get the results back.  In the event we need to put our members
in some type of a respiratory respirator or air-cooled or whatever it
may take, we have the ability to do that right then on the spot.  We
don't have to wait, per se, for results.

		MR. FREDERICK:  Just one last question on that.  In your opinion and
in the opinion of your members that you represent, is industrial hygiene
air monitoring readily available for your workplace?

		MR. HITCHCOCK:  Once again, Robert Hitchcock.  Yes, it is readily
available.  In fact, one thing I'd like to add is in the event a
decision would be made to put someone in a respirator due to a
circumstance, they will remain in a respirator until there are two
successful tests taken. 

		What I mean by that is, let's say they put me in a respirator today,
and next week, we take a sample of the area.  It comes out good.  I
still have to be in a respirator until a second test is done.

		That second test cannot be done for a minimum of one week later.  At a
minimum, the employee is going to be in that respirator for a minimum of
two weeks to ensure that the air quality has improved.

		MR. FREDERICK:  A question for Mr. Wafford.  You described the
warning signs in the workplace.  Do the warning signs that are included
in your workplace about silica provide good notice to the coworkers you
work with, others from the employer that may not normally work out on
the shop floor, any other folks that might be coming in, vendors or
others, about the hazards of silica?

		MR. WAFFORD:  No, there's no signs that say there's dangerous silica
in the facility.

		MR. FREDERICK:  Okay, I apologize.  I misunderstood your statement
earlier.  My apologies.

		My final question just about education and training on silica.  Again,
this was touched on in a couple of folks' testimonies.  Has anyone
currently experienced where there's not training on silica hazards in
your workplace?

		MR. GRAVES:  My name's Rodney Graves.  We're probably the perfect
facility for that because most of these guys are foundries and stuff. 
In my actual plant, we make gears.  It's mostly machine.  Like I said,
we've got the largest heat-treat facility in the North America area.  	

		We put the gears in it.  So our company is not really up on the silica
because they just buy the product already made and bring it in.  The
skilled tradesmen put it in.  It's not like a part of everyday process
for them for their production.

		MR. FREDERICK:  Okay, so it's not been included in your hazardous
communication training?

		MR. GRAVES:  No, none of the oils and stuff, no.

		MR. WAFFORD:  Matt Wafford.  Indiana OSHA decided -- we're not
training on silica.

		MR. RAGLE:  Shawn Ragle.  I'd like to speak to that, if I could.  At
our plant there, I don't remember the last time we had any silica
training at all.  In fact, I think you could go out and ask the employee
on the floor what the hazards were or where can you even pick it up at?

		I don't think anyone could really tell you.  It's one of those things
that, if it is educated, it's because somebody else has told you on the
shop floor, you probably ought to wear a respirator because bad things
can happen to you.  

		We've had several people that worked there that's lost lungs, one lung
or part of a lung, in our plant.

		MR. FREDERICK:  Thank you.

		JUDGE SOLOMON:  Thank you.  

		MS. TRAHAN:  Good morning.  Chris Trahan with the Building Trades,
T-r-a-h-a-n.  I wanted to ask about, I believe it's in the Defiance
plant.  You gentlemen have talked about measuring velocity, doing
velocity checks.  Who does that?

		MR. HITCHCOCK:  Rob Hitchcock.  We have skilled trades that perform
them tests.

		MS. TRAHAN:  Is it the same as the IH techs that we're talking about?

		MR. HITCHCOCK:  No, it's not.

		MS. TRAHAN:  What kind of training does it take to do a velocity
check?

		MR. HITCHCOCK:  Well, we've got that incorporated into the millwright
trade.  Millwrights, as they go into an unfamiliar area, we have safe
job procedures and other procedures that they have to read through and
educate themselves on.  

		If there is something such as that, that would require some additional
training, even outside the plant, we would even send them to that.

		MS. TRAHAN:  I guess what I'm trying to figure out is, can a
construction worker be trained to do a velocity check?  Is that
something that you could envision a skilled construction worker doing
outside of your plant environment, based on what you know about what it
takes to do that?

		MR. HITCHCOCK:  I believe --

		MS. TRAHAN:  Can I clarify that question a little bit?  Not in your
plant, but with equipment they may use in the field.

		MR. HITCHCOCK:  I believe if anyone's properly trained, anyone could
probably do the job.

		MS. TRAHAN:  And what type of instrument is used?

		MR. HITCHCOCK:  It's just a small digital type meter system, a
handheld meter.

		MS. TRAHAN:  Just to try to close that up a little bit, could you
venture how much time it would take to train somebody to use that small
handheld velocity meter?

		MR. RAGLE:  Shawn Ragle, Local 974.  I actually take velocity
readings.  Basically what you're using is you're using a handheld meter.
 I use a hotwire, and a hotwire is a probe that you put in the duct and
take your reading.

		Basically, it's very simple.  Probably you read through the manual,
have probably a 45-minute class or something on your meter.  Learn what
you're seeing on your meter, and that's pretty much what you have.

		MS. TRAHAN:  All right.  Well, thank you very much.  I appreciate it.

		JUDGE SOLOMON:  Finally --

		MS. REINDEL:  Hi.  My name is Rebecca Reindel, R-e-i-n-d-e-l.  I am
with AFL-CIO.  Thank you very much for coming today.

		I just have a couple of quick questions.  Mr. Ragle, you mentioned
that there's limited air monitoring in your plant.  I was wondering, as
a safety rep, have you ever been allowed to observe the air monitoring
that has been done?

		MR. RAGLE:  Shawn Ragle, Local 974.  Excellent question.  Actually,
I've requested to be an observer for air monitoring, and the company has
denied me that access.  They've chosen to go with the employee that they
put the monitor on.  

		Really, if you're doing your job, how are you going to monitor your
monitor to make sure everything is going correctly?  I really think that
we need to have a little more voice, or at least some validation that
the monitoring is being done correctly.

		We shouldn't put that on the employee wearing the monitor.

		MS. REINDEL:  Thank you.  Mr. Ragle, one more question for you
specifically.  I was wondering what kind of medical surveillance that
the foundry workers currently get.

		MR. RAGLE:  Our plant got it through bargaining.  An employee that's
been at the foundry can get a chest x-ray every four years if he goes
and requests it.  The company won't tell the employee that he's able to
get it at this time.

		What they do is they take your chest x-ray.  They send it to a third
party in Ohio.  They don't put on the chart that you're a foundry
worker.  They just send the chest x-ray to a facility in Ohio.

		MS. REINDEL:  Okay.  Thank you.  Then, I have one more kind of
question for anyone who would like to answer.  We've heard from some
workers, and other workers in this hearing, about dust in the workplace
being taken home with them on their clothes or maybe it goes into their
vehicles, and heads to their home.

		I was wondering that if you were only to rely on respirators in the
workplace, would that help relieve any of the dust that you take home
with you?  Can you talk about any experience you have of dust ending up
in other areas of your life that are not at work?

		MR. RAGLE:  Shawn Ragle, Local 974.  I can talk a little bit on that. 
Foundries are tall buildings, and there's lots of stairs.  If you have
to wear a respirator and try and climb stairs, as you could probably
guess, that's probably not practical.

		There's a lot of instances where a respirator is kind of a hindrance. 
Even though it may be trying to provide you cleaner air to breathe, it's
a hindrance to a lot of tasks that are being performed in the foundry. 

		A lot of your employees are going to choose not to wear it just
because they just don't want to wear it because of how it affects their
job.

		The dust, like I said, in foundries, there's dust everywhere.  You go
over, and you tap on a beam, you're going to get dusted.  The building,
at least in our facility, to walk into the foundry and not take home
dust with you, I think it's unrealistic.

		MR. MERCER:  Andy Mercer, Local 8.  Part of your question, as far as
bringing it home, it's a recognized hazard in our facility.  We have, in
our contract, the company has to maintain locker room facilities for
people to be able to shower.  We also have our uniforms are provided
free to prevent bringing home dirty clothing, having to wash it with
other family members' and stuff like that.

		There are measures taken for us, as far as that goes, as it's
recognized.

		MS. REINDEL:  Okay.  Is that the experience of other facilities?  You
have those facilities at other workplaces?

		MR. ESSEX:  Greg Essex, Local 226, Indianapolis.  It's the same as our
facility.  They provide shower facilities and locker rooms and the
uniforms in most cases.

		MS. REINDEL:  Okay.  And as some of you have mentioned in your
testimony, the increased ventilation in your facility would decrease the
overall amount of dust that's around in the facility.  Do you see less
dust ending up on you, if that is the case?

		MR. RAGLE:  Shawn Ragle, Local 974.  Yeah, the dust collection, we use
in there, what I see is they may have dust collector, but they don't use
a lot of local capture on a lot of stuff.  They'll want to go with a
general capture, and it's not very efficient.

		There's a lot of things that we could do to make improvements on that.
 One thing would be to get more local capture on where the problem areas
are.  Then, we can monitor whether magnehelic, it's a gauge you use to
measure differential pressure.

		Just a lot of things could be done to maximize the equipment that we
have now.  

		MS. REINDEL:  Okay.  Thank you.

		JUDGE SOLOMON:  Okay, Ms. Ryder.  

		MS. RYDER:  Okay.  OSHA will begin our questioning now.  We will start
with Jessica Schifano.

		MS. SCHIFANO:  Thank you so much for your testimony this morning.  We
really appreciate it.  

		My first set of questions is for Mr. Wafford.  I just wanted to
follow up about the use of the respirator that you demonstrated earlier.
 

		Is that respirator connected to a supplied air source or a
battery-powered fan?

		MR. WAFFORD:  It's a supplied air source.

		MS. SCHIFANO:  You mentioned that while you're wearing the respirator,
that it can fog up.  Really, that requires you to take it off.

		Can you just talk a little bit more about specifically if you are
using the respirator, how it affects your vision and ability to do your
job?

		MR. WAFFORD:  When you're using it, just like your safety glasses,
it's going to get fogged up.  It's going to get dust all over it.  You
ain't going to be able to see to do your job with it.

		MS. SCHIFANO:  Does the respirator also impact your ability to
communicate?  If so, how does it do that?

		MR. WAFFORD:  It makes it so your face is covered, your whole head is
covered, your ears are covered.  You want to say something to the guy
beside you.  He can't hear you.  He can't read lips, so you're going to
have to take it off and talk to him and then tell him to take his off,
too, so he can hear you.

		There's three of us downstairs that has to break these casts.  The
first guy flips it, and then the other two gun it.  Well, if your first
gunner misses some, you're going to have to holler at the other guy and
say, hey, you're going to have to get these because I missed them.  

		Say your gun breaks, you've got to holler at him.  Or the person
running the shakers, if they don't flip it, they'll have to holler at
you, hey, I didn't get that part done.  You're going to have to pick it
up, the next two guys.

		MS. SCHIFANO:  So you would say communication is vital to performing
your job effectively?

		MR. WAFFORD:  Yes.

		MS. SCHIFANO:  And that the respirator impairs your ability to do
that?

		MR. WAFFORD:  Yeah.

		MS. SCHIFANO:  And so, the effect of all of these complications is
that you do not wear the respirator regularly in performing these
operations?

		MR. WAFFORD:  Correct.  I wore it for like an hour.

		MS. SCHIFANO:  Okay.  Great.  Thank you so much, Mr. Wafford.  

		My next set of questions is for Mr. Boecker and Mr. Hitchcock.  We
were talking earlier about the IH technician at your facility.  I just
wanted to know if they were trained specifically about silica hazards. 
If so, what that training consisted of.

		MR. HITCHCOCK:  Rob Hitchcock.  Yes, they were.  Mike Bassinger (ph.)
is our IH tech.  He's been around for quite a few years.  He's probably
one of the most knowledgeable persons probably within the UAW, to be
honest with you.

		Over the years, he has taken several courses.  I couldn't tell you
right now exactly what those courses were.  I'm sure the UAW could get
that.

		MS. SCHIFANO:  Great, thank you.  With this specific knowledge that he
has of silica hazards, is he able to identify places in the facility
where there may be problems with silica, inform workers that are
performing those operations of silica hazards, and just generally
provide information to other workers in the facility about silica
hazards?

		MR. BOECKER:  Yeah, that's part of the plan, the yearly plan that Mike
will put together -- sorry, Rick Boecker, Local 211.  That's part of the
plan that I spoke of that they put together on a yearly basis and
jointly agreed upon with the UAW and General Motors.  

		Since Mike is trained with silica, he knows the areas that potentially
are more at risk than others.  He'll go out and canvass employees and
put the pumps on them or whatever material he needs to perform the
tests.

		MS. SCHIFANO:  Great, thank you.  And Mr. Boecker, you mention again
the plan that is put together on an annual basis.  Does that plan also
include information about exposure controls and what sorts of things are
supposed to be in place to protect workers from silica hazards?

		MR. BOECKER:  Yeah, I can't 100 percent answer that.  But I know Mike
is involved in, I guess, preventative measures, that's what you're
asking.

		MS. SCHIFANO:  Yes.

		MR. BOECKER:  Yeah, so based on that, he'll give recommendations to
engineering, or they'll work together to try to rectify the issues. 
Does that answer your question?

		MS. SCHIFANO:  Absolutely.  And more generally, are all of -- is all
of the information in this annual plan or document that you put together
about silica hazards, is that used to inform workers?  And what is the
value of that plan to a regular worker?

		MR. HITCHCOCK:  Rob Hitchcock.  What that does for our workers, for
one, it provides a sense of security, I'll call it.  As a week goes by
and an area is tested, Mike will go out and inform the area what he's
doing, why he's there.  

		And when he's done and the results are back, he also goes back and
informs them of the results.  So it's a continuous educational piece
that is in direct correlation of the schedule that we put together on a
yearly basis.

		MS. SCHIFANO:  Great.  Thank you so much.  That concludes my
questions.

		MS. IANNUCCI:  Okay, thank you, Jessica.  I'd like to continue on with
some questioning.  This is Annette Iannucci.  

		I'd like to follow up with Mr. Ragle, first of all.  You mentioned
that your chest x-rays get sent to a third party in Ohio.  Do you know
what they do with the x-rays there, what the third party does?

		MR. RAGLE:  No, I have no idea what they do in Ohio.  They basically
read the x-rays.  And if they send them back to the plant, after the
plant receives them, they'll call us up there.  And they'll have the
plant doctor tell us if there's anything wrong or not.

		And, typically, I've not heard them ever say anything was wrong with
anybody.

		MS. IANNUCCI:  Okay.  So the results do get reported back to your
employer in that case?

		MR. RAGLE:  Yes.  And then the employer reports back to the
individuals that had the x-ray taken.

		MS. IANNUCCI:  Okay.  Are any workers intimidated by this, by having
the results go back to the employer?

		MR. RAGLE:  No.  Actually, at Caterpillar, they don't advertise that
this is available.  I've asked the company if they would notify the
individuals on the shop floor, when their four years are up, to have
it -- have the x-ray completed, and they've elected not to so far. 

		MS. IANNUCCI:  Okay.  Thank you.  

		Now, I'd like to ask if any of the other companies have any medical
surveillance, and what that surveillance program involves.

		MR. P'POOLE:	Jeff P'Poole, 523.  Once a contract year, our company
allows us to have a physical done.  And they will do one x-ray on that. 
I don't have any knowledge of anybody going through it and having it
done.

		They do try to make us do it.  But a lot of people, you do it on your
own time.  Most people won't do anything on their own time, even if
it'll save your life, I guess you could say.

		MS. IANNUCCI:  Okay.  So having it on company time would probably
increase participation?

		MR. P'POOLE:  Oh, yes, definitely.

		MS. IANNUCCI:  Okay.

		MR. BOECKER:  At Defiance, we do yearly physicals -- I'm sorry,
Richard Boecker, 211 -- we do year physicals, and I guess the pulmonary
testing where you blow.  We did that on a regular basis, yearly basis,
in the past.  But now, I don't know what the sequence is, we -- maybe
three years or four years, I'm not real sure.  But we do some testing.

		I guess chest x-rays, we have done them in the past, too, so optional.

		MS. IANNUCCI:  Optional.  Okay.  And do those results go back to your
employer?

		MR. BOECKER:  Yes.

		MS. IANNUCCI:  Yes.

		MR. P'POOLE:  I left that out, too.  Every year, we do the pulmonary
and our hearing test and stuff, where you, you know, blow to make sure
your lungs can, you know, function properly in our plant. 

		And that's the same time of year that they make us have a respirator
fit test.  So it'll --

		MS. IANNUCCI:  Okay.  It is mandatory at your company?

		MR. P'POOLE:  Yes.

		MS. IANNUCCI:  Okay.  For the other companies where it's not
mandatory, do you have any idea of how many employees participate, a
percentage?

		MR. P'POOLE:  That test is mandatory, but the physical per contract is
not mandatory.  Okay, just to clear that up.

		MS. IANNUCCI:  Okay.  And do you have any idea how many employees
participate, take advantage of it, a percentage or --

		MR. GRAVES:  Well, in our facility, we have our own health clinic and
stuff.  And we do -- we actually were here lately unaware of all the
silica stuff, and that's why I'm here.

		The UAW health and safety knows -- trying to -- what these standards
and stuff, and letting our plant know that we do have problems in the
facility with that.  

		So at the moment, we currently don't have anything for that.  We just
do the yearly physicals that -- and they do pay you to go do that.  And
it actually lowers your monthly insurance cost, if you do their
physicals.

		MS. IANNUCCI:  Okay, Mr. --

		MR. GRAVES:  Rodney Graves, I'm sorry.

		MS. IANNUCCI:  Yes.   

		MR. WAFFORD:  Matt Wafford, 2239.  We don't have no yearly physical or
no checks necessary where I work at.  I've worked there seven years.  I
ain't never had nothing done.

		MS. IANNUCCI:  Okay.  Thank you.  Is more medical testing something
all you would like to see?

		MR. MERCER:  Andy Mercer, Local 8.  We have nothing, and I would
definitely encourage that, absolutely.

		MS. IANNUCCI:  Okay.  And would you have any preference of how the
results are reported?  What gets reported to your employer?

		MR. MERCER:  Personally, I wouldn't.  I mean, that's something I think
you'd want a member consensus on within the facility, you know, how they
want that handled and how it's most applicable.

		MS. IANNUCCI:  Okay.  And can I please direct a question to
Dr. Sivin?  I think you might be the right person.  I know in your
comments, you mentioned that the only thing that should get reported to
the employer is whether the employee has limitations on respirator.  

		I was just wondering if you have any experience.  Like if someone
needed to be referred to a specialist, how you'd be able to handle that
and maintain confidentiality.

		DR. SIVIN:  Well, I think you would give the referral directly to the
employee, excuse me.  If it were a question of who was paying for it, et
cetera, the employee would have the right to share that information with
the employer.

		If, for some reason, I mean, they'd have the right in any case, and
they'd certainly have the right if it were somehow procedurally
necessary in order to get the service.  

		If the employee decided that the information about the referral were
somehow a threat to their ability to maintain their current position
with the employer, it should be the employee's right not to share that
information.

		MS. IANNUCCI:  Okay.  Thank you.  

		I just wanted to check on a couple of other things.  I think I've
heard several of you mention that your family members have worked at a
certain place maybe their entire work duration.  Is it uncommon to see
somebody work at a location maybe for 45 years over their entire working
lifetime?

		MR. BURKEEN:  Stan Burkeen, Local 523.  As a person and personally, my
family's been at CCMA since it opened in 1950, so I've had several
family members there.  And it's not uncommon.

		We have a gentleman, Rob Low (ph.), who works there now.  He's been
there approximately, what, Jeff, close to 45 years.  He's in his
seventies.  He's still not ready to retire.  So it's not uncommon in our
place.

		Now, there's been a time throughout the years that they wouldn't hire
family members because there was so many family members in the plant. 
They've recently got away from that and went back taking referrals from
the employees.  And so, there's more family members there now.

		MR. HITCHCOCK:  Rob Hitchcock.  I just want to make a point here.  A
few years ago, we just had an employee retire.  I believe he had 53
actual years.  So yeah, people do tend to stay for quite some time.

		MS. IANNUCCI:  I'm sorry, Dr. Sivin, did you have something else?

		DR. SIVIN:  I'm fine, thanks.

		MS. IANNUCCI:  Okay.  Thanks.  One other question, we've heard some
commenters say that at some locations, it's not possible to use a
cleaning method other than compressed air or dry sweeping.  Have any of
you found that to be the case?

		MR. RAGLE:  Shawn Ragle.  Yeah, we looked at how could we get away
from using compressed air.  And we were able to change a few jobs, but a
lot of the jobs that we used compressed air, they still use it just
because of the process.  It simplifies it.  There's less cost, and it
seems to work good for the company to keep the compressed air the way it
is.

		MS. IANNUCCI:  Okay.  

		MS. RYDER:  I guess just a follow-up on that really quickly -- Anne
Ryder.  

		Do you know whether or not it's impossible to use some other method of
cleaning? 

		I think yesterday we heard that cleaning around molten metal is just
too hard to do, to use wet methods.  I'm not sure if any vacuum methods
were mentioned.  But do you know whether it's actually possible to use
another method of cleaning in that situation?

		MR. RAGLE:  Yeah, we come up with other methods for some of the jobs
we looked at.  But after it was discussed, we just -- the company chose
to go ahead and stay with the air.  But yeah, it's definitely possible.

		MS. IANNUCCI:  Okay.  I know some of you mentioned that they regularly
check the ventilation controls to make sure they're meeting EPA
requirements.  And is this because of EPA laws that require a certain
frequency of checking those controls, or does anyone know?

		MR. RAGLE:  Yeah, you'll have controls.  I think the controls, just
going off memory, was around 2007 when they were enacted for foundry
melters or dust collectors that pull from the melters that are treated,
when treatment is put into a melter.

		Those collectors, you have a daily check, and it's just a general
check.  And then, you'll have a quarterly check that you have to file
and keep on hand for, I want to say, it's 5 or 10 years.

		But, anyway, those checks are more in-depth.  Those checks are the
ones that check for holes in the bags and to ensure that the collector
is actually performing like it should be doing, and that we're not
emitting contaminants more than what the permit would allow for.

		MS. IANNUCCI:  Okay.  So there's OSHA regulations requiring things
such as air monitoring for a certain frequency.  Your employer would be
likely to follow those?

		MR. RAGLE:  Yeah.  Personally, I think that would be a good thing to
do anyway.  If you've got a happier workforce inside, we know our
collectors are working.  We've proven in our plant that, you know, a
little maintenance goes a long way on our collectors.  And that cleans
your air up a lot inside there, and you get a lot less complaints.

		And actually, the productivity, you know, if people are not having to
blow black stuff out of their nose and hack stuff up, they're a lot
happier with their job, if they can go in, do a job, and be able to go
home.

		MS. IANNUCCI:  Okay, thank you.  That's it for me.  I'd now like to
turn it over to Joe Coble.

		DR. COBLE:  Good morning.  Thank you all very much for being here
today.  I'd like to compliment you on some of those videos and pictures.
 I think the saying is, "A picture is 1,000 words," and really
communicated effectively, I think, some of the challenges and the
sources of the dust.  To actually hear from people that work in those
settings is extremely valuable.

		We saw lots of examples of engineering controls that were in place but
appeared to be leaking or maybe not doing what they were designed to do.
 How much of the problem is related to the lack of controls versus just
ineffective controls that are in place, in your experience?

		MR. RAGLE:  Shawn Ragle.  At least in our plant, one of the issues you
have would be like a pneumatic sand transporter.  You're basically
blowing your sand through a pipe across the plant.  Essentially what
you're doing, you're sandblasting the inside of the pipe.  That creates
a lot of little leaks that you're always chasing.  It's always eating
through the pipe.  So you've got a constant leak basically on that.

		As far as your shaker tables go, they jam up a lot.  So a lot of
times, maintenance people, the facility, don't want to put covers on
those shaker tables because that's just an additional burden to try and
fix the problem.  So controls would help out a lot, if we could get some
sort of, you know, hard control we have to follow.

		DR. COBLE:  We saw the shot core booth, I think it was, in one of the
earlier ones.  I don't know whether it was Andrew -- but it was -- you
said they were using steel shot to clean the parts.  Was that instead of
using silica sand?  I mean --

		MR. MERCER:  Andy Mercer.  No, the steel shot is blasted, same as like
a sand blaster would do.  It's blasting the residual sand off the
casting, making it more a clear visibility on it for inspection and
finished purposes.

		DR. COBLE:  But is the steel shot being used instead of blasting it
with sand, or it's always been steel shot?

		MR. MERCER:  As far as I know, it's always been steel.  It's fighting
against the sand, so --

		DR. COBLE:  And then we're interested in this -- I think it was
Mr. Hitchcock talking about the challenges involved at the GM plant,
where it wasn't always like it is today.  And you mentioned back in the
early '90s, there was a commitment made to go ahead and try to address
this issue.  

		And I suppose at the time, there might have been some skepticism as to
whether you'd be able to get there.  Was it -- there are numerous
sources.  So was it a matter of systemically going out and identifying
those sources, one by one?  Or could you just kind of control some major
sources and that got you there?

		MR. HITCHCOCK:  Well, Rob Hitchcock.  I think the initial plan started
out as first, both parties sat down and collectively agreed that we're
going to move forward on health and safety, okay.  And once that was
decided and agreed to, I think what really took place is they identified
probably the worst areas of the plant to go start out with, okay.

		And as they would knock one area out, as far as meaning make the
improvements needed, then they would move onto the next.  And over the
years, those areas have been consolidated down to a very manageable
number of areas, where we can continue to monitor and make adjustments
as needed.

		DR. COBLE:  And then, you mentioned you have routine air monitoring. 
Does that indicate that you've achieved levels below 50 in most of the
plant?

		MR. HITCHCOCK:  Absolutely.

		DR. COBLE:  Did you --

		MR. HITCHCOCK:  Now, we do have a few areas that we still have to
manage, we keep close eye on.  It's right there at the 50, but those are
the areas that we'll continue to do the dust collector bag changes and
things like that, to keep them to that standard.

		DR. COBLE:  And how are the air monitoring results communicated to
employees?

		MR. HITCHCOCK:  Face-to-face.

		DR. COBLE:  Face-to-face, so that happens on a routine basis?

		MR. HITCHCOCK:  That's correct.  Each area of our plant has a safety
talk each week.  And as results come out, it's either taken to that
meeting in front of the group that works in the area or staying into the
specific individual that was tested.

		DR. COBLE:  Right.  And then my final question is for the panel in
general regarding the secondary exposures.  And to what extent, there
are some jobs that are directly working with the source of the sand.  

		But what percentage of jobs there are exposed primarily through
secondary exposures that could potentially be reduced by controlling it
at the source?  I mean, are there jobs in which you don't really work
with the sand, but you still get exposed anyway?

		MR. RAGLE:  Shawn Ragle.  Yeah, there's -- obviously like you said,
there's some jobs that have a high potential to be exposed.  But there's
also a lot of jobs, such as a trucker, that may be trucking through an
area that'd pick it up, just a truck driving across the floor, kicks up
dust.  

		You've also got, like we discussed, about blowing off parts or
inspecting parts.  You've got dust floating around there that could
affect the inspectors at the end of the line.

		So as far as a percentage goes, I just think about it real quick here,
I would say probably -- we probably have about, say, 70 percent of our
employees would have the second-hand exposure, and the other 30 would
probably be more of the higher risk category.

		DR. COBLE:  Okay, thank you.  That's it for me.

		MS. IANNUCCI:  Okay, thank you.  Tom Mockler will now ask questions.

		MR. MOCKLER:  Yeah, I wanted to follow up on the Defiance plant.  Is
it possible to get more detailed information on sort of the before and
after, in terms of like the cost of installing the controls and some of
the exposure measurements before and after, the number of employees
affected?

		MR. HITCHCOCK:  Rob Hitchcock.  I don't know if I could get you that
information or not.  I'll be honest with you.

		MR. MOCKLER:  Okay.

		MR. HITCHCOCK:  You know, the improvements, if we're looking at from
the time where the standard was lowered to 50 to today, you're talking,
you know, over 20 years.  I don't believe I could get that information
for you.

		MR. MOCKLER:  Is it possible to get parts of the information?

		MR. BOECKER:  Rick Boecker.  What specifically were you looking for?

		MR. MOCKLER:  Well, really, I mean, like for a particular job, how
much an engineering control costs to install, for example.

		MR. BOECKER:  Yeah, we don't have that access available to us.  You
know, I don't think that anybody's withholding it, but, you know, I
guess through Andy or something, we can maybe try to get you guys
something.

		MR. COMAI:  So we can work with you on that.  We have a -- there's a
national agreement.  There's a national Joint Committee on Health and
Safety.  I sit on that committee.  I sit on the Industrial Hygiene
Review Committee for UAW General Motors.  And we'll do the best we can
to fill in anything you can ask for.

		MR. MOCKLER:  Okay.  All right.  Thanks.

		MS. IANNUCCI:  Okay.  Anne Ryder will now ask questions.

		MS. RYDER:  I have a couple of questions.  I think first, I have a
question for Mr. Hitchcock and Mr. Boecker, since you guys talked
about your GM plant in Defiance.

		You said that most of the areas are staying below 50, but in some of
those areas, you're right at 50.  Are those areas demarcated in any way,
so that employees know, you know, this particular area has higher
exposure levels, and you need to take certain precautions because of
that?

		MR. BOECKER:  I believe those areas that are respirator-required,
that's posted as part of their job instruction.  They'll have like a, I
don't know, they call it the "little man" or something that'll have your
PP, your glasses, your respirator, that type stuff.

		MS. RYDER:  So how would an employee who's just walking around the
plant know that there's an area that has higher levels of respirable
crystalline silica?

		MR. HITCHCOCK:  In the event there would be an area where someone
could just walk into, it would be well-defined, signage.  We have all
kinds of signages for any type of hazard that we have out to the plant. 


		Our joint team is very good at identifying and ensuring that the
placards are updated and clear to see.  And anything we can continue to
do is what we do.  

		I mean, in the event there would be a particular area that would
create -- or that could potentially cause a hazard to anyone coming to
an area, most generally, we have a 100 percent contact at our plant. 
And it would be communicated through a safety talk in our weekly safety
meetings. 

		And that way, everyone understands that, hey, if you would go over to
this area, we have a situation right now.  Stay clear of the area, et
cetera.

		MS. RYDER:  Okay.  And you said there's usually a sign in that area or
something to notify people?

		MR. HITCHCOCK:  Yes.

		MS. RYDER:  Okay.

		MR. HITCHCOCK:  Yes.

		MS. RYDER:  All right.  I think maybe Mr. Ragle will be able to
answer this, but some of the rest of you might be able to, also.  We
heard yesterday from the National Association from Manufacturers that
when repair work or maintenance needs to be done on equipment,
contractors are coming in to do that type of work.

		So I was wondering if that's your experience.  I think you mentioned
that you do repair work.  And if there are contractors who are coming in
to do any repair work, who is telling them about the hazard, you know,
the silica hazard?  And who's in charge of those workers?

		MR. RAGLE:  Yeah, it depends on the job.  If it's a smaller job, as
far as in our plant, people that come in every day, knowing they're
going to work on the dust collectors are myself and a partner.  When you
get into a bigger job, such as a rebag, they'll usually hire an outside
contract crew that can bring in a sizable workforce to do that.

		Essentially, when an outside contractor comes on site, they have
to -- the company gives them basically a handbook that tells them all
of our safe practices to follow while they're on site.  Now, as far as
in the actual dust collector or maintenance of a fan, that's left up to
the employer to ensure that his employees have been properly trained and
notified of any hazards.

		MS. RYDER:  You mean the contractor would be notifying those
employees?

		MR. RAGLE:  Yeah, the meeting between Caterpillar and the contract
worker, they typically would only send their safety representative from
the contract company.  The safety representative of that contract
company would go back and relay any safety measures that we take or
follow for our plant.

		Now, as far as what's inside the collector fan or any hazards, that's
up to the employer or the contract employer to relay to his employees. 
I don't believe -- I haven't been part of their meetings, but I do work
with them out there.

		And as far as I can tell, it's mostly -- the stuff that they get told
is mostly about where to park on the facility, stop at all the stop
signs.  It really doesn't deal directly with what's in the dust
collector.

		MS. RYDER:  At that meeting, that pre-meet that they have, would
Caterpillar be able to tell them about, you know, you're working on
equipment that's got silica dust in it?  Do you think that's something
that's doable?

		MR. RAGLE:  Yeah, that's very doable.

		MS. RYDER:  Okay.

		MR. RAGLE:  The company is aware of what's going inside the dust
collector.  We know what comes out of it, you know, through -- and
where we get rid of the waste from the dust collector.  So, yeah, that
information's there. 

		MS. RYDER:  Okay.  Does anyone have anything else?  I think that's all
that we have.

		JUDGE SOLOMON:  Yeah, I don't want to prolong the agony, but I guess
curiosity and maybe the interest of justice require that I ask.  You
have a national contract with GM.

		DR. SIVIN:  Correct.

		JUDGE SOLOMON:  So is the Defiance situation replicated in your other
facilities?

		DR. SIVIN:  For the most part, yes.  And for example, there was a
question asked about ventilation checks earlier.  Those ventilations are
per the occupational health and safety language of the national
contract.

		JUDGE SOLOMON:  Do you track the health of the employees?  I assume
that you have claims that are filed.  You have the variations on the
theme of insurance claims, claims for benefits even for governmental.

		DR. SIVIN:  General Motors also provides pension and health care to
the employees until death.  So yes, in fact, and General Motors collects
all the death certificates.  And many epidemiologic studies have been
published out of that information.

		JUDGE SOLOMON:  All right.  And I just wonder whether OSHA had been
provided all of that information.

		MS. RYDER:  Not that I'm aware of.

		DR. SIVIN:  Some of it was in Dr. Mirer's testimony, and we'll be
happy to provide all other relevant information in post-hearing
comments.

		MS. RYDER:  Okay.  Great.  Thank you.

		JUDGE SOLOMON:  Okay.  Does anybody have any questions?  Do you have
any questions based on what I just --

		MS. RYDER:  No, I just wanted to mark for the record and enter into
evidence.  I think I have a copy of your PowerPoint presentation.  I'm
going to mark that as Exhibit 53.

		DR. SIVIN:  Yeah, we hope you have it electronically, too, with all
the videos.

		MS. RYDER:  Okay.

		DR. SIVIN:  I think -- I believe you have that all ready.  If there's
any part of that you don't have, we'll be -- we're happy to provide it.

		MS. RYDER:  Okay.  I'll mark that as Hearing Exhibit 54 and -- the
videos.  And I'm going to reserve Hearing Exhibit 55 for pictures of
that PAPR or the respirator that you -- 

		JUDGE SOLOMON:  All right.  Since there's no objection, Exhibit Number
54 is entered into evidence. 	

(Whereupon, the document referred to as Hearing Exhibits 53 and 54 were
marked and received in evidence.) 

		JUDGE SOLOMON:  Anything else, Ms. Ryder?

		MS. RYDER:  That's it.  Thank you.

		JUDGE SOLOMON:  Okay.  If there's nothing else, we're going to go off
the record until 1:00.

		(Whereupon, at 11:55 a.m., a lunch recess was taken.)

A F T E R N O O N   S E S S I O N

(1:00 p.m.)

		JUDGE SOLOMON:	 Okay.  We're back on the record.  This is the
afternoon presentation from the American Chemistry Silica Panel.  

		And, Ms. Lindberg, do you want to enter your appearance for the
afternoon?

		MS. LINDBERG:  Sure, Judge.  Kristen Lindberg from the Office of the
Solicitor here at the Department of Labor.

		JUDGE SOLOMON:  You have some exhibits?

		MS. LINDBERG:  Yes, Your Honor.  I have a number of exhibits that the
American Chemistry Council has provided, which I would like to mark now.
 

		Number 56 will be the presentation materials from Neil King and Samuel
Boxerman.  Number 57, if I can get it off the sheet here, will be the
presentation materials from Dr. P. Morfeld.  58 will be the testimony
of Dr. Peter Morfeld.  59, the presentation by Paul K. Scott.  Number
60, the written testimony of Paul K. Scott.  Number 61, the presentation
by Kelly S. Bailey.  Number 62, the oral testimony of Kelly S. Bailey
and Richard J. Lee.  Number 63, the presentation materials from
Jack Waggener.  And last but not least, the presentation materials from
Stuart Sessions.

		JUDGE SOLOMON:  So that is 64.

		MS. LINDBERG:  That is 64, yes.

		JUDGE SOLOMON:  Okay.  With objection, they are entered into evidence.

		MS. LINDBERG:  Thank you.

(Whereupon, the documents referred to as Hearing Exhibits 56 through 64
were marked and received in evidence.)

		JUDGE SOLOMON:  Okay.  Mr. Morrill, are you going to introduce the
members of your panel?

		MR. MORRILL:  Yes, Your Honor.  May I start now?

		JUDGE SOLOMON:  Sure.

		MR. MORRILL:  Okay, great.  I have a few opening remarks, and then
I'll introduce the panel as we go.  Okay, great.  Thank you.

		For the record, my name is Jackson Morrill, last name spelled
M-o-r-r-i-l-l.  I'm the Director of the American Chemistry Council's
Crystalline Silica Panel.  And on behalf of the panel, I want to thank
OSHA for the opportunity that you've provided for us to give testimony
today.

		Since 1989, the panel's mission has been to ensure the regulatory
actions regarding crystalline silica are grounded in sound science and
are properly designed to protect the health of workers and the general
public, without eroding the economic viability, competitiveness, and job
base of the many sectors of American industry that produce or use
crystalline silica or products in which it's contained.

		In fulfilling this mission, the panel has supported and continues to
develop and support cutting edge scientific and technical research on
silica-related issues that we believe are directly relevant to this
rulemaking.

		The panel's current membership, which you see here on this slide, and
apologies for the formatting there, but it covers a broad range of
companies and associations in a varied set of industries.  I think we
have a very broad representation in our panel, many of whom are impacted
by the proposed rule.

		In addition to supporting the panel's comments and hearing testimony,
various members of ours have filed their own comments and provided, or
will be providing, testimony during this hearing process, as well.

		Getting to our presenters for today, the Silica Panel hearing
testimony began last week actually with one of our witnesses, Dr. Tony
Cox, who provided a critical analysis of OSHA's preliminary quantitative
risk assessment.

		Today, we will be providing the balance of our testimony with the
witnesses you see up here today.  And there's some actually in the
audience as well.

		We provided to the docket copies of the CVs of most of our experts, so
I will only briefly introduce our panel before turning it over to our
first speakers.

		To my right here, Neil King and Sam Boxerman will kick off our
testimony with a summary of the panel's positions on the major issues
raised in this rulemaking as set forth in the panel's comments submitted
to the docket last month.

		Neil King is a self-employed legal consultant, who has provided
support to the Crystalline Silica Panel for nearly 15 years. 
Sam Boxerman is a partner at the law firm of Sidley Austin.

		Dr. Peter Morfeld, who is to my far right here, will follow with a
discussion of the epidemiological issues related to the proposed
rulemaking.  Dr. Morfeld is a noted epidemiologist based in Germany,
who's authored over 328 publications, a number of which directly
addressed silica, including several notable publications on the German
porcelain workers cohort, among others.

		He currently works at the Institute for Occupational Epidemiology and
Risk Assessment, a public-private partnership with Cologne University
and Evonik industries.

		After Dr. Morfeld's presentation, we will turn to the issue of
measurability.  And directly behind me is Paul Scott, who will provide
testimony on the adequacy of sampling and analytical methods for
measuring respirable crystalline silica at exposure levels of 25 and 50
µg/m3.  Paul's a qualified environmental professional in supervising
health scientists at Cardno ChemRisk, LLC.

		And then, behind me here are Kelly Bailey and Rich Lee, who will then
continue the discussion of measurability, focusing on the results of a
blinded laboratory performance study sponsored by the panel.  

		Kelly is chair of the ACC's Crystalline Silica Panel and also the
Director of Safety, Health and Environmental matters in the corporate
office of Vulcan Materials Company.  He is also one of our founding
members.  And, again, we've been around for 25 years, so Kelly's been at
this for quite some time.

		Rich Lee is the CEO of RG Lee Group, which is a well-known scientific
consulting and lab facility headquartered in Monroeville, Pennsylvania. 


		Finally, we will conclude with two presentations on economic
feasibility.  Our first expert, who is Jack Waggener, right there in the
audience, will be providing a discussion of OSHA's cost models for the
proposed standard, and an explanation of the modifications to those cost
models made by Jack and his colleagues at URS Corporation.

		Mr. Waggener, senior principal and engineering manager at URS, has
over 40 years of professional experience and is a recognized expert in
managing projects involving compliance with OSHA PELs.

		Finally, Stu Sessions is right here in the audience as well.  He will
conclude the panel's presentations with a detailed discussion of the
economic impact of the proposed rule, focusing in particular on impacts
to 19 sectors in general industry.

		Mr. Sessions is President of Environomics, Incorporated, an economic
consulting firm that provides analysis of the benefits, costs, economic
feasibility, economic impacts, and cost effectiveness of policies,
programs, regulations, and legislation involving the environment,
energy, and occupational safety and health.

		So, once again, we thank OSHA for this opportunity to present the
panel's positions on the proposed rule.  And I'll turn it over now to
Neil and Sam to begin our testimony.

		JUDGE SOLOMON:  Okay.  Mr. King, before you start, there was an
agreement off the record that the presentation would last about three
hours.  And then there would be questions. 

		We were advised that there would be some extensive questions and that
OSHA has some questions.  So, hypothetically, we'll be able to be done
around 5:00 or 5:30.  Okay, thank you.  

		I'm sorry, Mr. King.  Go ahead.

		MR. KING:  I'm going to just provide an overview of the Crystalline
Silica Panel's positions on the major issues in this proceeding, broken
down by significant risk, economic feasibility, and measurability.

		As I believe it's been made clear earlier in the proceedings, OSHA has
to make three showings to support a reduction in the permissible
exposure limit, that is the PEL, for crystalline silica:  

		First, showing that there's a significant risk of material health
impairment at the current PEL that would be substantially reduced by
reducing that level to the proposed PEL.  

		Secondly, that the standard would be technologically and economically
feasible in all affected industry sectors.  

		And, thirdly, that exposures can be reliably measured with an
acceptable degree of accuracy and precision at the levels of the
proposed and action level.

		Beginning with significant risk, as has been pointed out earlier in
these hearings, there has been a dramatic reduction in silica-related
mortality and morbidity in the four decades since OSHA's current PEL was
adopted.  That includes more than a 90 percent reduction in silicosis
mortality rate from 1968 to 2010.  Similar reductions in silicosis
morbidity, and a 99.5 percent decline in annual average silicosis
respiratory tuberculosis deaths from the five-year period 1968 to '72 to
the five-year period 2002 to 2006.  

		Now, some people at the hearings have suggested that silicosis
mortality may be underreported.  I don't know that that's correct, but
even if it is correct, it would not alter the dramatically declining
trend for silicosis mortality over these last 40 years because any
underreporting that may have occurred presumably would have been at
least as prevalent in the earlier years of this period as in recent
years.

		People have noted that cases of silicosis continue to be observed.  We
don't deny that, but we believe they're attributable to two factors. 
First, the over-exposures to crystalline silica that occurred decades
ago and are now manifesting after the latency period of 20 to 30 years
or more, as has been indicated earlier in these proceedings.  Most of
the workers diagnosed with silicosis in the last 10 years or so were
first exposed before OSHA's current PELs were adopted.

		The second factor is that there are continuing exceedances of the
current PELs, often by very large margins.  These occur at a rate of
about 30 percent or more year after year, as evidenced by OSHA's
compliance sampling results, and by its current estimates of worker
exposures in this rulemaking.

		OSHA's finding of significant risk is based on estimates of mortality
from lung cancer, nonmalignant respiratory disease, that is NMRD, and
nonmalignant renal disease, as well as cases of silicosis morbidity, as
these things arise after a continuous 45-year occupational exposure to
the respirable crystalline silica levels of 100 µg/m3 and 50 µg/m3.

		Those, however, are not the correct long-term exposure levels to
evaluate in assessing potential risks under permissible exposure limits
of 150 µg/m3.  OSHA, as it noted earlier in these proceedings, applies
its PELs on a daily basis.  So an employer is deemed to be out of
compliance with the standard if his employees are exposed above the PEL
on any given day.

		In order to be 95 percent confident that the PEL will not be exceeded
on any given day, an employer must maintain the long-term average
exposure level in his workplace at a fraction of the PEL.  That fraction
likely would be less than 50 percent of the PEL in many workplaces
characterized by typical exposure variability.

 		With that in mind, one can see that to ensure compliance with a PEL
of 100 µg/m3, long-term average exposures in many workplaces would
probably be maintained at a level below 50 µg/m3.  And to assure
compliance with a PEL of 50 µg/m3, as OSHA has proposed, the long-term
average exposures in many workplaces probably would have to be
maintained at a level below 25 µg/m3.

		In fact, they'd probably have to be maintained at a level below 20
µg/m3 because OSHA switching from the old ACGIH sampling protocol to
the new ISO/CEN sampling protocol will mean that about 20 percent more
respirable dust or respirable silica is collected than would've been the
case under OSHA's current sampling model.

		OSHA's risk assessments for all four of these endpoints did not
evaluate potential silica-related risks at these long-term average
exposure levels.  So its conclusions regarding risks under alternative
PELs of 100 µg/m3 and 50 µg/m3 are based on an evaluation of the wrong
long-term exposures.

		Apart from evaluating the wrong long-term exposures, the risk
assessments on which OSHA's conclusions regarding significant risks are
based suffer from a variety of serious shortcomings.  These are
discussed at length in the comments that the ACC Crystalline Silica
Panel and its various consultants filed on February 11, 2014. 
Highlights are discussed in the testimony that Dr. Tony Cox presented
last week and in the testimony that Dr. Peter Morfeld will present
later this afternoon.

		Among the shortcomings in OSHA's risk assessment are the following: 
OSHA's risk assessment assumes that silica exposures cause lung cancer
in the absence of silicosis.  OSHA's risk assessment assumes there is no
exposure threshold above 100 µg/m3 for any adverse health effect of
silica exposure.

		OSHA's risk assessment assumes a linear exposure-response relation at
low exposure levels, rather than considering possible thresholds or
J-shaped exposure-response relations.  OSHA's risk assessment fails to
model exposure estimation error.  OSHA's projection of renal disease
mortality is based on limited data from studies with large exposure
uncertainties.  It runs counter to numerous studies finding no causal
association between silica exposure and renal disease mortality.

		Finally, OSHA's risk assessment and the studies on which it is based
fail to control a number of biases that Dr. Tony Cox discussed in his
testimony last Wednesday.  These are listed on the slide.  I don't think
I have to go into them in detail because he went through them fairly
carefully last week.

		Now, I'm going to turn the microphone over, such as it is, to Sam
Boxerman, who wants to talk about technological and economic
feasibility.

		MR. BOXERMAN:  Thanks, Neil.  Yes, I'm going to briefly address the
points we've raised on technological and economic feasibility.

		The starting point, I think, is that the burden of proof in this
matter, on these issues, rests with OSHA to demonstrate the
technological and economic feasibility of the proposed standard on an
industry-by-industry basis.  Now, with regard to technological
feasibility, you've already heard presentations.  Comments have been
submitted by individual industries about the technological feasibility
of the PEL.  We support the submissions of the ones we've noted.

		But beyond that, and really the focus of the panel's consideration of
technological feasibility, is that exposure measurements at the level of
the PEL must be accurate and reliable.  You'll hear testimony.  You've
already seen comments on why we submit that it is neither accurate nor
reliable in this circumstance.

		With regard to economic feasibility, number one, I think it's pretty
well established a standard is not economically feasible if compliance
costs may threaten the existence, competitive structure, or long-term
profitability of an industry.  Likewise, that is one of OSHA's burdens
to establish economic feasibility in meeting that test.

		Now, the rule of thumb that OSHA has used to screen for economic
feasibility has involved this comparison in annualized compliance costs
to an industry's revenues profits, the 1 percent of revenue and 10
percent of profits.  In evaluating that analysis, we submit very
strongly, and you'll hear additional testimony about this, that the full
compliance cost of the proposed standard should be compared to the
industry's revenue and profit figures.

		I mean, after all, it is the full cost that facilities will actually
incur if they have to comply with proposed standard.  It's the full
costs that are going to impact a firm's profits, that will impact a
firm's financial viability.  It's the full costs that are going to
affect competitive structure in an industry, long-term profitability.

		It simply belies common sense to use anything other than
full-compliance cost.  As Lee even analyzed what OSHA has done, it has
excluded a very significant portion of the cost in its analysis.  And
Mr. Waggener and Mr. Sessions will discuss that further at length.

		With regard to the specifics of the analysis that we have done, in
comparison to OSHA, who estimated for 19 of the general industry
sectors, an incremental compliance cost in the range of $100 million. 
After making adjustments and corrections to OSHA's cost model, URS has
estimated an annualized compliance cost in those same 19 general
industry sectors to be $6.1 billion on a full-cost basis and $4.7
billion on an incremental-cost basis.  That's a combination of the
engineering cost and the cost associated with provisions.

		Now, there are a number of reasons that Mr. Waggener will summarize
in his testimony.  We've listed a few of them on this slide, but I'll
leave it to his expertise to provide those explanations to you in more
detail.

		So what's the result of this gross underestimate by OSHA of the cost
associated with this standard?  We submit that it's the economic impact
assessment as a result is not credible.  It's predicated on a gross
underestimated compliance cost.  But, moreover, as we'll hear further
from Mr. Sessions, it's comparison to revenue and profits from a time
period as many as 14 years ago.  We would submit that's not
representative of how that standard is going to be implemented, if it is
implemented, going forward.

		So when the proper cost numbers are used and an appropriate
methodology is applied, we have found that the rule of thumb for a
profit and revenue is exceeded in just about every general industry
sector.  Indeed, for many industry sectors, the ratio to profits is so
high that the proposed standard can be judged to be economically
infeasible without further analysis.

		MR. KING:  I'm going to just turn back just before we conclude to the
issue of measurability.  The data that OSHA and others have presented in
this proceeding show that employers will not be able to obtain accurate
and reliable measurements when respirable crystalline silica levels are
at or below 50 µg/m3.

		OSHA relies on a March 2013 study conducted at its Salt Lake Technical
Center.  But that study is not representative of precision in results
from real world sampling and analysis.  In fact, it's not even a
representative of precision in analyses conducted at the Salt Lake
Technical Center itself.

		It fails to account for several sources of analytical error, and it
does not reflect inter-laboratory variability.  Moreover, it is far out
of sync with the variability reflected in results from the American
Industrial Hygiene Association's PAT Program, where the effects of
interfering minerals and inter-lab variability are taken into account.

		Confirmation of this point is provided by the results of a blinded
performance study of five commercial labs sponsored by the ACC
Crystalline Silica Panel.  In that study, which will be described at
greater length later this afternoon by Mr. Bailey and Dr. Lee, the
analytical results for respirable crystalline silica filter loadings
that are equivalent to eight-hour personal samples collected at exposure
levels of 25 and 50 µg/m3 clearly were not sufficiently accurate and
precise to allow reliable determinations of whether the proposed PEL and
action level were being exceeded.

		In conclusion, our view is that OSHA has not carried its burden of
showing that there's a significant risk at the level of the current
general industry PEL of 100 µg/m3, or that any such risk would be
substantially reduced by lowering the PEL to 50 µg/m3.  Nor has OSHA
shown that the proposed standard would be economically feasible in
general industry, or that worker exposures at a level of 50 µg/m3 and
below can be reliably measured within an acceptable degree of accuracy
and precision.

		With that, we're going to turn the table over to Dr. Peter Morfeld,
who will discuss the issues of epidemiology.  Peter, why don't you come
up here?

		(Background comments.)

		JUDGE SOLOMON:  Technical difficulties beyond our control.

		MR. KING:  We insist on more time to allow for this.

		(Laughter.)

		UNIDENTIFIED SPEAKER:  It's not our computer.

		DR. MORFELD:  So yes, Your Honor, ladies and gentlemen, colleagues, my
name is Peter Morfeld.  I'm head of the Institute for Occupational
Epidemiology and Risk Assessment of Evonik Industries, who is in a
public-private partnership with Cologne University.  So I'm
simultaneously at the university and lecturing there.  I'm a member of
the medical faculty.

		I'm the field for dust epidemiology for more than 20 years.  I
submitted already a detailed written testimony on 7th of February of
2014.  My comments deal mainly with the mechanism of action, lung cancer
and silicosis, lung risks and threshold, silicosis risks and thresholds,
renal disease mortality and nonmalignant respiratory diseases, measures
for a variation of health effects.  And I will summarize some of the
main points.

		First, I'd like to talk about mechanism of lung cancer.  There is one
current and very interesting paper published by Paul Borm and colleagues
in 2011 in the peer-reviewed literature, which is missing in OSHA's
documentation, which makes clear what the potential mechanism may be how
interested in silica dust can induce lung cancer.

		So this paper makes clear, by a comparison and review of all evidence
available, that the mechanism that is identified is secondary
inflammation-driven genotoxicity.  This is indirect via inflammation. 
In particular, there are no crystalline silica particles found in the
nucleus of the cells.  There is nothing going on with particles in the
epithelial cells inside the lung.

		So these findings from the mechanistic point of view clearly argue in
favor of a threshold for any causal association between silica exposure
and the risk of lung cancer.

		I'd just like to note that this is widely accepted in the scientific
community.  I just cite here two longer papers Scribe (ph.) 1999 based
on MAK commission in Germany and the SCOEL 2002, which is the threshold
limit value committee of the European union.  If I understand this
correctly, OSHA appears to share this view, so in the proposal, we are
making a passage about that.  I understand this as a support of this.

		So from this point of view, I'd like to look at lung cancer
epidemiology.  Silicosis, I think this is clearly a primary effect of
excess exposures to respirable crystalline silica, resulting from
inflammation in the lungs.  This is the same principal mechanism that is
also discussed for the risk in lung cancer.

		There is absolutely no doubt that silicotics are at advanced risk of
developing lung cancer.  There is still a lot of debate how silicosis is
linked to lung cancer mechanistically.  The role is unclear.  There are
a lot of papers about this.

		It could be that silicosis is a kind of biomarker in-between, between
the exposure and the lung cancer.  It can, of course, also give us
information about this condition.

		Importantly for the discussion, when you look through all the papers
published, it appears that the subjects who did contract silicosis
appear to have no excess risk of lung cancer.  Though this was
demonstrated in a meta-analysis and, in particular, this is so when you
take smoking habits into account.  So evidence is still missing that
crystalline silica dust per se, without having developed silicosis, is
producing an excess in lung cancer risk.   

		I know that there's a new Chinese study published in 2013, which is
cited in some papers as giving new information about the relationship. 
I'd like to inform you that there is a major drawback with this 2013
study.  This study did not adjust for occupational confounders. 

		Why is this important?  There is a case control study published on the
same study group, Chen et al., 2007, that made very clear that pottery
workers were exposed to a higher level of PRHs.  Some iron and copper
mines are also included in this new study at higher levels of PRHs --

		In this case control study, the authors showed that they showed a
clear association with crystalline silica dust exposure and the lung
cancer mortality.  But this association is gone after adjustment.  

		So it is unclear to me why this paper is not really discussed in the
Liu et al. and why the authors did not address this point.  This is
something that is already under consideration among the different author
groups who have dealt with this study.  

		And so, I'd like to inform you that the co-authors tried a reanalysis
to try to clarify this confounder problem.  So without having clarified
this major problem, I think it is really premature to draw any
conclusions from this study, as done in the Seened (ph.) et al. paper
that was published recently.

		One pivotal study that is really important, if you discuss lung
cancer, is the Steenland et al., 2001.  Kyle Steenland pooled 10 --
cohorts, more than 60,000 workers.  And looking at the results, he
judged that crystalline silica dust is only a weak carcinogenic in
comparison to well-established ones when looking at milligrams per cubic
meter.

		However, there is a lot of heterogeneity involved.  This heterogeneity
cannot be fully accessed statistically by doing it in tests.  You have
to look into details of comparison.  And it's easily seen that there is
a big heterogeneity within the study when you compare the aboveground
workers with the underground workers.  The underground had much less
risk in lung cancer.

		If you put this together, then this study demonstrates excess risk
convincingly only at high exposures greater than 6 mg per cubic meter
years.  Furthermore, the SMR is rather low.  It's in 20 percent excess. 
This study does not account for smoking.  And, of course, this is no
silicosis information involved, so the role of silicosis is unclear.

		We have a new study now from Germany from 2012, which is of the same
size, almost 60,000 German uranium miners.  What is very interesting is
that this study can rely on a very good assessment of exposures, and
there were extensive side-by-side measurements done with original
historical equipment so that you can really see what happened in the
'50s and in the '60s.

		And after adjustment for a radon and arsenic exposures, a piecewise
linear spline function fit to the data showed a naught at 10 mg per
cubic meter years.  This means this study provided evidence of a
threshold at 10 mg per cubic meter years and supports the view that
excess lung cancer risk occur only at high crystalline silica exposures.
 This study also is not mentioned in the OSHA document.

		What is going on in the European discussions?  Based on the
considerations of mode of action, I've discussed first so the Borm paper
and others, and the clear experience of elevated lung cancer among
silicotics, but no such observations among workers without silicosis. 
The German committees, the U.K. -- the HSE -- and the European committee
argued that minimizing silicosis could also minimize lung cancer risks
due to crystalline silica dust.

		OSHA presented estimates that 45 years of occupational exposure to
crystalline silica dust, at the level of 0.1 mg/m3 will increase the
risk of lung cancer by an amount between 13 of 1,000 to 60 of 1,000.  So
because these analyses ignore the concept of thresholds and, of course,
do not take into account the role of silicosis, I believe that these
estimates are unreliable.

		Now I'd like to go to silicosis.  It is clear that no other malignant
health effect is associated so clearly with crystalline silica dust
exposure and specific as silicosis.  So if you look into OSHA's
document, OSHA says that there is little quantitative data available
with which to estimate a threshold exposure.

		And they point at the paper from Eileen Kuempel from NIOSH in 2001,
who used rat-based toxicokinetic-toxicodynamic models.  I'd like to
comment that these models, of course, are based on what we call
transitional toxicology.  This kind of branch is critically discussed by
recent papers.  Can we really rely on these translations from rat
experiments into humans?

		Importantly, Steven and Patel (ph.), in their analysis, they tried to
use this model of Kuempel and showed that there is no better
relationship to lung cancer or silicosis than modeling, as it's usually
done in epidemiology with cumulative exposures.  In particular, this
study did not search for concentration thresholds.

		I'd like to discuss whether this is really the only information we
have about the potential threshold for silicosis.  I'd like to show some
slides about a new study we did in Germany about a large porcelain
cohort with workers employed at more than 100 porcelain manufacturing
plants who attended a medical screening program in 1985 up to 1987.

		Silicosis assessment was based on more than 120,000 radiographs.  We
defined the endpoint as profusion score at least 1/1 ILO 1980 on the
basis of a new consensus reason. 

		So here is an overview of the dust exposures.  More than 17,000
workers were involved.  Just 40 incident cases popped up.  And what you
see, the mean of the cumulative exposures measured in milligram per
cubic meter years is at 1.5 among the controls, so the non-silicotics. 
And it is, what, 9 among the incident cases, so very clear relationship
between the dust exposure and the onset of silicosis.

		If you model this, it is usually done in most of the papers by a
so-called loglinear model, you get something like this, a very steep
increase at the beginning, across the exposure scale.  And if you read
this, even at an exposure of 1 mg per cubic meter years, the log, which
is at the y-axis, the log is about 1.5.  And so you have something about
an eight-fold risk already, up to 1 mg per cubic meter years.

		Most of the analyses stop here and say, okay, we have clear evidence
of a dose response, and such curvature shows us that there is even a
large increase of rates at the low level range.  Unfortunately, this
kind of modeling forces the curvature into the results and doesn't allow
to be more flexible and to show whether this is really a steep increase
at the beginning.  

		As with a couple of other papers, we also use them, restricted to
explants, one of the procedures to check for non-monotonic increase, and
to see that there is a kind of little dip at the beginning after 2
mg/m3.  Used other procedures that are also able to model non-monotonic
increases.  You see this fractional polynomial is almost the same
finding, up to 2 mg per cubic meter years.  There is no increase in risk
to see.  So this is motivation to search for a threshold. 

		Now, how to do it, this is quite interesting to see that almost all
regulatory groups in the world look for thresholds and limits.  But in
epidemiology, the studies do not address this question usually.  So
there are just very rare studies around that try to estimate these
thresholds directly.  The problem is how to do it.  There are procedures
available.

		I'd like to present this very briefly to you, that this is available
and can really be used to help the regulatory agencies.  One procedure
to do is the so-called likely profile approach.  You define a grid of
potential threshold values and sequentially compress the exposure by
these potential values.  The best fit stop points at the threshold value
estimate.  So this is very abstract.

		I will run through a few slides so that you can see what this
procedure is doing.  This is a simple example, with X as exposure, from
0 to 5, with no increase in Y.  The responses are 0.  After 5, it is
increasing linearly with a slope of 1 with some error about it.

		If you look at this, this is the graph showing the points I
constructed.  If you draw a line through it, going through the origin,
you see that this is a worse fit.  Now, we compress the exposures by 1,
but do not move them to the negative side. 

		This is the next picture.  We subtracted 1.  But if the result is
negative, the point sticks at 0, so this is a little bit better.  If we
subtract 3, it's getting better again.  If we subtract 5, this is where
the threshold was in our curve, it's a perfect fit.  If you do too much,
it can be worse again.  Now, this is 9.

		We can do this in a formal way, measuring the kind of variance or the
error of the model with this statistic, and where the statistics, the
error is smallest, you get the indication where the threshold is.  This
is exactly 5.

		This machinery identified the threshold we have seen before.  This can
be used now in epidemiology in the more complex situations.  What we did
is, with the German porcelain study, we applied this machinery to the
concentration, not to the cumulative exposure, to the concentration data
we have, and then rerun all the modeling and recalculated all the
cumulative exposure after having modified the concentration data.

		What you see is that you get a rather similar curve than shown before
by the illustration.  This ends up at the best estimate at around
2.5 mg/m3, clearly significant because the non-threshold model is here,
the one with the zero.  So this means it's clearly outside the
non-threshold model from the confidence interval.

		So, to summarize this, we found with this machinery, from the German
porcelain study, a threshold estimate for the concentration of
respirable crystalline silica at .25 mg/m3 with a lower 95 percent
confidence limit at .15 mg/m3.  I think that every quantitative risk
assessment of the association with crystalline silica dust exposure and
silicosis or lung cancer should allow for the existence of threshold and
really try to estimate them.

		So all other studies around don't do this, so they are not very
informative about what is really happening at the lower exposure end,
and what in particular is happening when you think about the scale in
milligram per cubic meter.  Steenland and Deddens, they also searched
for threshold for cumulative exposure in the pooled cohort study.  These
were just thresholds for the cumulative things.  They never looked into
detail into the concentration thresholds, so they are not very
informative about this, what we are really interested in.

		Therefore, I think that the risk estimates given for silicosis,
ignoring the threshold effects, the potential threshold effects, are not
very convincing.  We do not know whether these numbers are really
reliable because we do not know what will happen if all shifts are
really kept below 0.1 mg/m3.

		Renal disease mortality, just a short glance on that.  One important
study is the pooled analysis from Kyle Steenland from 2002.  There are
three studies pooled, all U.S. studies.  They saw a monotonically
increasing exposure-response relationship.  One of the studies, Attfield
and Costello, should be replaced because we have a new study, an update
of this by Vacek et al. in Vermont granite workers.  This update could
find no association, so this should be replaced, I think.

		Another point I don't understand is that this pooled study on renal
disease mortality was based on the pooled lung cancer analysis that Kyle
Steenland did when he was at IARC.  This was on the 10 cohorts, but only
three cohorts were pooled for renal disease mortality.  It's unclear
why, what about the other seven cohorts?  Why did they not get into this
pooled analysis?  This, of course, may have introduced study selection
bias.

		And if you go into another further analysis of this stuff done by
Steven and Patel in 2003 and 2004, they concluded that this amount of
data is insufficient to provide robust estimates of risk.  This is from
the original author, so it was quite surprising that OSHA now is using
these data against the warning of the principal investigator and
estimated excess risk due to renal diseases.  I think this is an
over-emphasis.

		Now, nonmalignant respiratory disease, a short look into that.  One
major study and important for OSHA's assessment is, of course, by Robert
Park et al. in California in diatomaceous earth workers.  And what I
think is critical in this study is that the authors truncated the
cumulative dust exposures at about half of the maximum. 

		The high exposures showed no effects, so there were no cases, and they
were ignored.  If you do this, it means, of course, that your exposure
response estimates are biased effort.  It is better if you think you'd
like to deal with healthy worker survivor effects or things like that
with this kind of trick, then it is better to use particular procedures
which are well-developed to deal with this, like GST NIOSH.

		There are some problems with smoking, of course, because the data on
this is sparse, and which is important again, Robert Park did not switch
for the threshold effects.  It's unclear to me whether this is really
reliable at the low level of exposures.

		We know that, as I already mentioned, the Vacek et al. did an update
of the Attfield and Costello study and performed a nested case control
analysis, and they could not see any clear effects of respiratory
crystalline silica dust in the nonmalignant respiratory diseases.  So I
think because this updated study could not confirm the findings reported
before in Attfield and Costello, this ought to shed some doubt on this
relationship.

		This taught, in principle, the findings of Vacek are consistent with
the results of diatomaceous earth worker study by Checkoway et al.  They
showed that the SMR for the nonmalignant respiratory diseases were
significantly elevated in only workers who started work very early,
before 1950 when the exposures were rather high.  

		If you use the unlogged model that is preferred according to Park et
al., 2002, then you see that there are definitely higher risks for
nonmalignant respiratory disease mortality only in the highest exposure
category.  So this all is compatible with a threshold.  This wasn't
checked for it.  They didn't search for a threshold, but the results we
have do not rule out the thresholds.

		There's another point, of course, that makes it difficult with a
nonmalignant respiratory diseases.  We know that general dust exposures
can cause also these kind of disease: chronic bronchitis, emphysema,
pneumoconiosis.  Recent reviews on this, given Cherry out of the U.K.,
and when we look at what the toxicological people say, they argue in
favor of a threshold effect for these endpoints. 

		There is a paper by Paloo (ph.) in 2011, which was evaluated very high
in the toxicology societies.  Of course, the effects of the general dust
exposures were not taken into account simultaneously with the cross-dust
exposure.  It's unclear how much the risk estimate might change if
adjustments were made.

		So OSHA projected that 45 years of occupational exposure to
respiratory crystalline silica at a level of 0.1 mg/m3 will result in 83
excess deaths from nonmalignant respiratory diseases per 1,000 workers. 
I think this is unsupported because there is no threshold estimates
taken into account.

		Some discussion and conclusion.  I would like to emphasize one point. 
The epidemiological studies on which OSHA relies focused on cumulative
exposure.  This is correct as this is the driving variable for most of
the diseases.  But we are here interested in the concentration.  And
concentration thresholds have to be taken into account additionally.

		So the problem with the cumulative exposures is that the same value of
the cumulative exposure can be reached when you are exposed over a long
term to low concentration or over short term to higher concentration. 
So this is a clear drawback of this metric.  In particular, if you read
from the mechanistic papers, that is very important for health effects
caused by crystalline silica exposures, if the concentrations are high
or not.

		Most of the authors of epidemiology studies perform simple calculation
to derive statements about concentrations, although they, within the
studies, only deal with cumulative exposures.  So for instance, you see
in Steenland and Ward, 2013, where they said for a comparison exposure
of the 45-year working lifetime at the current OSHA standard of 0.1 mg
per cubic meter years, resulting in 4.5 mg per cubic meter years.  

		You can see that Kyle Steenland here is really changing the scale.  He
is no longer talking about milligrams per cubic meter, but he is talking
about milligram per cubic meter years as the OSHA standard.  In
principle, this calculation is right.  But because the same cumulative
exposures can occur under varying exposure intensities, this has to be
taken into account and is important for all judgments.  Then, of course,
this leads us to a serious issue then if you do such simple
recalculations.     

		We know from the OSHA document that there is a lot of overexposure
going on.  So I'm not sure whether this is true.  I just take it for
granted as it is written up in the documents.  I was quite surprised
seeing that about 25 percent of all workers are calculated to have
exposures currently higher than 0.1 mg/m3.  And even about 12.5 percent
of the workers are higher than the value we identified in the German
study as the limit value for the development of silicosis, 0.25 mg/m3. 

		So if you, of course, take the German study, and I try to evaluate
this situation from our results, I would say, yes, of course.  I would
expect silicosis cases under this situation.  And because the situation
was worse in former times, it is clear that we will be confronted with a
lot of diseases among the workers exposed to silicosis, exposed to
crystalline silica dust because of these overexposures.  That's very
clear.

		So if you take all this together, then of course, I think there is no
convincing epidemiological evidence that when the occupational
crystalline silica exposures are really maintained at or below a level
of 0.1 mg/m3 in every shift, that we have to expect to see any excess
risk of silicosis, lung cancer, renal diseases, or nonmalignant
respiratory diseases.  Thank you.

		MR. KING:  Thank you.  Paul Scott will now discuss measurability
issues.

		MR. SCOTT:  All right.  Today, I would like to discuss why the
analytical methods for respirable crystalline silica are inadequate to
accurately measure exposures at the proposed PEL and the proposed action
level of 25 µg/m3.

		I'd like to discuss why justification presented in the Preliminary
Economic Analysis is misguided and inaccurate.  This is primarily
because it fails to adequately characterize the various sources of
sampling and analytical error associated with the sampling analysis
methods.  The result is that OSHA has not made this a supportable
showing that silica exposures can actually measure with acceptable
precision at the concentrations of the proposed PEL and action level.

		By way of introduction, my name is Paul Scott.  I am a qualified
environmental professional with over 23 years of experience in applying
statistics in the areas of environmental and occupational health.  I
have specific experience with statistical methods for evaluating
accuracy, precision, and reliability for total and hexavalent chromium
for environment air sampling, as well as PCBs in sediment.

		The existence of sampling and analytical methods that can accurately
and precisely measure respirable crystalline silica at the proposed PEL
or action level is a first step in determining the technological
feasibility of the rule.  If the sampling analytical methods do not meet
the necessary standards of accuracy and precision, then there is
uncertainty associated with trying to determine whether our
concentrations are above or below the PEL or action level.

		If the methods do not meet the necessary standards of accuracy and
precision, then it is difficult to accurately determine the
effectiveness of engineering controls needed to meet the proposed PEL. 
I just want to point out in the next couple of slides an overview of my
comments.  I'll let you take a look at those, but I'm going to discuss
each of these in detail.  

		Starting with my first comment here on limits of quantification, the
limits of quantification for various x-ray diffraction and infrared
methods range from 5 to 10 µg.  These limits of quantification are
based on ideal conditions where substances that interfere with the
analysis are not present.

		Quantification in the presence of interferences increases the
potential error because additional measurements have to be made to
compensate for changes to the background under the measurement peak or
changes to the peak profile because of coinciding peaks.

		In most industrial settings, materials that interfere with the
analysis of quartz and crystallite are often present.  For x-ray
diffraction methods, the presence of iron, mica, feldspar, and other
minerals will interfere with the quartz peak and degrade detection
limits significantly.

		For infrared methods, silicates and other minerals can interfere and
affect the accuracy of analysis.  The range of reporting detection
limits in the paper by Eller et al. of 5 to 50 µg reflects some of the
variability associated with these different matrices.  But it's limited
because it's based on test data from the PAT Program and not actual air
samples collected in industrial setting.

		Depending on the setting, it can be expected that the impact of
interferences may lead to significantly higher limits of quantification
than 10 µg, depending on the concentration of particulates containing
substances known to interfere with x-ray diffraction and infrared
methods.

		My second comment is focused on use of sampling analytical error
versus precision in the PEA.  In the PEA, OSHA defines sampling
analytical error, or SAE, as a 95 percent confidence limit using the
critical value of 1.645.  This is the standard normal distribution value
for probability of 5 percent and is typically associated with a
one-sided confidence limit.

		However, unacceptable accuracy and precision measurement of plus or
minus 25 percent in 95 percent confidence indicates the need for a
two-sided confidence limit.  Also, any measure of sampling and
analytical variability should be able to evaluate whether you have an
under- or over-estimate of the true air concentration.  It is also
indicative of the need for a two-sided confidence limit.

		Using a one-sided confidence limit will result in an underestimate of
the actual variability because it only accounts for the variability
associated with values that are overestimates of the true concentration.

		Also in the PEA, precision is defined as a confidence limit using the
critical value of 1.96, which is the standard normal distribution value
for a two-sided confidence limit and is based on probability of 2.5
percent, which is half of 5 percent, for a 95 percent confidence limit.

		And in the preamble of the proposed rule, OSHA uses the SAE and not
the precision when making statements about how sensitive and precise the
sampling analysis methods are.  However, this is inappropriate because
it's only a one-sided confidence limit, and not a measure of both, of
two-sided, both underestimates and overestimates of variability.

		And my third comment is on OSHA's estimates of SAE and precision
reports concentrations at the proposed PEL and at the proposed action
limit.  Essentially, they're incorrect because they don't account for
all sources of sampling and analytical variability. 

		In the PEA, OSHA defines the precision as a function of the analytical
error characterized by the coefficient of variation associated with
analytical method, this is the CV1, and the coefficient of variation
associated with sampling error, CV2.  For the CV1, relative standard
deviation from a study by the Salt Lake Technical Center, which involved
an evaluation of 10 replicate samples at quartz loadings of 21 and 40.6
µg, and CV2 of 5 percent associated with sampling pump flow variability
were used to estimate the percentage SAEs and precisions.  Our release
estimates of CV1 and CV2 are not adequate because they failed to
characterize all sources of sampling and analytical variability. 

		Other sources of sampling variability, for example, include
variability between cyclones of the same type, performance of the
cyclones of different dust particle sizes for a single dust species,
with different dust species and with real-world multi-species
environments.  Also, effect of loading and cleaning on the cyclone
performance and the effect of electrostatic properties of the dust.

		There are data, however, that exist to quantify two additional areas
of sampling variability.  The first is inter-sample sampler variation
based on samplers of the same type.  In a study by Gottman and
Shrenofnat (ph.), which is also discussed in the ASTM method for
respirable dust and workplace atmospheres, they looked at variability
between samplers of the same time and estimated an average relative
standard deviation of 6 percent due to this source of error.

		In addition, the ASTM method also discussed variation using different
types of samplers in the same environment and found this to be about 5
percent.  If you estimate a modified coefficient of variation or CV2
based on these additional sources of error, you end up with a value of
9.3 percent.

		This figure presents the precision estimates associated with the Salt
Lake Technical Center evaluation as proposed action level and PEL, using
these alternative assumptions for sampling error or CV2.  Also, the line
indicates plus or minus 25 percent, NIOSH standard for precision and
accuracy.

		As can be seen from this figure, the estimated CV1 -- from Salt Lake
Technical Center 2013 evaluation is used for the corrected CV2 of 9.3
percent.  This value is 25 and 23 percent respectively for the action
level and the proposed PEL.  This is an increase in variability of about
6 percent.

		The use of the Salt Lake Technical Center evaluation to estimate
analytical error by OSHA fails to account for several sources of
analytical error that include the effective differences in particle
sizes on the analysis of silica by x-ray diffraction and infrared
methods; the effect of potential interferences on these two analytical
methods; the effect of inter-laboratory differences in sample
preparation methods, calibration standards, and the implementation of
the x-ray diffraction and infrared methods; and, finally, the effect of
inter-laboratory differences caused by differences in analysts and
sample preparation methods, as well as variability in the actual
analysis run.

		By failing to account for these additional sources of analytical
error, OSHA calculates an overly optimistic value for precision based on
artificially low value of analytical error.  To be specific, with
respect to the Salt Lake Technical Center 2013 evaluation that is
presented in the PEA, it does not account for the variability associated
with the analytical method or the differences in precision between two
methods, which have been shown to be important in the published
literature.

		Also, it doesn't account for inter-laboratory variability, which has
also been identified as major contributors to sampling analysis error. 
Also, the sample size of 10 samples per loading is fairly small, given
the purpose of the study, which is to characterize analytical error for
two methods used by at least about 57 labs.

		Also, it fails to account for the variability associated with the
existence of potential interferences by using alternate sample matrices.
 The variability associated with the silica being in different matrices
can have a significant impact on the analytical error and precision as
demonstrated by Eller et al. analysis of the IH PAT data from 1990 and
the most recent analysis of 2003-2013 PAT data.

		Interestingly, the results of a similar evaluation of accuracy and
precision performed by the Salt Lake Technical Center in 2010, using
twice as many samples per loading level, are included in the OSHA docket
but not discussed in the Notice of Proposed Rulemaking or the PEA.

		While the documentation for this study is limited, it appears to
include the analysis of two sets of 10 samples of cristobalite and two
sets of 10 samples of quartz by x-ray diffraction methods at loadings of
around 20 and 40 µg.  Relative standard deviation in associated
accuracy and precision values were estimated for each set of 10 in the
same manner as the Salt Lake Technical Center evaluation that was
presented in the PEA.		

		Based on what is in the docket, the estimates of the analytical error
for the study are much higher than those from the evaluation presented
in the PEA.  For quartz, the estimate of analytical error is 16.1
percent at 20 µg and 12.8 percent with 10 samples at 40.  Using those
estimates for 20 µg loading, the precision ranged from 33 percent to 36
percent, depending on the value of sampling error that is used.

		For the 40 µg loading, the precision ranges is from 27 percent to 31
percent, depending on the value of sampling error used.  These precision
estimates are higher than those presented in the PEA with a smaller Salt
Lake Technical Center evaluation, with precision at 20 µg being 14
percent higher, and the precision at 40 being 10 percent higher, for a
sampling error of 5 percent.

		These data indicate that the estimated precision at Salt Lake
Technical Center couldn't be much higher than the values estimated using
the 10 samples per loading discussed in the PEA and are well above the
25 percent level considered to be acceptable.

		Another source of data that could be used to characterize the
analytical variability is data available from the AIHA IH PAT Program
administrated by industrial hygiene professionals from the AIHA and
NIOSH over the last three years.  This program serves as a standard for
certification for silica analyses for all IH laboratories in the U.S.
and many others around the world.

		For each round of the program, samples are prepared.  The silica
loadings are embedded in different matrices on a rotating schedule and
sent to labs four times per year.  The matrices are typical of what
might exist in the samples collected from industries that monitor silica
exposures.  

		These matrices include calcite, coal dust, talc dust, and a
combination of coal and talc dust.  Each laboratory analyzes a sample
based on standard methods that can include x-ray diffraction, infrared
and colorimetric methods, though very few labs have used colorimetric
methods in recent years.  Thus, the data associated with the PAT Program
can account for variability due to intra-laboratory differences over
time, and analytical method differences and differences in sample
matrices.

		The PAT data are well suited to characterize the analytical error for
analysis of silica because the program evaluates several sources of
analytical error that the Salt Lake Technical Center evaluation does
not, specifically, variability due to inter-laboratory differences,
analytical method differences, and differences in sample matrices.

		The data from the IH PAT Program are inherently more variable than
data from the Salt Lake Technical Center evaluation because of the
additional source of variability characterized by the data.  Compared to
the Salt Lake Technical Center evaluation, an analytical error estimate
for the dataset would provide a better characterization of analytical
error and results that are more representative of real world conditions.

		One of the drawbacks of the PAT Program data is that the lowest silica
loading used in the program is typically around 40 µg, which
corresponds to the proposed PEL.  However, it's been demonstrated that
relative standard deviation increases with decreasing silica loading for
x-ray diffraction and infrared methods.  Any estimate of relative
standard deviation calculated from the lower end of the silica loadings
used in these data will be lower than that, than the actual value that
might be estimated at 40 µg and certainly for 20 µg samples that
correspond to the proposed action level.

		Another drawback is that after round 158, the relative standard
deviations for labs with relative standard deviations greater than 20
percent were set equal to 20 percent.  This will buy us the overall
relative standard deviation low for any round after 158, because labs
that had relative standard deviations greater than 20 percent would be
lowered to 20 percent.

		This figure presents precision estimates for sampling error of 5
percent, calculating from the Salt Lake Technical Center 2013
evaluation, the Salt Lake Technical Center 2010 evaluation, the PAT
Program data that were included in the OSHA docket, and the PAT data
from 2003 to 2014 -- 2013, excuse me.

		The PAT data that are available to estimate for all standard deviation
associated with silica analysis in the OSHA docket are from rounds 156
to 180, and range in time from April 2004 to February 2010.  In order
to focus on mass loadings that's close to the 20 and 40 µg, as
possible, this dataset includes only samples with loadings less than 70.
 That's about 23 out of the 100 samples available in those rounds.

		Recently, Harper et al. evaluated the PAT data for 2003-2013 in a
manner similar to that used by Eller et al., 1999, who looked at the
data from 1999 and 1998.  This dataset included data from rounds 152
through 194.  They estimated a relative standard deviation for this data
that included all rounds of data and all mass loadings, ranging from 40
to 250 µg, of 20.8 percent.

		The precision estimates at the proposed action level increased from 19
percent for the Salt Lake Technical Center 2013 evaluation, to 33
percent using the 2010 data, to 37 percent for the PAT data in the PEA,
to 42 percent for the most recent PAT dataset.  A similar increase can
be seen for the proposed PEL, going from 17 percent to 27 percent to 37
percent and then finally to 42.

		It's clear from these analyses that the precision becomes poorer as
more sources of analytical error are accounted for.  The precision is
best for the data from a single laboratory, analyses pure silica samples
from one preparation and analysis method.  It is poorer for the data
from multiple sample matrices that are analyzed by multiple laboratories
using different sample preparation and analysis methods.

		On my next slide, I show a similar trend in precision using a sampling
error of 9.3 percent.  Once again, for both the Salt Lake Tech Center
2013 evaluation, the Salt Lake Tech Center 2010 evaluation, the PAT data
from the OSHA docket, and the PAT data from 2003-2013.  

		Finally, another source of data that can be used to characterize the
analytical error and the overall precision are results from a respirable
crystalline silica round robin performance testing program sponsored by
the American Chemistry Council Crystalline Silica Panel.  This study
will be discussed in more detail by my colleagues after my talk.

		As an overview, filters containing three different loadings of
respirable quartz dust were sent over a period of several months to five
different AIHA-accredited laboratories for analysis as part of a blinded
testing program.  Each of these five laboratories used x-ray diffraction
method to analyze for crystalline silica.

		As part of the blinding process, the laboratories were not informed
that they were participating in the performance testing program. 
Filters were submitted with a standard chain of custody forms as if they
were collected during ordinary workplace monitoring of silica exposure
by commercial customers. 

		An analysis of the variance of the round robin data indicated that
there were significant differences with respect to loading level. 
Because of this, the relevant standard deviations were calculated for
all the data at 20 and 40 µg loadings respectively, and separate
estimates for precision were calculated.

		The estimate of analytical error at a loading of 20 µg was 37
percent, and precision values ranged from 72 percent to 74 percent
depending on whether a value of 5 or 9.3 percent was used for sampling
error.

		At a loading of 40 µg, the estimate of analytical error was 32
percent.  The precision values ranged from 63 to 65 percent, depending
on the value of sampling error.  These estimates of precision are
substantially higher than those estimated using the Salt Lake Technical
Center 2013 evaluation for the PEA and the Salt Lake Technical Center
2010 evaluation in the two IH PAT datasets.

		This figure shows the difference between the Salt Lake Technical
Center 2013 evaluation and the estimates from the American Chemistry
Council round robin study.  Precision estimates associated with the
proposed action level and proposed PEL using the round robin study are
almost four times larger than those estimated using the Salt Lake
Technical Center 2013 dataset.

		My fourth comment discusses a miscalculation that was presented in the
PEA.  In the preamble of the proposed rule, OSHA states that most
laboratories achieve good agreement and results for samples having
filter loads just above 40 µg quartz.  This statement is based on
analysis of a number of laboratories and sample results that were within
plus or minus 25 percent of the reference value in the PAT Program for
rounds 156 to 165.

		It is unclear why OSHA limited its analysis to these data from these
rounds, which are not the most recent.  Whatever the reason, the PEA
represents the percentage of laboratories reporting values within plus
or minus 25 percent of the reference value in these rounds.

		OSHA states that for all reference values, laboratories achieve good
agreement because 80 percent are within plus or minus 25 percent of the
reference value.  In addition, OSHA states that for reference values
less than 70 µg, 81 percent of the laboratories are within 25 percent. 
However, recalculation of the results from the table in the PEA
indicates that the actual percentage of the labs within plus or minus 25
percent is not 81 percent, but is actually 73 percent.

		Of the reference values of 70 µg, 83 percent of the labs were within
plus or minus 25 percent of the reference values.  This demonstrates
that precision deteriorates at lower silica loadings, and that three out
of four labs achieve good agreement results for samples with less than
70 µg of silica.

		My fifth comment discusses how past data from the PAT Program has been
used to characterize differences, method differences and
inter-laboratory variability.  

		In Chapter 4 of the PEA, OSHA states that the PAT data were not
suitable for characterizing the method differences and inter-laboratory
variability because PAT sample preparation errors may contribute to the
analytical error because reference value is based on either the average
of a subset of laboratories or all the laboratories reporting for the
round, depending on the round be evaluated.

		However, the PAT data had been used to evaluate the accuracy and
precision of respirable crystalline silica method in two separate
studies that had been published in the peer-reviewed literature, Shulman
et al. in 1992 and Eller et al. in 1999.  Both papers evaluated the
accuracy and precision of the different analytical methods, as well as
inter-laboratory and intra-laboratory differences in precision, healthy
analytical methods based on the different time periods of the PAT data.

		Thus, members of the scientific community have considered the PAT
Program to be suitable and useful for evaluating the precision of
analytical methods for crystalline silica.  In addition, OSHA scientists
have also used this data to evaluate accuracy and precision in the past,
in a study by Matson (ph.) et al. in 1995.    

		My sixth comment talks about use of high-flow rate sampler.  On page
443 of the PEA, OSHA states that the higher flow rate device, such as a
BGI GK 2.69, using the recommended flow rate of 4.2 L/min, can collect
more than 10 µg of dust for a one-hour sample at the proposed PEL.

		The implications of this statement is that even if the accuracy and
precision of the sampling analysis of silica does not meet the standard
of plus or minus 25 percent and a 95 percent confidence level when
silica filter loadings are 40 µg and below, there are samplers that are
capable of collecting a greater amount of silica at the same air
concentration due to the increased flow rate.

		For example, the proposed PEL, a high-flow rate sampler at 4.2 L/min
would collect 101 µg of silica for an eight-hour shift compared to 41
collected using the current Dorr-Oliver sampler at 1.7 L/min.  101 µg
silica loading is about 20 µg higher than what we collected using the
Dorr-Oliver sampler at the current PEL of 100 µg/m3 for general
industry. 

		However, there are several limitations with the use of high-flow
samplers.  First, the accuracy and precision of the high-flow rate
samplers for measuring respirable crystalline silica have not been
evaluated, although performance of these samplers for collecting
respirable particulates in quartz has been evaluated relative to
low-flow samplers in several studies.

		These studies have focused on sampling efficiencies of the respective
samplers relative to the ISO/CEN particle size convention and how the
amount of mass the individual high-flow samplers collect compared to the
low-flow sampler.  No big studies evaluated the accuracy and precision
of these samplers using the methods recommended by NIOSH for sampling an
analytical method development.

		Second, the need to evaluate accuracy and precision release methods is
important because studies by Lee et al. in 2010 and 2012 indicate that
the high-flow rate samplers tend to collect a higher proportion of
larger-size particles than the lower flow rate samplers currently use. 
Lee et al. in 2010 found that the BGI GK 2.69 high-flow sampler had a
large bias for particles with a large mass mean aerodynamic diameter.

		In general, all three high-flow samplers evaluated by Lee et al.,
2010, tended to have a substantial bias towards collecting more
respirable particulates than the low-flow samplers, collecting between
12 percent and 31 percent more mass than the low-flow samplers.

		Third, while the high-flow samplers collected more quartz mass than
the low-flow samplers, standard deviations associated with the mass
ratios, net mass ratios, were high, indicating a potential increase in
sampling and analysis error.

		Because respirable silica in occupational settings tend to have a
greater proportion of smaller size particles, and high-flow samplers
tend to over-sample larger particles compared to the low-flow rate
samplers, it seems likely that the high-flow samplers will collect a
greater proportion of non-silica particles that can potentially
interfere with the analysis of respirable crystalline silica using x-ray
diffraction or infrared methods.

		In addition, the collection of a larger mass of particles may lead to
overloading problems for environments with a high concentration of
particulate matter that contain a lower proportion of silica.  The
increase in the probability of interferences and the potential for
overloading will both create accuracy and precision problems.

		A further consideration is that the use of a high-flow sampler likely
will necessitate use of a pump with a higher flow rate than those
currently used by most industrial hygienists and dust samplers,
particularly if the industrial hygienist is collecting samples at a
sampling rate of 4.2 L/min versus current rate of 1.7 L/min using the
Dorr-Oliver sampler.

		For these reasons, before recommending the use of high-flow samplers,
OSHA needs to develop a better understanding of the accuracy and
precision of these samplers under real-world conditions, using particles
that may be found in the workplace.

		My concluding thoughts:  I would like to conclude by saying that the
estimates of precision presented in the PEA are overly optimistic and
inaccurate because they fail to account for a variety of sampling and
analytical errors, because only the Salt Lake Technical Center 2013
evaluation was used to define the analytical method error used to
estimate the precision for the analysis method of the proposed PEL and
action level.  

		The precision estimates fail to account for inter-laboratory
variability, differences in the two methods used, the effect of
procedural interferences due to the sample matrix, or the effect of the
analyst and analysis variability over time at the same laboratory.  If
these additional sources of error are included in the precision
estimates, the accuracy and precision sampling analytical methods for
silica will not meet the plus or minus 25 percent at 0.95 standard
recommended by NIOSH for these silica loadings at 20 and 40 µg that are
associated with the proposed action level and PEL.

		In addition, the use of high-flow rate samplers may not address these
problems with precision at low silica loadings because the accuracy and
precision of the high-flow rate sampling methods are unknown, and the
high-flow rate sampling methods tend to collect a higher percent of
large particles that will tend to interfere with silica methods and may
overload the sample over an eight-hour period.  Thank you.

		MR. KING:  At this point, Kelly Bailey and Dr. Richard Lee will
describe the ACC-sponsored five commercial lab performance study.

		MR. BAILEY:  Good afternoon, Your Honor, OSHA, all you folks bearing
with us.  My name is Kelly Bailey.  I am the current chairman of the
American Chemistry Council's Crystalline Silica Panel.  I work for
Vulcan Materials Company as the corporate Director of Safety, Health and
Environment.  Vulcan is the largest U.S. producer of crushed stone, and
sand and gravel.  

		I am a board-certified industrial hygienist, and I've been in the
profession for 40 years.  With me, over to my left is, Dr. Richard Lee
who is President and CEO of RJ Lee Group.  We appreciate the opportunity
that OSHA has provided the panel of witnesses to participate in these
important hearings.

		RJ Lee group is a niche laboratory specializing in asbestos, silica,
and other environmental testing.  RJ Lee Group is widely regarded as
experts in these fields.  They have been pioneers working with EPA,
OSHA, NIOSH, and MSHA in the development and improvement of methods for
the analysis of these materials.

		RJ Lee Group has worked for and collaborated with NIOSH in assisting
other countries in setting up crystalline silica analytical
laboratories.  They are also experts in the broader field of material
testing, especially with respect to the automated electron microscopic
analyses.

		Why was this done?  Why did we do this study?  Well, our testimony
describes a recent study funded by the American Chemistry Council's
Crystalline Silica Panel, hereinafter called "the panel."

		The study was designed to ascertain how accurately and precisely a
sample -- a small sample of AIHA-certified commercial laboratories
could analyze respirable crystalline silica, or RCS, masses that
correspond to the current general industry permissible exposure limit of
approximately 100 µg/m3 of air, the OSHA-proposed PEL of 50 µg/m3 of
air, and OSHA's proposed action level of 25 µg/m3.  Based on an
eight-hour work shift and a pump flow rate of 1.7 L/min, these masses
would equate to 80, 40, and 20 µg per filter, respectively.

		The objective of the study was to evaluate how labs would perform in
terms of accuracy and intra- and inter-laboratory precision at these
relevant silica loadings.  The study was limited to five laboratories
using a muffle furnace for sample preparation and x-ray diffraction for
analysis.

		The study is unique in the extent that to which it was conducted on a
blinded basis.  In a typical laboratory evaluation or performance test,
a laboratory is usually well aware that it is participating in a program
designed to evaluate the accuracy and precision that the laboratory can
achieve in analyzing samples.

		This can introduce bias into the evaluation from the very beginning. 
In the study we are describing today, a conscious effort was made to
eliminate this bias by having third parties prepare and forward the
sample filters in a manner that made them look as though they were
actual air filters from an exposure-monitoring event.

		The labs were unaware that these samples were spiked with known silica
masses and that their reported results would be evaluated for accuracy
and precision.  Dr. Lee will explain the details regarding the sample
preparation that his laboratory performed for the panel.  First, I will
quickly explain how the process was carried out.

		First, polyvinyl chloride filters, PVC filters, were prepared using
Min-U-Sil-5 as the crystalline silica source.  Respirable feldspar and
respirable kaolin were selected as interfering minerals for the mixed
matrix filters prepared at a 50/50 mix with silica.  PVC filters with a
0.8 micron pore size were used to receive the deposited material at the
sample weights of 20, 40, and 80 µg.

		As I noted earlier, the five labs selected for the study used a muffle
furnace for sample preparation and XRD for analysis.  They were
geographically dispersed, as well, representing labs of different sizes.
 All were AIHA-certified for silica analyses.

		For each lab, three sets of 10 filters were prepared.  Each set
consisted of one blank filter, three pure silica one at each of the
three reference weight levels, three with silica plus feldspar, and
three with silica plus kaolin.  RJ Lee Group forwarded these filters to
Sandler Occupational Medicine Associates, or SOMA. 

		SOMA added dummy plant codes and sample numbers to the filter
cassettes to make them appear to be normal field air filter samplers. 
Each of the five labs got one set of 10 filters three different times
over the course of nine months.  The labs reported the results to SOMA,
which forwarded the results to Dr. Louis Anthony Cox, Jr., for
statistical analysis of the laboratory performance.

		That analysis was prepared in a blinded matter, as well, in the sense
that neither Dr. Cox nor the panel was aware of which results were
reported by which lab.  This is how the study was conducted.  

		I will now turn the microphone over to Dr. Lee, to describe the
process that was used to prepare the reference filters for this study.

		DR. LEE:  Thank you, Kelly.  I, too, would like to express my
appreciation to OSHA for being able to testify at this deliberation. 
The key goals of our participation in the study were to validate the
loading methodology, and to ensure that the laboratories received blind
samples.

		The methodology for depositing known quantities of materials on
filters is well understood and practiced every day in laboratories such
as ours.  A known quantity of material is weighed out and filtered,
suspended in the alcohol or other appropriate media, and filtered onto
the appropriate filter media.

		We defined recovery in the standard form as shown in this question
here, which is just the amount deposited minus the amount measured over
amount deposited and expressed a percentage.  The RJ Lee Group measured
the recoveries, that is the amount of silica deposited on the test
filters, used x-ray diffraction, using all these scores through primary
reference standards.  

		We did attempt to compare the pre-imposed weights as a backup method
for evaluating the amount of material deposited.  But we found the
alcohol partially dissolved the filter, extracting about 70 µg, plus
or minus 20.  This caused a random weight loss which rendered the direct
measurement of weight unusable.

		It would have been ideal to use the 5 micron PVC filters that we
normally use in air sampling for silica.  Unlike air sampling, the 5 µm
filters proved to give very poor recoveries of the deposited material,
as the liquid media carried the respirable particles through the filter.
 Using 0.8 µm filters solved this problem.  However, we found that the
.8 micron filters did not dissolve completely in the THF used in the
chemical method for processing filters.  Consequently, the RJ Lee Group
had to eliminate from the study laboratories which used the chemical
method.  

		The choice of the reference material was complicated by the
unavailability of the standard reference silica NIST 1787a.  RJ Lee
Group elected to use the Min-U-Sil 5 as the reference material. 
Min-U-Sil 5 has a medium particle size, a 1.72 µm versus about 1.68 for
the NIST standard.  We used the sedimentation process on both the
Min-U-Sil 5 and on the interference materials to ensure that we did not
have oversized particles on the filter.

		As Kelly described, multiple sets of cassettes were prepared by RJ Lee
Group and sent to SOMA.  The cassettes contained pure silica or a 50/50
mixture of silica and one of the interferences.  The filters were
preconditioned, pre-weighed, and assigned a tracking number.  The
reference material was loaded on a pin on a microbalance and using an
acceptable limit of plus or minus 1 µg from the nominal weight that was
to be deposited.

		The contents of the pin were suspended in isopropyl alcohol and
deposited on a filter using vacuum filtration.  The filters are
desiccated and post-weighed.  The prepared filter was loaded in a
cassette, seated, labeled, and shipped to SOMA.  Finally, a subset of
filters were analyzed in house.  The reproducibility of the analysis was
good, and the recoveries, acceptable, averaging about 95 percent.  You
can see from the x-ray diffraction patterns that are presented
underneath, the similarity between the different replica samples.

		Finally, a set of samples was prepared, shipped to SOMA, and returned
to us, the RJ Lee Group, in a blind manner.  The estimated recovery is
because we did not have the code between the original sample and the
blind sample.  We used the nominal weights.  Using that, the estimated
recovery was about 97 percent.  We found no effect of shipping in the
weight.

		With that, I'll turn the microphone back to Mr. Bailey.

		MR. BAILEY:  So, what were the results of this five-lab performance
study?  As an industrial hygienist, I was quite surprised by the
findings.  The results I am going to share exclude those samples where
the laboratory reported the silica mass to be below the detection limit.
 

		Thirty-four percent of the non-blank filters, that is the filters with
silica loadings of 20 µg or greater, were reported to have silica
masses below the detection level.  Had these non-detect results been
included in Dr. Cox's analyses, the accuracy of precision value would
have been worse than those that I will show you.

		In this table, you can see that in two instances, the labs reported
silica masses on blank filters.  One of these was reported to be 52 µg
of silica.  None of the mean values for the sets of 20, 40, and 80 µg
silica loadings came within 30 percent of their respective reference
weights.  

		This table includes only the pure silica filters where the lab
reported the result above the LOD, or limit of detection.  When all of
the filters were there, all the filters that were there were reported
with silica levels are examined, the next table, one finds that the
results are not much better.

		Looking at the entire set of results, whereas detected silica mass was
reported, the data indicate that these labs could not reliably
discriminate amount different reference levels.  In fact, for those
filters in the higher quartile of the reported results, all of which
should have come from the 80 µg reference level, there was a 50 percent
chance that the filter came from the 40 or the 0 blank reference level
set. 

		For those filters in the lowest quartile of reported results, all of
which should have come from the 20 µg reference level set, there was a
28 percent chance that the filter came from the 40 or the 0 blank
reference level.  In effect, the results of the analyses by these five
commercial laboratories were not sufficiently accurate to reliably
distinguish between concentrations that differed by a factor of two. 
For example, an 80 µg versus a 40 µg, which would be the current PEL
versus the proposed PEL.

		Moreover, the inter-laboratory variability in this study was high. 
The results show that a filter with a given loading of silica could
yield a reported silica mass that varied by nearly a factor, based
solely on which lab performed the analysis.  If these results are
representative of laboratory proficiency, they indicate that
inter-laboratory variability alone would make it impossible to reliably
distinguish concentration as it differed by a factor of two.

		Intra-laboratory variability was also high.  With the exception of one
of the labs, the coefficients of variation ranged from 20 to 66 percent
at the various reference levels, indicating substantial problems with
intra-laboratory precision.  Within some of the laboratories, there was
substantial overlap among the 95 percent confidence interval around the
mean-reported silica masses for the different reference levels.  The
variability of reported results within individual labs was such that, in
some cases, the reference levels could not be reliably distinguished
from each other.

		In conclusion, based on the results of this study, obtaining reliable
silica measurements to determine regulatory compliance, even at the
current PEL, is problematic.  The results reported here would have been
even more problematic if Dr. Cox had not excluded from the analysis
those filters, more than one-third of the total, that had silica
loadings of 20 µg or more but were reported by the labs to being below
detection level.

		Based on this study, achieving the conventional 95 percent statistical
confidence of an analytical method was not reached.  Inter-laboratory
variability was so high that reported results could not be used reliably
to discriminate among filters to reflect eight-hour exposures to silica
at levels of 25, 50, and 100 µg/m3 of air.

		Within a single laboratory, there was enough variability and reported
results so that two-fold variations in exposure could not be
distinguished.  Based on the results of this small study, the panel
highly recommends that OSHA conduct a larger blinded study for all labs
intending to analyze crystalline silica before adopting exposure limits
that very likely cannot be measured reliably using current technology
and methods.

		Thank you for the opportunity to participate in this important
hearing.

		JUDGE SOLOMON:  Mr. Waggener, yeah.

		MR. WAGGENER:  Good afternoon, Your Honor, OSHA, and the audience.  My
name is Jack Waggener.  I'm a senior principal and professional
engineering manager with URS Corporation.  I'm here today to talk about
our critique of the OSHA cost model for general industry, the general
industry sectors, and our modifications that we made to that model to
make it real world in its estimate as we see it.

		I'll begin with an introduction.  The American Chemistry Council's
Crystalline Silica Panel asked and engaged URS to perform this process
for the general industry.  URS is approximately a 60,000-employee
professional engineering and scientific company that is involved in all
types of industries throughout the country, and essentially for all of
the industries in this particular category.

		Additionally with that experience, which includes mine, we have also,
for more than 25 years, owned one of the PAT-certified laboratories that
just got discussed in the discussions as to the accuracy of the test.  I
personally have more than 40 years' experience dealing in this type of
issue, beginning with the lead PEL in the 1970s, which I was highly
involved in on behalf of the industry and helping the industry getting
compliance, and actually during the generation of that PEL.  More
recently, in the 2000s, I was heavily involved in a similar fashion, as
I am today, with the hexavalent chromium rule and also in the cadmium
rule.  

		URS prepared comments which are part of the docket, and that is what I
am here to talk about today primarily.  Along with that, there is an
incredible amount of Excel sheets and backup data that goes along with
all of this costing procedure, which is a very large effort on our part,
and I'm sure OSHA's part, to put it together.

		In consultation with our team at URS, we also were involved with more
than 100 different industry CIHs and engineers throughout these
particular industries that we are going to be talking about today.  To
begin the process, when the rule was proposed, we reviewed the different
documents, the PEA and so forth.  In the record, we did not have the
OSHA spreadsheets for the cost model, which we were trying to review. 
We asked for that to be provided, and we appreciate the fact that you
guys promptly provided that to us.  That is what we reviewed and
critiqued.  In doing that, we found a number flaws that went along with
that, in which we made corrections.  

		We'll call this the executive summary.  We will get the bottom line
first, and then talk about how we got to the bottom line.  We looked at
19 general industry sectors.  There are actually 29.  We chose the 19
because we felt like we had at least more credible data in the record to
ascertain what the actual cost would be.  I'll talk about that a little
bit more as we go through here.

		The URS annualized cost we estimated to be, for these 19 industry
sectors, over $6 billion, some 56 times more than what OSHA estimated at
approximately $100 million.  We feel like, based on the information we
did and the analysis we did, and how we chose to make changes, that we
are still conservatively low.  Even though this number far exceeds
OSHA's number, it probably is much higher than this.

		Many of the professionals in these industry categories felt like to
actually reach, and this is an important issue, the 50 versus getting to
the original 100 PEL is probably at least five times more difficult and
costly.  We found a number of flaws that created this when we tried to
make those corrections by introducing different assumptions and
different things of that nature to get to what we believe to be more
real-world corrections.

		The industries that we focused on, the 19 general industry sectors,
these are the ones that we felt like we had reasonable data to do so:
the asphalt paving products, asphalt roofing, concrete products, custom
jewelry, cut stone, fine jewelry, flat glass, iron foundries, mineral
processing, mineral wool, nonferrous sand casting foundries, non-sand
casting foundries, other ferrous sand casting foundries, other glass
products, paint, pottery, ready-mix concrete, refractories, and
structural clay.  Those would be the focus of what we'll be talking
about today.

		In the next slide, I presented the ones that we felt like we didn't
have good information to make the estimates on.  These were captive
foundries, dental laboratories, dental equipment and suppliers,
railroads, refractory repair.  There were five different porcelain
amalgam categories that were identified.

		An example of that issue, while we didn't think we had the good data
to do it, the data was not representative that was in the record, is
what I should say.  For the porcelain amalgam sectors, OSHA estimated
some more than 9,000 facilities as being out in the United States.  The
Porcelain Amalgam Institute, which represents this industry, estimates
there are 125 facilities.  There was obviously a disconnect there.  I
know the Porcelain Amalgam Institute is submitting comments to make that
correction.

		In our next slide, we break down, again, this is the URS estimate for
the cost.  It's broken down by industry category, as I just mentioned
earlier.  If you go across the top of the page, you have engineering
control costs to meet the 50 PEL.  These are emphasized.  This is the
URS estimate after we changed the model, or the assumptions and the
inputs of the model.  

		The 100 PEL, and then we also had the incremental cost.  We have
ancillary provision costs.  Then we had the incremental costs and then
the full cost.  Well, of course, as you have heard previously, we
believe the full cost is the real cost that the rule will be impacting
on industry and should be part of what is really looked at here as we go
forward, not just the incremental cost.  Again, at the bottom right-hand
side of the table, you can see the total cost of them, $6.1 billion of
annualized cost that we are predicting.

		The next table is what is by the OSHA model, which is the incremental
cost, which without going into a lot of detail, it still comes up to the
$108 million versus what we had on the previous page.  Just as an
example for concrete products, the URS model cost is about 42 times
higher than what OSHA has gotten for iron foundries.  We are about 67
percent.  For structural clay, bricks, and things of that nature, we are
about 24 times higher.  We looked at each industry separately to get
what it really would approximately be relative to what was really going
on in that industry.  It goes up and down in terms of the multiples of
how much more that it would actually be.  So just briefly, for the 19
sectors, the URS estimate is about 44 times higher than the incremental
cost, as based on what OSHA got, and the URS full cost is about 56 times
higher.  

		The next slide deals with the sizes of the facilities, which was taken
into account by OSHA and URS in the cost model.  I am preaching to the
choir here, I guess, but at least for the very small entities, that
represents less than 20 employees.  For the small entities, as we
analyzed, it is more than 20 or equal to 20, and up to less than 500
employees.  That is the SBA definition.  Then, we also had large
entities, which were larger than 500 employees.

		So we looked at each one of those.  That's important in looking at
what the potential cost impact is and what the economic impact to the
industry is, which Stu Sessions after me will be talking about.  The
OSHA cost model lumped, however, the very small and small entities
together.  So all the companies that were under 500 employees were all
lumped together in their analysis.

		We believe that the size, in looking more specifically also at the
less than 20, was very important.  So we did an analysis to keep those
separate and present that data separately.

		And the next slide deals with the engineering cost per facility size. 
Again, this is lumping all these 19 industry sectors together,
aggregated together.  We show, in terms of millions of dollars, for that
grouping like to the left here, the $518 million for very small
facilities under URS' estimates.  OSHA had comparatively $15 million. 
Then, for small entities, you can see the numbers.  It was considerably
higher, and then for OSHA, without going into all the details there. 
Anyway, that shows that.

		Probably the most important one here, from my view of looking at what
we did here, or at least one of the most important things, this slide
shows what the average cost is for each facility, one individual
facility.  Not aggregated together, but what the average is based on the
models.

		And I've chosen two sectors, just as examples.  We have it for all of
them obviously.  Again, these are numbers.  We've been talking about
millions of dollars and billions of dollars.  These numbers are in
thousands of dollars.

		This is by facility.  I think the importance here is shown here.  If
you look at the iron foundries, which is the top half of the table here,
the bottom half is structural clay, you can see these are facilities
less than 20 employees.

		Actually, the average employee is like less than 10.  It is eight
employees or so forth, so you are looking at a very small business that
may be looking at an annualized cost under the URS estimate of $218,000
a year, versus under the OSHA's estimate to get into compliance with the
50 PEL, less than $3000.

		Obviously, if you think about it, I'm sure you know many small
businesses spending less than $3000, you know you can buy a pretty good
computer that goes on your desk.  That is not an unusual purchase.  But
when you start talking about having to spend $218,000 a year as a very
small mom and pop company, that is a big deal, just for reference. 
Likely, you're going to go out of business.  That is a very likelihood
that that would be the case.

		Even for the small facility, it's a big margin.  Again, these are
average numbers; some are higher, some are lower.  These are iron
foundries.  The average one point, or I should say, $1,874,000 versus
about $40,000, what OSHA indicated.  Again, these are people that
average actually less than 100 employees per facility.  That is a heck
of a bite to be taking out of a small company as you go forward here. 
Anyway, we recalculated these and came up with what we believe the real
numbers are, closer to what reality was than what the OSHA numbers were.


		From this point, I want to talk about that there were a number of
flaws, that at least we believe were flaws, in the OSHA model that we
have identified, and we tried to make connections to that are more
realistic.  I am going to talk about those flaws, and kind of what we
identified as flaws, some of the major flaws and corrections that we
made.

		We will start with engineering controls, and then we will go to the
ancillary provisions.  Pardon me.  Why such a large difference in the
cost?  Well, one item to begin with was how OSHA estimated the number of
controls that would be needed to be put in their cost model.

		They did not actually look at the types of facilities out there.  They
didn't look at the size of the facilities really or look at the number
of facilities.  They looked totally at whatever the over-exposed
employees were on a national basis.  Basically, through analysis, they
decided that this rounds it off, just about for all of the -- they
divided.  They said, you have four exposed workers per control.  So they
took that national average and divided it by four.  

		Now, that assumes you've got four people together all the time in
these locations and so forth, which is just clearly not realistic.  We
tried to take it and move it into a different field, to where we would
look at each individual facility, the sizes, the characteristics of that
particular industry as identified by OSHA, and use that data to come up
with a different amount of a number of employees.  Not a different
number of employees that is overexposed, but a different number of
controls that would be needing to cover those facilities.

		And we took a statistical approach, a very typical standard approach
using a binomial distribution, which is commonly used.  We looked at
every industry and the sizes and every job differently because every job
has different exposures.  We generated more than 400 binomial
distributions to come up with what we believe would be the number of
controls that would be needed, spread out over the plants individually
and throughout the United States.

		I'm not going to go into statistical education here, but this is just
one of the typical curves that you would get.  This center of the curve
there at the top represents that line or the percentage of workers that
OSHA predicted to be overexposed.  The far right on the horizontal axis
demonstrates the overexposed or the total workers in that job on the
average for that size.  It's just an example.

		And you go through this, and you determine how many controls and how
many people are being exposed at certain levels and different size
facilities.  That was the way we did that.  Excruciating details with
regard to that in the record that we submitted.

		So the next slide deals with other problems that we found.  Again,
they failed to account in the model for the facility sizes, the number
of shifts that different companies operate.  They failed to recognize
the physical limitations of some of the controls. 

		You've heard lots about variability of monitoring and testing.  None
of that was taken into account either when coming up with these
estimates.  As a practical matter, we cannot address only the
overexposed workers in a given job under certain situations for
controls.

		OSHA's model, we believe, frequently undersized the size of the
controls that have been put forth by OSHA.  Also, the incremental costs,
which we will talk about in more detail here, which is actually a huge
issue, the cost of LEV, local exhaust ventilation, which is just one of
many controls that are used in these types of plants.  You have also
heard mention the ISO/CEN.  The impact of the ISO/CEN was not taken into
account in the model.

		Moving forward, the first error we saw was in, again, looking at very
small and small facilities in terms of OSHA saying there are four
overexposed employees per control.  The matter of fact is that many of
these very small and small facilities have one employee where that
control would be.  And that needed to be accounted for.  And this is
very important obviously to coming up with the cost of the very small
and small facilities.

		Frankly, sometimes it's the reverse.  There are facilities, such as in
the iron foundries, that may have eight controls per employee, not the
reverse, four employees per control.  By choosing that number or
sometimes different numbers, it depends on what they were looking at,
they really underestimated what the cost of the rule would be.  So the
bottom line, we had to make corrections with regard to that.  And the
binomial distribution helped us make that -- one of the things that
helped us make that correction. 		

		In terms of work shifts, OSHA assumed that there would be two shifts
per facility.  The fact of the matter, most very small facilities, and
in some cases, a lot of the small facilities, they only work one shift. 
And so, we had to make that adjustment in the model to account for the
cost.

		The variability in the monitoring has been spoken to before.  The
additional -- OSHA has really identified that you only need to do this
first monitoring.  You are going to find out whether or not you have
individuals that are -- that you have to have controls on or not.

		The fact of the matter is, and this is common knowledge, you know, you
do an analysis.  You go in and do sampling one time.  And the next --
and you find out that Person A is out of compliance, and Person B who is
20 feet away from him is in compliance.  You go back three months later,
and the reverse happens.  So you've got all these different people, and
you can't just say, okay, I did one sampling, and I'm okay.  I know what
the statistics are.  That's not the case.  That's not the real world. 
So there's more people that need controls than just doing this one
event.  And, therefore, this four overexposed worker issue is flawed.

		And, again, kind of going along with the same thing, just from a
practical matter, you can't just address what you identify once or twice
to be overexposed workers.  If you have a very high percentage of
overexposed workers in a job, and that's a consistent thing --
consistent in a certain job area, you can't just pick and choose. 
You're pretty well going to have to put in consistent controls
throughout that area, not just one control.  It may involve two or three
controls.  

		And what URS did in our model, if you had more -- if the data, which
OSHA put together in a PEA, had more than 50 percent of the employees in
a given job that were overexposed, we assume that all of those workers
needed to have a control, not just half of them.  And that's more
realistic of what would go on in the real world.

		The physical limitations of controls, this is just an example that we
have given here.  The OSHA model, as I have already said, is a general
rule to assume these four workers for control.  Well, a good example of
where that does not hold is in a lot of the mobile equipment, front-end
loaders, tow motors, and so forth that operate in these facilities.

		If they are put into an enclosed cab with an air-supplied system to
protect them from the silica, there is typically one person that
operates that on a shift.  And at most, if they're running two shifts,
there's two people.  OSHA assumed there was -- that was going to cover
four people, when in fact, it typically is only going to cover one, two
at the most.  So those are -- for those types of examples, we adjusted
the model to reflect that type of thing.

		As I -- another situation I kind of alluded to earlier, iron foundry,
the muller operation project, part of the job or the mixing areas may
have one worker that is running around, taking care of five, six, seven,
eight mullers in a given shift.  Every one of those has to have a
control and a proper control.  So there you've got eight controls per
person.  So, again, just another example of why we had to make
adjustments.

		All right.  URS incorrectly used -- excuse me.  What did I just say? 
URS?  Thank you.  It is getting late in the afternoon, okay.  OSHA
incorrectly used exactly the same controls to achieve the 100 PEL as
they did the 50 PEL.  And this is just not technically correct.  

		In other words, it takes much more sophistication to get from -- to
50 or even to 25 action level than you are just to get to 100.  And this
is what's in the model, so there's no additional or more effective
controls to get to 50 than it is to get to 100, so this artificially
reduces the cost.

		It's clear that the four workers that are estimated to be exposed
above 50, but less than 100, it is reasonable to assume that most cases
that controls for the 100 PEL are already in place to meet the PEL. 
Clearly, additional more effective controls are going to be required.  

		So URS and their model increased the sizes and the efficiency of some
of the controls, which increases the cost.  It made them larger,
stronger, such things as the LEV, local exhaust ventilation, increased
the cfm to overcome and collect more of the dust or the tweaking that
may be done to those things, so that was taken into consideration.

		With regard to the model controls that were used by OSHA, URS felt
like they were considerably lower than they needed to be for many of
these industries.  We looked at this on an industry-by-industry basis in
sizes.  The capture velocity, such as for LEV, typically 100 all the way
up to maybe as much as 250 by OSHA.  For many of these industries, that
is simply not doing the job.  It would have to be sometimes two, three,
four, five times higher than that in many cases. 

		OSHA -- I mean, URS did increase those to take into account that when
OSHA did their assessments, well, I should say, in the PEA, the
feasibility determinations for meeting the 50 PEL are anecdotal at best,
often based on a very few selected samples slightly below the 50 PEL,
and often after much trial and error that has taken place at that plant,
which is important to the cost aspect.  

		I will just say the trial and error process, after you make these
corrections in plants and go through the process, you add a new local
exhaust ventilation.  You make a change in the table that the person's
working on.  You go back and sample again.  No matter what a great job
you do, you're going to find people that are still not in compliance. 
And you have to go back and further adjust, and maybe even replace those
things.  And those are not -- those are costs that are not even built
into the URS model or the OSHA model.  So those alone could raise the
price by 25 to 50 percent.

		When OSHA looked at the -- it appears that when OSHA looked at the
data that's represented in the PEA, where they did have numbers less
than the 50, they never did a statistical analysis of it like's been
talked about, the 95 percent SAE method that's commonly used by OSHA
field personnel in doing inspections.  The result is that it's an
over-optimistic view that OSHA has taken to identify what can be
accomplished by these controls.

		OSHA's cost model did not always use all the controls recommended in
the PEA.  In the cost model, there would be things recommended that
would be necessary to meet it.  They simply think in some cases, those
are just left out of the model.  They would have been needed to reduce
the contamination or the exposure rate that was projected.

		How are we doing on time?  Okay.  All right.

		JUDGE SOLOMON:  Well, if Mr. Sessions is going to get a half an hour,
you've got about 10 minutes.

		MR. WAGGENER:  Okay.  All right.  I've got more here than I can
possibly say in that, so we'll scoot through.

		Because OSHA incorrectly calculated the incremental cost of the
proposed, OSHA subtracted the entire cost for all workers over the 100
PEL, and there were a number of people over the 100 PEL, to obtain the
incremental cost, even though the controls used by OSHA were intended to
achieve the 50 PEL.  In effect, all of these workers that are over 100,
which is more than half when you get down to the total amount that OSHA
cost, have been excluded from the cost model completely.  There is no
cost in there for those guys to reach things.  This is, to us, a huge
issue.  Stu will talk a little bit more about that.

		Anyway, we made corrections with regard to that.  The cost of the LEV
or the cfm, OSHA used a number, and we went through a process to come up
with a number that is almost double what OSHA came up with in terms of
the basis for the annualized cost for LEV, for instance.  And it's
backed up by a number of things that are in the record.

		The next thing is dealing with the ISO/CEN, which roughly we'd assume,
based on the analysis, that we would pick up an additional 20 percent of
silica, which all the PEA data is really based on the old method, not
the new method with the ISO/CEN, so we increased the number of exposed
employees.

		All right.  Moving right along, ancillary costs.  Getting out
engineering, ancillary costs.  We looked at these different items.  We
made changes to the initial exposure monitoring, periodic monitoring,
respirator use, medical surveillance training, regulated areas,
professional cleaning.

		Okay, the initial exposure monitoring, we believe OSHA underestimated
the time and cost for the CIH or CIH assistant or helper to actually
perform the task of taking samples, evaluate the results and create a
report.  URS added one day's time to that cost estimate.

		URS, excuse me, OSHA's low estimate for workers requiring this simply
used one-fourth of the workers in at-risk job.  URS has increased this
for very small facilities.  We said, you need to test all of the
employees, where we have only eight -- maybe eight different employees
on average.  And those employees are doing a lot of different things. 
They're not doing one job.  They're doing two or three or four jobs,
which involved jobs that have exposure.

		For the small facilities, you would use one-half of the workers in
at-risk jobs, I should say, over a two-shift basis.  And then we adopted
OSHA's one per four method that they actually used.  We amortized
because of what actually occurs in facilities.  We amortized this over 5
years instead of 10 years because there is changes in production and
controls that go on, and it's more likely to happen over a 5-year period
than over a 10-year period.  It means you have to go back and do
monitoring.

		For the periodic monitoring, OSHA, who we believe is unrealistically
low, assumed that 15 percent of the workers would be over the action
level and that no worker would be over the PEL.  Well, based on the
discussions you've heard earlier and today, that is just simply
unrealistic.  We expect many people to be over the PEL and many more
people to be over the action level.  The proposal has that if you are
over the PEL, you will have to do quarterly sampling for that job, and
semi-annual monitoring if you're over the 25.  And there's, as I've
already said, lots of trial and error that goes into that.

		So we've increased the cost model significantly.  We assume that, for
people over the 50 PEL, at least half of the workers that are currently
over the 100 PEL that have not been able to get under 100 have not
accomplished this.  Let's just remember, this has been around a long
time.

		These companies have been working very hard to be under the 100 PEL. 
They don't want to be over the 100 PEL.  And they have not been able to
get there, so it's not realistic to assume we're going to go to a level
that we can hardly measure accurately and be able to assure ourselves
that we're going to be under the 50.

		And respirator costs and use, it was somewhat the same thing, without
going into all the details.  Again, these companies, I've got an example
here, 44 percent of the iron foundry cleaning and finishing operators
are still over the 100 PEL.  And these facilities have been trying to do
it for years.  It's just unrealistic to assume you're going to be down
to a very low level and be only 10 percent over the PEL as OSHA assumed.

		Medical surveillance, again, you're going to have a lot more people we
believe that are going to have to go onto medical surveillance because
it will be at a higher level than the 25, and particularly the 50, PEL
that OSHA has estimated.  For ongoing medical, we've done somewhat the
same thing.  We've assumed that half of the workers initially over the
100 PEL would still exceed the 50 and need to have the ongoing medical. 
And training had some labor issues with regard to the labor costs that
were used by OSHA.

		Regulated areas, this is where you have areas where people are over
the 50 PEL, and you have to have regulated areas.  And we'd made
different estimates than OSHA did based on the fact that we believe that
it will be much higher than what OSHA has estimated, such as having 1
visitor in very small facilities, 5 and 20 in large facilities.

		One major issue and very costly, in the PEA, OSHA estimated -- or
indicated that, to meet the 50 PEL, you would have to have professional
cleaning annually or more frequently for both the concrete products,
foundries, mineral processing, porcelain enameling, pottery, and
structural clay.  And then, in the OSHA model, I think this is just an
oversight by OSHA, the OSHA model, the engineering cost model, indicates
that they are going to do it in an ancillary cost.  But when you go to
the ancillary cost model, they didn't do it.

		So URS added that to the cost of the rule, based on industry's
estimate.  That's about a dollar per square foot of facility.  So you
have all these things adding up to be the last slide, which is like the
one I began with almost, which basically says that we are at about $6
billion for the 19 industries, some 56 times higher.  We found many
flaws, and we've tried to make real-world corrections to those to come
up with a model.  

		With that, I'll turn it over to Stu, who is sitting right by me.  You
wanted to change places? 

		MR. SESSIONS:  We're party to an unusual event here, which is our
attorneys are handling the tech issues that are associated with the
presentation.  

		JUDGE SOLOMON:  There's a false assumption that attorneys aren't
versatile enough --

		MR. SESSIONS:  Yes, I'm quite impressed.

		MR. WAGGENER:  If we had our grandchildren here, they would've figured
it out a long time ago.

		MR. SESSIONS:  But we're not equating attorneys with our
grandchildren.  

		JUDGE SOLOMON:  Is this any different than your report that you gave
the other day?

		MR. SESSIONS:  Yes, it is.  It is different, same red-colored
PowerPoint.  

		I'm Stuart Sessions.  I still am the President of Environomics,
Incorporated, which is a small consulting firm in Bethesda, Maryland. 
We work on economic analysis of regulatory and policy issues involving
the environment, occupational safety and health, and energy.

		Perhaps most important -- and I'm testifying on behalf of the American
Chemistry Council, which has supported my work.  Perhaps most important,
in terms of my background for this presentation, I also, for a number of
years, had management positions in the Office of Policy Analysis at the
Environment Protection Agency.  And I ran a branch there which
essentially evaluated the impact of EPA regulations on individual
plants, on industries, and on the national economy.  And then I managed
a division which, in a sense, oversaw the agency's procedures for
analysis of regulatory impacts of all sorts.

		That experience leads me to believe and to offer suggestions at a
number of points that, when OSHA has run into sort of difficulties or
inaccuracies in handling particularly regulatory analysis issues, there
are a variety of other ways that other agencies have undertaken to
pursue handling these difficulties.  And in many cases, I think our
experience working for and with other federal agencies informs our
comments and suggestions to OSHA, both about methodology and data
sources.

		In the interest of time, I'm going to skip around pretty substantially
and just sort hit some highlights of the testimony.  A first issue that
I want to talk about is a topic that had been mentioned actually several
days ago, and that both Neil in his introduction and Jack have
referenced, which is this question of what we call full versus
incremental cost.

		And I want to kind of portray exactly what we believe OSHA has done in
the current analysis in terms of estimating the costs for general
industry to comply with the proposed rule.  In looking at this, we
divide the world into two groups.  There's the group of workers who OSHA
currently assesses as being exposed between the proposed PEL and the
existing PEL, workers who are exposed at between 50 and 100 µg/m3.

		There are, according to OSHA's exposure analysis, 41,500 of them --
sorry, let me back up with another interjection.  All of these numbers
and all of my presentation -- almost all of my presentation, all of
Jack's presentation, have addressed a general industry, with the
exception of hydraulic fracturing.  At one point, I will mention
something about hydraulic fracturing.  But all that we have been saying
is general industry less hydraulic fracturing, so these numbers in
particular are less hydraulic fracturing.  They're not counted in that. 


		In any case, so Group A are those exposed between the existing PEL and
the proposed PEL, and there are 41,000 of them.  Group B is all the
workers who OSHA estimates to be currently exposed above the current
PEL, and there are 81,000 of them, for a total of 122,500 workers who
OSHA estimates will need additional protection if the PEL is reduced to
50 µg/m3.

		So OSHA's cost analysis essentially estimates the cost for all 122,500
workers' exposures to be reduced below 50, and then subtracts from that
total the cost for reducing the 81,000 workers who were exposed above
the current PEL to below 50.  And OSHA's, what we call, "incremental"
cost estimate or, in some cases, they call it their "additional cost
estimate," is in fact only the cost to address the exposures of the
41,000 workers who are exposed between the proposed PEL and the current
PEL.  

		And in the cost estimates that are reported by OSHA, in the preamble
or in the PEA, there is zero cost counted for the 81,000 workers who
were exposed above the current PEL.  So we contend, among other things,
that cost ought to be estimated in some manner for all 122,500 workers
OSHA reports cost.  OSHA estimates the need for controls for only the
41,000 exposed between the two PELs.

		Now, OSHA portrays their analysis as attempting to capture the
incremental costs attributable to moving the PEL down from its current
level to 50.  But I submit that some of those 81,000 workers -- let me
put it a different way, all of those 81,000 workers, their employers
would incur substantial costs due to the reduction in PEL because the
Group B workers, in order for their employers to meet the current PEL,
the employers would make efforts to reduce their exposures to just below
100.  Basically, you are in compliance with the current PEL if you get
those overexposed workers down below the current PEL.

		And so, the impact of -- the incremental impact of the reduction in
PEL will require, in fact, reductions among -- reductions in exposure
among those 81,000 workers from somewhere just below 100 to somewhere
just below 50.  So OSHA, in fact, has taken all of the costs for those
81,000 workers off the table, and arguably, the cost if OSHA were
correctly to do the incremental cost analysis that the Agency says that
it is intending to do, those 81,000 workers would have a substantial
cost.

		And, in fact, the cost per worker for those 81,000, if their exposures
were just below the current PEL, would be about the same as the cost per
worker for the 41,000 that OSHA does count.  So even in OSHA's
incremental analysis, OSHA has left out very roughly about two-thirds of
the cost that the Agency should calculate.  So that's kind of an
overarching point about the cost analysis. 

		And it's actually two points.  One point is that OSHA has done the
incremental cost analysis incorrectly.  And the second point is that
when you are talking about economic impacts, which is a question of how
affordable are these costs or what impact will these costs have on the
industries that need to bear these costs, the correct notion of cost to
address, if you are worrying about economic impact, is the full cost.

		It's not just the incremental cost of reducing the PEL that these
industries will need to bear.  It's the cost of coming into compliance
with the current PEL also.  And it's the combination of those two costs,
it's the sum of those costs, which determine the economic impact of
attempting to comply with the proposed new PEL.

		Another point that I want to hit very briefly, and this is the one
time at which I will talk about all of general industry, including
hydraulic fracturing, this is just a summary of what the various
industry groups that have been evaluating this regulation have added up
as the total cost of the rule.  And so, general industry, with excepting
hydraulic fracturing and focusing in particular on the 19 of the 31
sectors that Jack mentioned, have a total cost estimate from our point
of view, from industry's point of view, of about 6.1 billion per year.  

		Hydraulic fracturing, according to the estimate that the American
Petroleum Institute has prepared, estimates a cost of attempting to meet
the proposed PEL.  And API's cost estimate is very important to realize
that they don't believe that the technologies that OSHA puts on the
table for meeting the proposed PEL will, in fact, do it, and they don't
believe that those technologies are, in fact, available.  But assuming
that they were able to do it and that they were available, API estimates
those costs at about $365 million to $366 million a year.  And the
construction industry estimates the cost of the construction regulation
at $3.2 billion a year.  A total across all the industries for the
entire regulation is about $9.7 billion per year, in contrast to OSHA's
estimate, which is, if I can read the fine print up there, $686 million
per year.

		So it's not just the general industries that Jack talked about for
which industry is disagreeing sharply with OSHA's cost estimates.  It's
all of the affected industries.

		The key issue in estimating economic impacts or in assessing economic
impacts, or assessing economic feasibility of the regulation, is in a
sense comparing the compliance cost that any industry will bear against
that industry's ability to bear the cost and against, in particular,
that industry's revenues and profits.  And OSHA has taken this kind of
sensible sort of approach towards economic impact assessment for all of
the OSHA regulations that I have ever worked on.

		I don't know how long this practice has been in effect, but it's a
very sensible general approach.  And OSHA's general rule of thumb is to
get worried about affordability if the cost exceeds either 1 percent of
revenues or 10 percent of the industry's profits.  So costs below that,
there's a good case that the proposed regulation is economically
feasible.  Costs that exceed either of those two thresholds, there is
some question about economic feasibility and further analysis is needed.

		Well, to make this comparison, you need at least three pieces of data.
 You need an estimate of the cost of the regulation, you need an
estimate of the industry's revenues, and you need an estimate of the
industry's profits.  And so, Jack has talked extensively about
re-estimating the cost to the industry.

		I'll talk a bit about the process of estimating revenues and profits
for the affected industries.  And there's some amount of art and science
to estimating revenues and profits.  And there are differences of
opinion that can exist on those topics, and I'm going to talk a little
bit about some of those.

		In particular, on estimating profitability, basically OSHA took data
from a particular source on profitability for these industries, from the
years 2000 through 2006, averaged over those seven years, and said
that's the average profitability for each of these industries.  And
that's what we will run the comparison of compliance costs against.

		An easy objection or an easy criticism of what OSHA's done is
profitability for these industries now, particularly after the
recession, and profitability for these industries now, particularly
given that most of these general industries produce construction
products and the construction industry has had sort of a housing crisis
in addition to the recession to deal with, profitability now is much
different than it was for these industries between 2000 and 2006. 
Certainly OSHA recognizes that and OSHA will, as OSHA staff have said,
and as they always do with the regulations, will update the data to
reflect the most current data available.  In particular, the data source
that OSHA uses is something called the Corporation Source Book.  And
that is available now through 2010, so you can update the data through
2010.

		And, in fact, in some of the numbers that I will report later, we've
updated OSHA's data through 2010.  And I call this new profit series
something called "Revised Profits."  And so, I will show some
comparisons of costs against profits as OSHA estimated them and costs
against revised profits as we've estimated them.

		But another very significant issue with regard to how OSHA chose to
measure profits is what data on profitability OSHA took from this
reference, this IRS Corporation Source Book.  And, basically, the IRS
reference includes a sample of tax returns from corporations in every
industry in the country.  And OSHA chose various data from that which is
available in the Corporation Source Book to represent profitability.

		In particular, my Point 2, I want to stress this, is essentially,
reported in this IRS documentation, OSHA data chose data for each year
for only those corporations that were profitable.  And so, OSHA, in a
sense, estimated the profitability, the profit rate, each year for the
corporations in that year were profitable.  

		So in my equation under Number 2 here, essentially OSHA took total
profits for those corporate tax returns that showed positive profits,
and divided that by, quite interestingly, total revenues for all
corporations, whether they showed positive profits or not.  So that's
really kind of an unusual calculation where the numerator is profits for
one group of companies, and the denominator is revenues for a different
group of companies.  In some sense, that doesn't make sense.

		Basically, when you do the numerator and the denominator, you want
both apples in the numerator and apples in the denominator.  But,
anyway, the more important point is that the way we believe OSHA should
use this Corporation Source Book IRS data is, let's take the profits
from all corporate tax returns, whether the companies show a profit or a
loss, and let's compare that against the revenues for all the companies
in the industry, whether they show a profit or a loss.  And let's call
that the profitability of the industry.  

		In particular, what's wrong with what OSHA does?  Well, it's wrong to
estimate profitability, to calculate profitability, looking only at
profitable companies.  Every year, there are some companies that are not
profitable.  And from year to year, the companies that are not
profitable for a variety of reasons tend to differ.

		So, you know, the regulation applies to all employers in the covered
industry.  It does not apply to only the profitable employers in the
industry.  And to assess the affordability of the costs, you ought to
look at the entire scope of the industry, profitable and unprofitable
companies, in judging affordability.  

		And, in particular, one would think that if you're concerned about the
possibility that the regulation isn't affordable, that you would be
particularly concerned to include in the calculation the companies that
are unprofitable.  Those are the ones that are most vulnerable.  Those
are the ones you would most want to have in the calculation.

		So, basically, there are issues in terms of what OSHA has done with
regard to profitability involving the recency of the data, which can be
corrected to some degree.  You can go up to 2010, if you take the same
data source, but a big issue with regard to how OSHA is selecting data
from that source and how it's using it.

		I guess I'm not going to go through this, but the entire testimony
package will be in the record.  And I refer OSHA, in particular, to
this, which is a sample of this calculation for one particular industry
for one particular year.  And I've split the IRS data into the returns
for profitable corporations and the returns for unprofitable
corporations.  And this slide shows what OSHA did and what I would argue
they should have done.

		Another point, which is a different point, about estimating
profitability is this particular data source, the Corporation Source
Book, provides information at what's known as the three- or the
four-digit SIC level -- NAICS, excuse me.  So it provides information
for rather aggregated industries.  In fact, OSHA's cost analysis, and
much of OSHA's analysis, is more granular, and it estimates costs at the
six-digit industry level.

		When it comes, though, to comparing the cost for a particular
six-digit industry, I'll take a particular six-digit industry, ceramic
wall and floor tiles is a six-digit industry, which is within a
four-digit industry, which is clay and refractory and other non-metallic
minerals.  OSHA estimates costs for six-digit industries, but the
profitability data available via the particular chosen IRS source is
available only at the three- or the four-digit industry level.

		The net result is that OSHA represents the profits for the component
industries within a very large industry by the profitability for the
entire industry as a whole.  And you get results like for asphalt
paving, which is a very specific six-digit industry, OSHA represents the
profitability for that industry by the profitability of a larger
industry that includes petroleum refining.  And petroleum refining is,
you know, some 50 times larger than asphalt paving mixtures.  And there
is no particular reason why the profitably of asphalt paving mixture
should bear much relationship to the profitability for petroleum
refining.

		And this table just shows -- essentially its fine print, it shows
what a small percentage of the aggregate industries about which OSHA
draws profitability information, what a small percentage each of the
individual industries to which OSHA transfers that information, that
these small industries are a very small percentage of the large
industries.

		Now, this is a tough issue because profitability data is -- there are
reasons why OSHA chose the IRS source.  There are pros and cons to it. 
And my bottom line suggestion on this is there are other sources of data
that go to the six-digit NAICS industry.  And OSHA ought to, in my view,
to get things more correct, do the best job it can in comparing and
combining the profitability information from the different sources.

		This issue has been discussed before, and in fact, OSHA back for the
SBREFA proceedings in 2003 used a different source for profitability
information that, in fact, does have much more granularity at the
six-digit industry level.  And OSHA backed away from that, and I think
the answer ought to be some sort of combination of attempting to make
the best judgment.

		Okay, Neil says I've got 10 minutes.  I will say one -- two rather
quick points about how OSHA estimated revenues.  The difficulty in
estimating revenues is good data on revenues by industry is available
only for every five years.  It was available for 2007, which was before
OSHA did this analysis.

		So OSHA's revenue data for 2006 was sort of generated by taking 2002
revenue data and updating it in a way that doesn't work to 2006. 
Basically my answer on revenues is strong recommendation that OSHA
should wait for the 2012 data to become available.  It will become
available later this year.  And that will be far better than the
procedure that OSHA has.  Bob is shaking his head.  Maybe it won't be. 
Maybe you have some insight that it won't be available later this year.

		In any case, the way OSHA has done it, taking the ratio of payroll,
which is an annually available statistic to revenue, does not work.  And
we provide a couple of charts to that effect that the ratio of payroll
to revenue is not constant over time.  It is not constant across size
categories within an industry.  And, furthermore, the ratio of payroll
to the number of establishments has declined very sharply since 2006. 
And that's a reason why 2006 data certainly is not representative.

		Basically, what I want to spend my remaining time on is essentially
what do we get when we look at these comparisons of cost against profits
and revenues.  And the first table is basically comparing OSHA's cost
estimate, which is incremental costs as opposed to full costs, which
I've talked about, against OSHA's data and less than desirable estimates
of revenues and costs.

		And in the lower right corner, I apologize again, people can't see the
fine print, but basically for no industries doing this comparison does
OSHA find cost exceeding 1 percent of revenues or 10 percent of profits.
 The next comparison is, if you look at OSHA's full cost, i.e., the cost
to reduce exposures for all 122,000 workers, not just 41,000 workers. 
In fact, looking at full cost on the far right-hand column, 3 of the 19
industries exceed 10 percent of profits.  So even using OSHA's own costs
and OSHA's own estimates of profits, if you switch to full cost, you
begin to see potential affordability problems as measured by profits for
3 of the 19 industries.

		When you switch to URS' cost, as you might expect, if URS is
estimating cost 60 times higher than OSHA's incremental costs -- in
fact, on this table, I've put OSHA's full cost.  You can pull out of
OSHA's spreadsheets the Agency's full cost estimates rather than the
incremental costs.  URS' cost estimate is some 30 times the OSHA's full
cost.  And in the center two columns, it's mostly yellow, which means
that for most all of the 19 industries, URS' estimated costs exceed 1
percent of profits or 10 percent of revenues.  Did I say -- yeah, 1
percent -- did I say that wrong, 1 percent of revenues, 10 percent
profits?

		And the last table I want to show is what happens when we make just
the small adjustments to profits, which I've talked about, which is
extending the profit series from 2006 to 2010, but including in the
profit calculations each year both profitable and unprofitable
industries.  The middle two columns that have yellow, let's see, well,
let's see, the third from the rightmost column is a comparison between
OSHA's full costs and redefined profits, marginally adjusted or somewhat
recalculated profits. 

		And in this case, OSHA's cost -- OSHA's full cost exceeds 10 percent
of profits for getting close to half of the affected industries.  So
even a marginal adjustment of -- in the way in which profitability is
calculated, combined with looking at full costs rather than incremental
costs, leads to substantial questions about economic achievability, not
to mention what would happen if we did further adjustments to better
reflect revenues currently, which I believe on a per-facility basis are
substantially less than revenues in 2006.  And further adjusting profits
to reflect 2011 through 2013, putting them in the equation, also.  So on
balance, it's our balance that there is a substantial case that, for
most of these industries, that the proposed regulation is not
economically achievable.

		JUDGE SOLOMON:  You have five minutes.

		MR. KING:  I think -- the one last point that I think Mr. Sessions
might not have focused on quite as much is the last two columns in this
last chart show the full cost as estimated by URS as a percentage in the
one case of OSHA's profits and the other case as a percentage of the
revised profitability figures.  

		And when you look at that, you see that virtually all the said
industry sectors exceed the 1 percent or 10 percent margins.  Moreover,
you find that for 10 of the 19 affected industry sectors, the URS full
cost exceeds 100 percent of the profits for the industry sector.  And we
would suggest that, where that is the case, there really isn't much need
to do further economic analysis, assuming you accept the numbers because
that's clearly going to show, we believe, economic infeasibility.

		Thank you.  I think that's the end of our presentation.

		JUDGE SOLOMON:  Do we have some idea of how many questioners?  Okay.  

		MS. LINDBERG:  Your Honor, if you don't mind, I think we'd all
appreciate a short break, if our panelists don't object.

		JUDGE SOLOMON:  Okay.  We'll take five minutes.

		(Off the record at 3:55 p.m.)

		(On the record at 4:00 p.m.)

		MS. TRAHAN:  Hi.  Chris Trahan with the Building Trades.  It's
T-r-a-h-a-n.  And I have a question, I think, for Dr. Lee.  Your
laboratory analyzes crystalline silica with a limit of quantitation of
what?  What is the limit of quantitation with your lab?

		DR. LEE:  I think per AIHA, we use a reporting limit rather than a
limit of quantitation.

		MS. TRAHAN:  Okay.

		DR. LEE:  And we use that on the basis of the lowest standard.  So --
and that's 5 µg.

		MS. TRAHAN:  Okay.  And do you feel if one of your clients received
their report that a sample contained 9 µg, they would be able to feel
confident in that result?

		DR. LEE:  Depends on what lab you're using.

		MS. TRAHAN:  Your lab.  Actually, my other job, my real job, is with
an organization that does use your lab for research purposes, and you
analyze our crystalline silica samples.

		DR. LEE:  Well, I think at 9 µg, the question is whether there's any
interferences.  If there's no interferences, then you should have a very
reliable result from us.  But if there's interferences, you might have
to get up around 15 to 20.

		MS. TRAHAN:  Okay, thank you.  And I guess I'll just leave it at that,
and maybe with a request, if you can share this ability that, I think
you and several other commercial labs do have, to have good accuracy and
precision at those low levels with the rest of the world, that would be
great.

		DR. LEE:  Okay.

		MS. TRAHAN:  Thank you.

		UNIDENTIFIED SPEAKER:  Can I add just one other factor, and with
respect to R.J. Lee Lab, it's also a research laboratory versus strictly
a commercial laboratory.  So their -- you know, their processes may not
be typical that you see out there.  So that's another facet of your
question.

		MS. TRAHAN:  That's important.  I think they're a great lab, so that's
good extra information.  Thank you.		

		MS. LINDBERG:  Just a quick request, since there are so many of you
over there.  If you could state your name before you start answering a
question, that would be helpful.  Thanks.

		JUDGE SOLOMON:  Right.  Ms. Seminario.  

		MS. SEMINARIO:  Peg Seminario, 

S-e-m-i-n-a-r-i-o.  I'm from the AFL-CIO.  A few questions for
Mr. Waggener and then for Mr. Sessions on the work that you did on
your testimony.  

		Mr. Waggener, on page 20 of your written testimony, you indicated
that you conducted various surveys and discussions with industry
representatives concerning OSHA's assumptions and cost estimates in a
number of different aspects of the cost analysis.

		What kind of surveys were they?

		MR. WAGGENER:  Well, they were actual surveys sent out to the
industry.  Also, a number of meetings and conference calls with industry
that were put together.  And not -- they're professionals in the
different industries.  Does that answer your question?

		MS. SEMINARIO:  So there was a written survey that was sent out?

		MR. WAGGENER:  Yes, that was one method, yeah.

		MS. SEMINARIO:  Could you provide that for the record, in terms of the
kind of information that you were attempting to seek, the kind of
questions and information?

		MR. KING:  You're talking now about the kind of a blank survey form, I
take it, and not individual company responses.

		MS. SEMINARIO:  Right, just what were they asked?  What was the survey
instrument?  What was the information that you were trying to obtain?

		MR. WAGGENER:  We'll get back to you.  These were all done in
confidentiality and to be filled in.  And I'm -- very likely they can
be, but we'll have to get back to you on that.

		MR. SESSIONS:  The surveys were done typically by individual trade
associations rather than by the panel as a whole.  And the question --
one of the questions would be whether the individual trade
associations -- some of the individual trade associations, in fact,
were not even members of the panel.  

		MS. SEMINARIO:  Okay.

		MR. SESSIONS:  The question is whether they would allow for the panel
to release those.

		MS. SEMINARIO:  Okay.  But the question I had was, looking at the
instrument, and then who was surveyed, so it was trade associations and
then not the individual employers themselves?  So where was the
information solicited from, as far as the written survey?

		MR. SESSIONS:  The trade associations issued the surveys to their
individual company members, who responded as a company.  Many of them
were facility-based rather than company-based.

		MS. SEMINARIO:  Thank you for that clarification.  And how many of the
surveys did you receive back?  How many were sent out and how many did
you receive back?

		MR. WAGGENER:  I don't remember the exact number, but I'd say close to
200.

		MS. SEMINARIO:  200.

		MR. WAGGENER:  It was a very substantial amount.

		MS. SEMINARIO:  And was this representative of the different industry
sectors that are members of the ACC Panel?  You folks represent a broad
range of different industries that are impacted by this rule.  So was it
a representative sample of responses that came back from the different
sectors?

		MR. WAGGENER:  I would say it was certainly very representative of the
different industries it actually was sent to, yeah.

		MS. SEMINARIO:  Okay.  Thank you.  As part of this survey, did you
request and did you receive any exposure monitoring data from the
facilities about their actual exposures that they were seeing in their
facilities?

		MR. WAGGENER:  I'm not -- we didn't -- I don't believe that we
actually got the actual numbers that came from the facilities.  I'd have
to get back to you on that.

		MR. SESSIONS:  It varied from survey to survey.  One survey, I know,
asked each plant what's the highest number you've seen, what's the
lowest number you've seen.  Another survey asked the plant for a given
job category for -- each of the job categories within the facility,
roughly what fraction of the exposures that you've measured are above
the current PEL versus below the current PEL.

		So some of the questions were fairly specific with regard to job
category and reference -- results across multiple samples.  I don't
think anybody asked for, could you give us your exposure data, and
turning in the actual data points, but various levels of
characterization of the pattern and level of exposure.

		MR. KING:  This is Mr. King.  Just for point of clarification, I
believe that the analyses that both URS and Environomics did basically
were using the exposure assessment data that OSHA had prepared and not
separate data from individual industries.  They weren't challenging
OSHA's exposure data as reflected in the Preliminary Economic Analysis. 
I just asked them to confirm; is that correct?

		MR. WAGGENER:  That is correct.  In the URS estimate, we used exactly
what OSHA had for the percentage of overexposed workers.

		MS. SEMINARIO:  So in some cases, for that particular item with
respect to the exposure distributions and numbers of workers, you relied
upon OSHA's data.  But when you looked at other areas, as far as a
number of workers that might be impacted by a particular control, you
developed your own estimates.  Is that correct? 

		MR. WAGGENER:  Everyone's different, okay?  There's no simple answer
to that question, other than if we talked to individual plants and to
find out how many workers.  I'll give you a perfect example.  A perfect
example in my presentation, I talked about the enclosed cabs.  In our
interviews and in our surveys, that information would only be one person
per cab per shift; that was information that was gleaned from that
information, and we used that in our model.

		MR. KING:  Also, just one other clarification.  Although they were
using, starting with OSHA's exposure data, they did make adjustments in
doing the costing.  For example, I think Mr. Waggener referred to the
fact that because of the switch to the ISO/CEN sampling protocol, they
assumed that the number of workers who were going to be exposed above 50
would be higher than what OSHA had assumed because the greater dust
collection.

		MS. SEMINARIO:  Okay.  Thank you.  Just returning back to the survey,
I understand that these were sent out confidentially and returned
confidentially.  But if it would be possible to provide the results of
the survey with the identities masked, so that there's some sense of
what the underlying information that was used to make this assessment, I
think that would be very helpful.  You folks did an awful lot of work. 
And so getting at that level of information, I think, would just be
helpful to all of us.

		One set of questions for -- going back to Mr. Sessions, you
indicated in your statement that it was your view that OSHA should
include in the cost of this standard the cost of coming into compliance
with the current 100 µg PEL.  Is that correct?

		MR. SESSIONS:  I'd like to put a bit of a distinction on that.  For
the purposes of economic impact assessment, I believe it's important to
consider the full costs, including both the cost to come into compliance
with the current PEL and the cost incrementally due to the reduced PEL. 
For economic impact analysis in particular, it's important to calculate
the full costs.

		A second point that I made is that in attempting to calculate the
incremental costs of the reduction in the standard, I believe that OSHA
substantially missed the mark in what the Agency attempted to do.

		MS. SEMINARIO:  No, I understand that point.  But the cost of reaching
100, that's an obligation that exists now and has existed since 1971 in
general industry; is that correct?

		MR. SESSIONS:  Correct.

		MS. SEMINARIO:  And I think as was in your statement or a number of
the comments here, that there are perhaps 30 percent of -- I wasn't
clear if it was samples, establishment, workers that are above the
existing PEL.  So there's a substantial amount of noncompliance
currently, correct?

		MR. KING:  According to OSHA's IMIS data, that appears to be the case.

		MS. SEMINARIO:  And according to the exposure estimates that you used
in your economic analysis.

		MR. KING:  Well, those are just taken from OSHA.  Those are OSHA's
estimates.  We didn't make any estimates.

		MS. SEMINARIO:  Okay, so based upon that, there's a substantial amount
of noncompliance.  Is it your expectation that employers who are not now
complying and haven't complied for 40 years with the existing PEL, that
all of them will now comply with the current PEL?

		MR. KING:  That they will be able to?

		MS. SEMINARIO:  No, that they would, that they would in terms of the
actual economic analysis of compliance, not the theoretical, the actual
compliance.  If 30 percent are out of compliance now, would you assume
that all of them will comply with a new PEL?

		MR. KING:  No, I think the analysis that Mr. Waggener presented dealt
explicitly with that point.  He said, given the fact that after 40
years, a substantial of employees still seem to be exposed above the 100
µg/m3 PEL, he does not believe that everyone is going to be exposed
below this proposed 50 µg/m3 PEL.  Indeed, we assume they'll be more
people exposed above 50 µg/m3, if that is adopted as a standard, than
OSHA assumes.  OSHA, I believe, assumes only 10 percent would be exposed
above 50.

		JUDGE SOLOMON:  Two things.  First of all, Mr. King, you didn't
identify yourself again.  And, number two, why are you speaking for the
gentleman who is seated to your left?

		MR. MORRILL:  If I may, this is Mr. Morrill.  He is functioning as
sort of our quarterback for answering the questions.  And so, we'll work
through Neil, and then if there are other responses, then certainly we
can refer to others.

		MS. SEMINARIO:  But in costing out the -- your cost estimates on the
rule, we're costing out the proposal to get to 50 for all the workers in
general industry who are subject to the standard.  Is that correct?

		MR. WAGGENER:  What we -- excuse me, Your Honor, Mr. Waggener
speaking.  What we did was we determined for the people that were over
100, we also determined what it would take to get to 100, assuming the
world was perfect, okay.  And the reason they're not below 100 now is
because the world is not perfect.  

		And I want to stop right there and say these people are not
necessarily out of compliance.  They -- for 40 years, they've been
trying to get into compliance, installing equipment, trying things in
different jobs, and they haven't been able to reach to 100.  So they --
most of those people, if not all, are in respirators to be able to
comply.  

		And oftentimes, not always, but most of the time, there's some
agreement with OSHA on a local basis with inspectors and so forth,
either formally or informally, for that to occur.  And that's what's
been happening in the real world.

		MS. SEMINARIO:  So in the real world, there is no noncompliance.  It's
just that it's been infeasible for them to get to 100?

		MR. WAGGENER:  I didn't say no.

		MR. MORRILL:  Jack, if I could just jump in quickly.  Your Honor, I'm
just wondering, given the fact that OSHA also wants to ask a number of
questions, we want to make sure that we keep the questioners to --

		JUDGE SOLOMON:  You know, I'm a trial judge.  And if you did that in
my -- one of my trials, I'd probably refer you over to the Bar
Association.  So he was in the middle of making an answer, and you
stopped him, and you're not to do that, okay?  

		So if you can answer, go ahead.

		MR. MORRILL:  I'm sorry, I'm -- I was just --

		JUDGE SOLOMON:  Let him answer the question.

		MR. MORRILL:  Okay.

		JUDGE SOLOMON:  He was just -- he was answering the question.

		MR. MORRILL:  I understand that, but we're trying to also manage the
time, given the instructions that you --

		JUDGE SOLOMON:  I've already ruled.

		MR. MORRILL:  Okay, thank you.

		JUDGE SOLOMON:  Okay, go ahead.

		MR. WAGGENER:  All right.  Yeah, thank you, Your Honor.  With regard
to the cost model, we put in what we believed would be controls to
better approximate the 100.  And then we put in controls we estimated it
would take to get to 50. 

		It is a model, and you have to recognize it's a model.  And that's
what we did.

		MS. SEMINARIO:  Thank you very much.

		MR. WAGGENER:  Okay.

		JUDGE SOLOMON:  Who's next?  Mr. Comai?  State your name and spell
your last name.

		MR. COMAI:  I'm Andrew Comai.  Last name is C-o-m-a-i.  I had a
question for the industrial hygienist who worked -- who reported out on
the preparation of the samples.  I just wanted to know what was the OSHA
method used in these labs?  Were they all the same method?  What was the
number?

		MR. BAILEY:  The -- each of the labs was certified by AIHA to -- and
I would -- they're using NIOSH 7500, I would think.

		MR. COMAI:  7500.  So in that method, is there an acid wash procedure
that's available?

		MR. BAILEY:  Yes, it is an option.

		MR. COMAI:  It's an option.

		MR. BAILEY:  I do not know --

		MR. COMAI:  -- lab.  So you generated some samples in a -- you rigged
up some samples, I'm just going to say -- use the word "rigged" with
non-silica material?  Or you said kaolin and feldspar, was that --

		DR. LEE:  Rich Lee.  Yes, half the samples were pure silica and half
the samples were either 50/50 kaolin or 50/50 feldspar with silica.

		MR. COMAI:  So why didn't you report out the differences between the
pure and the obscured samples?  Was that --

		DR. LEE:  That data is in our report.  

		MR. COMAI:  Okay.  It just wasn't up on the slide, so I wondered
what --

		DR. LEE:  Yeah, yeah, it just wasn't up on the slide.

		MR. COMAI:  Was there a difference?  Did that really generate some of
the randomness or chaos in the lab?

		MR. BAILEY:  This is Kelly Bailey.  I believe that the error
associated with what we saw in the -- all the variability we saw in the
samples, regardless if it was pure silica or not, would probably cloud
that difference.  So it was -- basically, there was not much difference
between --

		MR. COMAI:  So was it regardless of whether it was unpure or pure
silica.

		MR. BAILEY:  Yeah, I think that what was going on there was so large
that any kind of distinction like that would probably be masked.

		MR. COMAI:  You also did some size selection, so you took out larger
particles.  Why did you do that?

		DR. LEE:  Just to avoid size effects, so that we're attempting to
mimic the size that the cyclone is picking up.

		MR. COMAI:  So was there a lognormal distribution in the size you did?
 Did you have --

		MR. BAILEY:  In terms of the?

		MR. COMAI:  The log -- the particle size, was the particle size
distribution lognormal when you --

		DR. LEE:  Lee.  We did not measure particle sizes independently.  We
just did a sedimentation --

		JUDGE SOLOMON:  Just to clarify things.  That was Mr. Lee.

		DR. LEE:  I'm sorry, Rich Lee.  

		MR. COMAI:  So you didn't do a size analysis of the raw material that
you were putting down on the sample filter paper?

		DR. LEE:  For the Min-U-Sil, there's a normal size distribution.  For
the others, we did a cut to respirable size.

		DR. LEE:  Cut using air?

		MR. COMAI:  No, cut using liquid sedimentation.

		DR. LEE:  Sedimentation, okay.  

		MR. COMAI:  So did you ever control for -- with those labs, did they
report out on how many times they did an acid wash, if they got
something from -- something comes in the mail, we're not telling them
where it's from, they're having some trouble determining it, what's --
wouldn't you have to ask them, or they would report back?

		MR. BAILEY:  This is Kelly Bailey.  We did not inquire as to what they
did with respect to their treatment of the sample, whether they
acid-washed it or not.  It is basically like if they felt like they
needed to acid wash it to perform the analysis, then they would have.

		We were more interested in it the XRD analysis and also that -- how
the sample preparation, with respect to whether a muffle furnace was
used, a plasma asher was used, or THF was used.  And in the process of
doing this project, THF was found not to be usable, perhaps because of
the 0.8 micron pore size.

		MR. COMAI:  So when the lab felt like doing something, wouldn't --
that would depend on where the material came from.  So there's a
communication between the lab and the facility.  This is coming from a
foundry with -- or this is coming from a smelter.  These are the types
of contaminants.  This is how you should treat it.  This is how we've
treated it in the past.  None of that kind of discussion happened.  And
so, variability would increase if people -- if you're not telling
people where the sample comes from, it's -- they're guessing, in
essence, as to how they should treat that sample, how many times they
should wash it with acid.

		MR. BAILEY:  This is Kelly Bailey again.  I think that if a lab saw
the filter and had difficulty in getting the peaks that would indicate
calcite or something like that, then acid washing would be the
appropriate thing to do.

		MR. COMAI:  But you never controlled for that in your report.  You
don't know how many times they did it.

		MR. BAILEY:  No.  Many samples come in without a designation of where
they're from.

		JUDGE SOLOMON:  Do you have any more questions?

		MR. COMAI:  Yes, I do.

		JUDGE SOLOMON:  How many more questions?

		MR. COMAI:  Different topic, one question.  

		JUDGE SOLOMON:  Okay.

		MR. COMAI:  The -- you mentioned the -- this is the economic
analysis where you have -- you mentioned the -- maybe an air
conditioner for a cab on a front loader that's loading dusty granite. 
You assume that there's one person per control, and then each person has
their own control, and that was worked into your cost estimate?

		MR. WAGGENER:  To be specific, we assumed that there would be one
worker in that cab per shift.  And that -- if he was an overexposed
worker, that cab was controlling him.  If there were two shifts, then it
would be one person on each shift, so --

		MR. COMAI:  So sharing one front loader wasn't in your cost estimate? 
Multiple people using the same cab --

		MR. WAGGENER:  Well, it was, but it was one person per shift.  So if
it was a two-shift operation, then it would be two people that were
being covered.

		MR. COMAI:  So is it when you do a control -- so a local exhaust
ventilation control, where -- do we estimate a separate control for
each person when, say, a larger shop floor exhaust ventilation system is
installed?  Do you again assume that each person has their own --  Each
person has their own --

		MR. WAGGENER:  No, no.  The example I gave was -- that's a very
specific example.  That's why we did the binomial distribution to
determine, in some cases, there may be four workers covered by one
control at a device, say a grinding operation.  At another facility,
there may be one person controlled by that.  It depends on the facility
size.

		MR. COMAI:  So what was the maximum number of people controlled --

		JUDGE SOLOMON:  Mr. Comai, I'm going to let you go, okay?  You're
over.  

		Next, Mr. Frederick, come forward, please.

		MR. FREDERICK:  Thank you, Your Honor.  I'm Jim Frederick with the
United Steelworkers Union.  It's F-r-e-d-e-r-i-c-k.  A few questions for
the panel.  Did URS use only certified industrial hygienists cost for
the cost data in the estimate of performing their monitoring?

		MR. WAGGENER:  Oh, no.  We have numerous engineers involved that do
this.  

		MR. FREDERICK:  Can you clarify?

		MR. WAGGENER:  I'm sorry, I misunderstood your question. 

		MR. FREDERICK:  For the cost itself, did you utilize the cost of a CIH
to perform the actual air monitoring calibrating the pump, hanging it on
a worker, observing the sampling?

		MR. WAGGENER:  I don't remember the exact, but we did not use just a
CIH, which would be the most expensive person.

		MR. FREDERICK:  I'd like to try to go back to the data that was just
being talked about, about the controls per worker.  If I understood this
correctly in the presentation this afternoon, there were, at some points
in your cost calculations, calculated to be eight controls -- up to
eight controls per worker into the cost estimate.  So am I understanding
that correctly?

		MR. WAGGENER:  That was an example that I gave -- a real-world
example where a specific job in a specific industry may have eight unit
operations, and one guy on a shift is operating all of those.  So there
is one person per eight controls.  That was a rare instance.  And
actually that did not -- I should emphasize, that example was not in
our model.

		MR. FREDERICK:  Okay.

		MR. WAGGENER:  We never used anything of that nature.  And that was
one reason I said we felt like our model was conservatively low in terms
of the cost estimate.

		MR. FREDERICK:  Okay.  And then the cost estimate of controlling
exposure to silica, the United Steelworkers represent workers in small
businesses, medium-sized businesses, and large businesses.  Some of them
are member companies of ACC as well as many other employer
organizations.  Almost exclusively, our employers say to us that they
wish to provide a safe workplace for our members. 

		And item that I did not understand, and again, I read this differently
in your written comments than I understood in the presentation this
afternoon, that small employers will spend more than medium employers on
controlling the same hazard.  And, again, I read that differently in
your written, but I understood it to be stated as I just did in your
presentation this afternoon.  Did I misunderstand it this afternoon?

		MR. WAGGENER:  I don't think I said that, but I'm not sure.

		JUDGE SOLOMON:  We have a transcript.

		MR. FREDERICK:  Okay, fair enough.  That's all I have.  Thank you.

		JUDGE SOLOMON:  Okay.  Now we have NIOSH revisited here.

		MR. HEARL:  Thank you, Your Honor.  I'm Frank Hearl, H-e-a-r-l, from
NIOSH, Washington, D.C.  I have questions for Dr. Morfeld.  I didn't
want to make you come all the way from Germany without getting a few
questions.

		DR. MORFELD:  I hope so.

		MR. HEARL:  I want to ask a few questions about the sampling for dust
in Germany.  How do you sample for respirable dust in Germany?

		DR. MORFELD:  As it is done now or historically?

		MR. HEARL:  How about let's go with the --

		DR. MORFELD:  What is done now at the moment.  We use gravimetric
sampling, usually personal-based.

		MR. HEARL:  Personal-based.  Okay, is it using a cyclone?

		DR. MORFELD:  Cyclone, yes.

		MR. HEARL:  What flow rate does that cyclone operate at?

		DR. MORFELD:  So I'm not sure at the moment because this is not my
field.

		MR. HEARL:  Okay.

		DR. MORFELD:  I'm doing the epidemiology, but I can give you the
references.

		MR. HEARL:  Great.  And, historically, how was the sampling done?

		DR. MORFELD:  So typically there were a couple of different
instruments from precipitation and colorimeter used in the older times. 
And, therefore, in our studies we did for the porcelain workers and for
the uranium miners, we did comparisons between the new instruments and
the old instruments in the labs, and at working places so that we had
those conversion factors.

		MR. HEARL:  Right, so the old sampling, was it done on a task-based
basis instead of full-shift?

		DR. MORFELD:  It was done on full-shift during regular working shifts
and done additionally on the different conditions in the labs.

		MR. HEARL:  Okay.  Was there an adjustment that was made for unsampled
periods when they would collect those samples?

		DR. MORFELD:  Yes.

		MR. HEARL:  And was that -- did they assume that the concentration
was the same over the whole shift, or did they assume zero for the
unsampled period?

		DR. MORFELD:  So I -- in that case, I would refer to the two papers
by Darmin and Gouldner (ph.) who described this in detail.  I can give
you these details.

		MR. HEARL:  And so, the estimates that were -- that you referred to
for the threshold value in your calculations in the presentation, were
those concentrations being reported in terms of current German samples
or, I mean, what was the basis of those estimates?  I'll just leave it
at that.

		DR. MORFELD:  So all instruments that were used, so of course, we
incorporated also the old values.  We had measurements in the '50s when
transformed to gravimetric data in the porcelain industry in some
workers in the range from 1 to 2 mg/m3.  So we used also these data.

		MR. HEARL:  Right, and so it was just done with a variety of
conversion factors?

		DR. MORFELD:  Yes.

		MR. HEARL:  All right.  So, actually, I came across -- I was looking
to try to find out how the sampling was done just on my own and in the
audience.

		DR. MORFELD:  I can give you --

		MR. HEARL:  And there was a paper by Sun,  Chen, Yang and Bochmann --

		DR. MORFELD:  Right.

		MR. HEARL:  -- that from -- it was called "Quantitative Assessment
of Historical Silica Dust Exposures among a Cohort of Chinese Miners."

		DR. MORFELD:  This is Chinese miners --

		MR. HEARL:  Correct, same one, correct.  And then, that one -- I
think they had the conversion factor between the German-type sampling
and the U.S. sampling was 1.31 times the U.S. sampling would give you
the German for copper, 1.5 for iron, 1.8 for tungsten, and 2 for
pottery.  Does that sound correct?

		DR. MORFELD:  Yes.  They made these comparisons with the U.S. samples
and with the German samples.

		MR. HEARL:  Right.  Was that the historical German samplers or the
current?

		DR. MORFELD:  Historical and the current.

		MR. HEARL:  Historical and the current, okay.

		DR. MORFELD:  As far as I remember, yes, both.

		MR. HEARL:  Okay, great.  That's what I had.  Thank you.

		DR. PARK:  Robert Park, P-a-r-k, NIOSH Cincinnati.  Another question
for Dr. Morfeld.  I think there's a fundamental flaw in the analysis of
threshold for silica.  The profile likelihood procedure is fine.  It's
from Dr. Olm (ph.) in Germany.  I've done it myself on hexavalent
chromium.

		The problem is using the proportional hazard model, which makes a
fundamental assumption, which is that there's an estimable non-zero
baseline risk.  There is not for silica.  The baseline risk is zero.

		So this says "consequences."  It means that the nuisance parameters
are very unstable.  In this analysis, it means that the estimate for the
hazard ratio will have a high variance.  It means that it is a badly
fitting model.

		Now, one way to make it fit better is to redefine some of the silica
cases as non-silica, as not having silica exposure, for example, by
applying a threshold.  In your optimum model, 17 percent of silica cases
now have no exposure.

		JUDGE SOLOMON:  Is there a question there, Dr. Park?

		DR. PARK:  How can you justify that procedure?

		DR. MORFELD:  So the first is, the Cox models break down if you have
no cases in the reference categories.  So this was an analysis done with
the continuous exposures.  And similar as in other analyses with the
lung cancers, with the continuous exposures, the Cox model can run.  And
we check that carefully.

		So we did not have any conversion problems -- problems with
convergency of the Cox models.

		DR. PARK:  It could still be a bad model, right?

		DR. MORFELD:  No, not necessarily so.  It has been shown that in a lot
of instances, the Cox model is the limit of a Poisson model, if you use
the same.  And this is often applied because this can do estimate in
absolute risks, not in relative risks.  So if the distribution of the
cases is not too skewed in such a way that there are large areas of
exposure without any cases, then you are not running into trouble with
this model. 

		The second question, I think the other was the question of precision. 
We did include a couple of covariates and also have thrown out these
covariates to see whether our estimates differ a lot, whether the
variances differ, but this was not the case regarding the threshold
estimates.  So it was rather stable.

		And the third question, I think, was on this 17 percent of silica
cases set to zero exposure after applying the threshold value.  This is
true.  So the problem is, in estimating thresholds, that usually the
people think that a threshold should be lower, or at least as high as
the lowest exposure of a case in a study.  

		This is not the case if you use a statistical procedure estimating the
threshold.  So it can be that some of the cases then are lying below
this estimated threshold.  Usually if you -- we have got some comments
in between and discussions whether it makes sense to look at the lowest
exposed case.  But usually you have to rely then only on the one case
with his data and not on the complete cohort data.

		So this is what we tried to do with this procedure, use everything,
and of course, not rely on a single case.  So it is possible that some
of these cases, if all the radiographs are rewrote again, it would
become non-cases, and other non-cases would become cases.  So all this
is unstable, of course.  And, therefore, we need a procedure to estimate
the threshold that takes all the data into account, and then of course,
you can end up with a result that, if you apply the threshold, some of
the cases are below the threshold. 

		DR. PARK:  The same problem exists with Poisson regression, by the
way.  I have two papers that I'll be submitting to the post-hearing.  

		One other question, have you done this same analysis for lung cancer,
and what did you find?

		DR. MORFELD:  So the lung cancer, no, we did not.  So the lung
cancer -- relative risk of lung cancer is not elevated in the study. 
So --

		JUDGE SOLOMON:  Well, if you didn't do it, you don't have --

		DR. MORFELD:  So we didn't do it.

		JUDGE SOLOMON:  You did?

		DR. MORFELD:  We didn't.	

		JUDGE SOLOMON:  Right, so you don't have to answer the follow-up
question because you didn't do it.  Okay, thank you very much,
Dr. Park.  

		So Ms. Lindberg?

		MS. LINDBERG:  I'm going to let Bill Perry take over here.

		MR. PERRY:  Sorry.  Thank you.  We do have just a few questions for
this panel, Your Honor, if we may.  I believe first to go will be Janet
Carter.

		MS. CARTER:  Good afternoon, Dr. Morfeld.  I appreciate you coming
all the way from Germany to talk to us today.  I just have a few
questions.  Mine are going to be health-based.

		There's been considerable testimony over the past week regarding
poorly soluble particles and the concept of overload, and a threshold
effect for inflammation leading to tumor genesis and nonmalignant
respiratory diseases.

		The European Centre for Ecotoxicology and Toxicology of Chemicals
recently released a follow-up report to the International Life Science's
workshop report on poorly soluble particles and lung overload.  Both
reports define poorly soluble particles in their evaluation as including
non-asbestiform talc, titanium dioxide, diesel soot, coal dust, and
carbon black.

		Now, while the ISLI study didn't specifically include silica, the
ECETOC report specifically excluded crystalline silica due to its
phytotoxicity and reactive surface characteristics.  I was wondering if
you were of this European industry report?

		DR. MORFELD:  Yes, I know it.

		MS. CARTER:  Okay.  I appreciate that.  Do you have -- how do you
think it applies to what we're talking about today?

		DR. MORFELD:  So this ECETOC report mainly deals with the overload
question.  The question is whether the transitional toxicology that is
used to translate results from rat studies to humans are reliable or
not.  We have a big discussion on this.  You may know that the MAK
committee in Germany has released a document on this, how to calculate
from rat overload experiment limit values for general dusts.  

		And the ECETOC report is a parallel approach to view whether the
overload results from the rats are really reliable and can be
transformed into the smallest negative in the results about this.  They
said they should not use it for classification and should not use it in
the limit value assessment.  It is a major outcome of it.

		MS. CARTER:  Yeah, it actually went further to say that they shouldn't
use it for risk assessment either.  And ECE actually talked about that
as well.

		Your testimony also discussed the concept of general dust causing
nonmalignant respiratory diseases.

		DR. MORFELD:  Right.

		MS. CARTER:  And you gave an example of carbon black, and I was
wondering if you thought it was appropriate to use carbon black as a
surrogate for crystalline silica for such diseases.

		DR. MORFELD:  No, this was not for crystalline silica.  It was one of
the dusts that we believe that it was -- in the old terminology, more
or less a nuisance dust.  So it's granular biopersistant, but of course,
carbon black is nanostructure, there may be something else going on.  So
there are other studies, there's titanium dioxide -- pigmentary
titanium dioxide and other use.  

		So we see that all this dust has some impact on lung function, in
particular.  We see some impact on chronic bronchitis, so it is
difficult, in particular in those studies, that this diatomaceous earth
workers and others, that where we have mainly general dust measurements,
but only crude information about the crystalline silica content, then to
use only this crude information about crystalline silica content to
calculate crystalline silica concentrations and to relate these to the
nonmalignant respiratory diseases, and then forgetting about the main
information about the general dust.

		This is the point that I made.  And all these papers, we have talked
ECETOC particulates paper, or look at the MAK document.  What they are
saying is be careful.  The general dust itself is no nuisance dust.  It
causes lung function change.  It causes chronic bronchitis and things
like that.

		MS. CARTER:  Your testimony also talked about providing a reference
list that OSHA should look at.  One of them happened to be the IARC
monograph 2012.  The IARC monograph actually specifically talks about
three mechanisms.  

		The primary mechanism could be an inflammatory response, but it
actually talked about two others which were derived from in vivo data. 
And they felt they were substantiated by in vitro data.

		Have you looked at these alternative modes of action in consideration
of your evaluation for crystalline silica?

		DR. MORFELD:  Yes.  See, I attended that workshop, and I was also in
that meeting.  And there was a long debate about the direct genetic
effects.  And you see in the figure that we're drawn, that this is a
dotted line because it was not clear whether there was anything going
on, and they don't want to exclude that.  And if I remember correctly,
the line is dotted.  And this was a result of these discussions.

		And I'd just like to refer to this Borm paper, which I think is really
interesting because they try to make a quantitative comparison, not just
some qualitative things, quantitative comparison where the inflammation
and direct genetic effects appear on a similar level of dose.  So they
found out that there is a ratio of about 50 between them.  So
inflammation is always going on before there can be a genetic effect. 
This was the general message.

		MS. CARTER:  Well, I'm glad you brought up the Borm study because that
particular study was actually extrapolating in vitro data to in vivo
data.  And those specific authors, actually two years prior, had
published a study saying that it's really not appropriate to do those
type of calculations.  So I was wondering what you thought about that,
about the limitations of those type of extrapolations from in vitro to
in vivo to symmetric calculations.

		DR. MORFELD:  Yes, I know the groups -- this is a group of Ken
Donaldson involved from Edinburgh.  They said they should be careful,
but there is a paper with Ken Donaldson and Overdurst (ph.) on it, and I
think Borm also, where they tried together with Tran (ph.) to calculate
coming back to the general dusting, whether the in vitro data and the in
vivo data in rats come to the same conclusion about those, that is a
minimum necessary to lead to inflammation.

		So I know that they are critical about this, but there are two
implications where they try to overcome this missing link between in
vitro and in vivo.  So it's not so that they are totally against this.

		MR. PERRY:  Just testing the microphone back here.  I think Steven
Schayer has a few questions.

		MR. SCHAYER:  Oh, yes.  So, Dr. Morfeld, I just have a couple of
quick questions for you.  So the first was just about the cohort and the
follow-up period in your study.  So I was just wondering, so were all of
the 17,644 employees followed up through 2005?  And if so, how often was
each worker assessed for silica during the follow-up period?

		DR. MORFELD:  Yes, they were followed up through that because it --
you see, if you go to the paper, it is a little bit more complicated
because in mortality, you can usually define an end date, up to -- that
the follow-up is preceded.  If you do it for morbidity study, you have
to take into account the last information you have about morbidity.

		So this must be sent at the endpoint of the last radiograph.  We did
so.  And all the person-years reported in, correctly sends it at the
last radiograph we had available.  So there are usually no radiographs
available, just for a few exceptions, but you can forget about it beyond
working at the companies.  Usually all these radiographs are taken
during being at work.  So -- but we sent it exactly, so it does not
that we expanded the person-years up to the end of the mortality
follow-up.

		I don't know whether you are alluding at that problem.

		MR. SCHAYER:  Okay, thank you.  The other question I have is a little
different.  And it was more on did you look at all at uncertainty in the
exposure assessment and the impact it might have had on your study?

		DR. MORFELD:  So we did not incorporate, as we said in the discussion
sector, and a formal uncertainty analysis on that.  So we used the data
as is and tried to find out the threshold value with confidence limit
based on this uncertainty that is then reflected in the confidence limit
of the exposure data.

		What could be done in future work is, of course, that you add
additional errors to the data as reported and documented in the study. 
But the exposure assessments were done independently from any
information about the response.  So this is where we started from, and
so this is the very best first information we had.  I agree that this is
one step we can move forward.  

		MR. SCHAYER:  Great, thank you.  One other question I had.  Did you
consider at all or look at an ILO category of 1/0 or greater?  I know
you looked 1/1, and we were wondering, you know, what effect you think
looking at 1/0 might have on the results of the power to detect a
threshold?

		DR. MORFELD:  So we are unable to do so because after explaining a
little bit about the process, we are running through to reread the
radiographs.  We have 120,000 radiographs that were originally read by
physicians responsible for the insurance company in Germany, doing all
this medical screening.  It is an insurance company of the employers, a
specific organization in Germany, we have so-called BGs, they are doing
this.

		So when we had all this data documented, we know from a lot of studies
we did before in coal mining and in other settings, and the UK has the
same information, that you do not simply pick up the information as it
is given in the documents because it can be very misleading.  So we try
to reread this.

		But imagine if you have 120,000 radiographs, who will reread this at
least two times?  So that is impossible to do.  So what we started is we
draw a random sample, and then try to find out whether we can rely on
the data as given or not.  And we find out if the data says it is
negative, you will never find a 1/1.  This is what the result of this
was.

		So we then went back and used all originally categorized 0/1, 1/0, and
higher, and reread all these radiographs independently by B readers. 
So according to this procedure, this two-step procedure, we are unable
to identify the 1/0 as an endpoint in our study.  For this, we have to
reread all the radiographs.

		MR. SCHAYER:  Okay, great.  And my last question was just, and I don't
know if this is possible, but we were wondering if you'd be willing to
make any of the raw study data available for other people to take a look
at.

		DR. MORFELD:  In principle, yes.  I personally have no problem with
this, but this, of course, has to be discussed with the sponsor and all
others involved.  You know, we have a lot of trouble in epidemiology in
Germany and in Europe with data protection discussions.  They are really
complicated.  And it is not only on the owners of the data or from the
companies, but it is more or less for the data protection people.  You
have to be -- personally, I mean, I have no problems with this.  So if
there is a solution, yes.

		MR. SCHAYER:  Okay, thanks.  Yeah, maybe if you could look into that,
I know we'd like to see it.  But if not, that's okay, too.  Thank you.	

		MR. PERRY:  Tiffany DeFoe.

		MS. DeFOE:  My questions are for Dr. Morfeld also.  Dr. Morfeld,
thank you for sharing your study with us.  It's very interesting.  

		There are a couple of areas of information that aren't reflected very
completely, either in Dr. Mundt's paper that you based the analysis on
or in yours.  And I'm wondering if you might be willing to submit more
information on these in a post-hearing comment.

		One of them, Steve already touched on, would be a more complete
description of the follow-up for silicosis morbidity, how it was
generally conducted in the industries where it was conducted -- or in
the facilities, and the extent of it over all, a description of what it
looks like with the cohort, if that makes sense.

		And the other area of information that would really help us to have
more detail on is the annual exposure means that served as a surrogate
for concentration and urinalysis.  In particular, well, if I had my
wish, I would request for each individual in the cohort their case
status, their highest annual mean exposure in the work history, and
whether or not they were a worker who had probably prior exposure.

		And if that's too much to ask for, because I know that's a lot of
data, then simply for the cases, the highest annual mean exposure in
their work history and whether they were a worker with probably prior
exposure.

		DR. MORFELD:  So the first question was, if I remember correctly, on
the silicosis and the mobility follow-up.  So usually, in Germany,
according to law, the workers are invited to get radiographs after
silica dust exposure after every second year.  So were invited for a
radiograph, and these radiographs were read by physicians and stored and
documented.  So this was the basis of our analysis.

		So we had a radiograph on the average of about every second year for
every worker.  And so we followed up on the incidence of silicosis with
this process.  Yeah, so, this is the basis.

		Then, your second question, I think, was about the annual mean
concentrations, and that was estimated.  And do you want to know
whether -- how this relates to the shift values or whether there is
more uncertainty involved or what are you addressing?  Or do you want to
have numbers or --

		MS. DeFOE:  Yes, I'm actually asking for data for --

		DR. MORFELD:  Yes, so the -- I think most of the numbers are
published in the Bick et al. paper.  I don't know if you have that.  I
think I've got this in the docket.  It's published and, of course, if it
is possible, if you ask specific things, we can do additional analysis
to show this.  So yeah, I don't have a problem with this.

		But I think the major things are for the jobs, and the distributions
are published.

		MS. DeFOE:  Yes, Bick et al. did publish the information that they
calculated.  But the data that I'm asking for has to do with the annual
mean exposures by individual.

		DR. MORFELD:  Yeah.

		MS. DeFOE:  Because I think that this would help us get a better
handle on how -- on the implications of your analysis.

		DR. MORFELD:  Yeah, for the -- yeah, so what I published were the
cumulative exposures, yeah, and you want to have all -- annual
distribution of the annual exposure estimates.

		MS. DeFOE:  No, what I would like is for -- ideally for each
individual in the dataset.

		DR. MORFELD:  Yeah.

		MS. DeFOE:  Case status and their highest annual mean exposure.

		DR. MORFELD:  Yes, I can --

		MS. DeFOE:  And their work history.

		DR. MORFELD:  Yeah, this is not published, right.  I can do this.

		MS. DeFOE:  Thank you.  That'd be terrific.

		DR. MORFELD:  Yeah.

		MS. DeFOE:  Secondly, looking at Table 3 in your paper, just a moment,
this is where you were looking at the effects of different offset
parameter choices on your model.  And it seems -- the choice of offset
parameters seems to have a strong effect, certainly on the confidence
intervals around your threshold estimates.  And in one case, on the
threshold estimate itself, with -- apart from your main model, the
choice of offset parameters seems to lead to fairly wide confidence
intervals.  

		And I'm wondering why does the choice of parameter have such a strong
effect on the model results?

		DR. MORFELD:  Yeah, so this is interesting because I've got this
question already a couple of times.  So the -- we used the log
transform of the data because some of the workers had exposures of zero.
 You can't directly apply the logarithm, so you have to put some offset
in. 

		And you can go to textbooks of statistics that discuss what kind of
offset should be used.  And if you go to the book I cited there from
Broyston (ph.), you get the information that you should identify the
smallest unit of change of interest that is occurring in your exposure
data.  And this should be used as the offset.  We used this from the
textbook.  We did not estimate this offset.

		We followed that recommendation and put that in.  This is exactly the
.01, which leads to our main model we are working with.  But having seen
this in our group, we discussed, of course, that we then just relied on
this recommendation coming from the textbook.  And then did sensitivity
analysis with other epsilons, with other offsets.  This is where the
table comes from.  So it is not finding out which epsilon fit it best,
but just doing a sensitivity analysis.

		So we defined a priori what is our model according to the textbook and
then did the sensitivity analysis.  And the interesting finding is, if
we follow the recommendation, the AICs measuring the goodness of fit
were best.  So in all other cases, it gets worse in two directions.  One
is that the general fit of the model went down, measured in the AIC, and
the confidence interval got wider if we used an epsilon not following
the recommendation.

		You see, it did not -- perhaps there's a misunderstanding.  I tried
to make it clear in the paper.  We started with an a priori model
following the recommendation from the textbook, and later on, just
checked whether other epsilon values would lead to better fits against
the recommendation.  But this was not the case.  Therefore, we stayed
with this model as the best.

		JUDGE SOLOMON:  What page of the -- of his --

		MS. DeFOE:  That is one Page 1031.

		JUDGE SOLOMON:  Which is different pagination than what we --

		MS. DeFOE:  Of his article. 

		DR. MORFELD:  This is all talking about the article and the table --
the article.

		MS. DeFOE:  And so you select -- you stuck with the original --

		DR. MORFELD:  Yeah, so if we -- so the interesting thing is we used a
priori the recommendation coming from the textbook, what kind of offset
we should choose.  We used this and did the calculation.  And then
sitting there and discussing, they just said okay.  This is just a
recommendation from the textbook.  Let's do a sensitivity analysis.

		So we put in other offsets.  And the finding was that we got the best
response measured in AIC as goodness of fit when we followed the
recommendation from the textbook.  So there was no indication that we
should change our a priori model.

		MS. DeFOE:  And did you elect to use a log transform on the exposure
data and to retain that particular offset parameter based on the AIC
value, based on model fit?

		DR. MORFELD:  No, we did not so.  So we did not decide on the offset
on the basis of goodness of fit.  We used it as a recommendation from
the statistical textbook.  And then, just did a sensitivity analysis,
whether this was a good recommendation with a view to our data, and it
came up, yes, it is because other offsets led to worse fits.  So,
therefore, we stick with the a priori model.

		MS. DeFOE:  Yes, thank you.  And the other part was about your choice
to log transform the data.  Did you choose that due to the model fit, or
was there another -- was there a different reason that you chose that?

		DR. MORFELD:  Yes, we also used the linear models.  We described this,
but the fit was so worse in comparison to the log models that we stick
with that.  There were a lot of recommendations going to that.  If you
go to Kyle Steenland, he always used the log because he found out that
the goodness of fit is far better if you take logs.

		MS. DeFOE:  Thank you so much.		

		JUDGE SOLOMON:  Mr. Perry?

		MR. PERRY:  We'll play a little musical chairs here, if we may.  And
somebody take this before I get feedback.  Thank you.  Brian Albrecht
from our Salt Lake laboratory.

		MR. ALBRECHT:  Thank you.  The first question I have is for Paul
Scott.  In your written submissions to the docket and in your oral
testimony, you pointed out that OSHA had made a calculation error in
Table 4-B9, and thank you for catching that.

		And you also recalculated the means and the relative standard
deviations for sample loadings less than 70 µg.  Did you also
recalculate the means and RSDs, for example, for loadings greater than
70 µg?

		MR. SCOTT:  No, I did not.

		MR. ALBRECHT:  Okay.  I've got a few questions for Mr. Lee and
Mr. Bailey.  In your presentation today, you showed us some data where
there were six samples that R.J. Lee had analyzed with recoveries of 94
percent and an RSD of 9 percent.  But I've reviewed the docket
submission, and I don't know what the loadings were for those samples. 
And --

		DR. LEE:  Those were the ones with the graphs?  I'm sorry, Rich Lee. 

		MR. ALBRECHT:  Yes.

		DR. LEE:  The ones on the left, I think, if you look at the graph,
were 20 µg, and the ones on the right -- there were two graphs on each
slide.  The ones on the right were 40 µg.

		MR. ALBRECHT:  Oh, no, I'm sorry.  I'm referring to an earlier slide. 
I believe it was the eighth slide, where it was just a general overall
accuracy and precision.

		DR. LEE:  If it was the original -- I think two samples at each of
20, 40, and 80. 

		MR. ALBRECHT:  Oh, okay.  Then in the other data that was also
presented, and now I am talking about the slides that you had previously
thought, that there were four samples at 20 and there were four samples
at 40.  And I'm also looking in the docket submission for data for
these.  Could you supply to the docket in your post-hearing comments the
data for the preparation of these?

		DR. LEE:  I'm sorry, can you repeat that?

		MR. ALBRECHT:  In your presentation today, there are several sample
results that have come up, all of which were analyzed by R.J. Lee.  But
I don't see the preparation data for those samples anywhere in the R.J.
Lee submission.  So I'm curious to know if you could submit your
information on how those samples were prepared.

		DR. LEE:  Certainly.  They would've been prepared in the same way as I
described, but I can certainly -- 

		MR. ALBRECHT:  So they would've been prepared in the same way as the
other 100 --

		DR. LEE:  As the other 150.  

		MR. ALBRECHT:  So if that is the case, then when did you choose to
compare to nominal values rather than actual values?

		DR. LEE:  Because the -- we had not yet received the actual -- the
code translating from actual to -- from nominal to actual.

		MR. ALBRECHT:  All right.  You discussed somewhat about how the
laboratories were selected to perform these analyses, and it was
laboratories were selected based on geographic region and capability to
use an XRD instrumentation.  Is that correct?

		MR. BAILEY:  That's partly correctly.  I mean, we also -- was an
AIHA-certified lab using the latest certification criteria of April
2010.  They had to use XRD.  They -- eventually found out we had to
have muffle furnace.

		MR. ALBRECHT:  Well -- and that's the point that I'd like to clarify.

		MR. BAILEY:  This is Kelly Bailey talking, by the way.

		MR. ALBRECHT:  Thank you, Kelly.  I'd like to clarify that, as well. 
So some samples were sent to laboratories that did not use a muffle
furnace.

		MR. BAILEY:  From the original study, we did not expect that they
would have trouble dissolving the filter.  So we didn't really take into
account whether they had a muffle furnace or were using chemical.

		MR. ALBRECHT:  Okay.  And is there data for those samples and the
preparation of those samples, as well?

		MR. BAILEY:  They could not run a silica analysis because they could
not dissolve the filter.

		MR. ALBRECHT:  I understand that there's no result data, but is there
data for the preparation of those samples that were not included in the
result set?

		MR. BAILEY:  Oh, from R.J. Lee --

		MR. ALBRECHT:  Yes.

		MR. BAILEY:  -- the preparation -- yes, I believe there is.

		MR. ALBRECHT:  Could that be also submitted to the docket?

		MR. BAILEY:  Certainly, but there's no laboratory result.

		MR. ALBRECHT:  I understand that.

		MR. BAILEY:  Okay.  

		MR. ALBRECHT:  With respect to the fact that these PVC filters were
not soluble with the chemical treatment THF, I'm confused as to how a
physical dimension like pore size would affect the solubility of a PVC
filter.  Can you speak to that?

		MR. BAILEY:  That is only speculation, and we do not know why they did
not dissolve.  But in both cases, the lab used THF.  Maybe perhaps
there's less activity when you have a small pore size.  They're probably
used to dissolving these 5 micron pore filters in, let's say just for an
example, 30 minutes, and this one may have to take an hour to do.  I
don't know the -- why they did not dissolve.

		MR. ALBRECHT:  Thank you.  When R.J. Lee performed analyses by XRD,
what preparation technique do they use to prepare their samples?  Do
they use THF dissolution?

		DR. LEE:  Rich Lee.  Thank you.

		MR. ALBRECHT:  Thank you.

		DR. LEE:  No, we do not.

		MR. ALBRECHT:  What method do you use?

		DR. LEE:  We use the plasma asher.

		MR. ALBRECHT:  All right.  Would R.J. Lee mind submitting manufacturer
and lot number information for these .8 micron, 37 mm PVC filters that
were used for these sample preparations?

		DR. LEE:  We would not.  I mean -- Rich Lee -- we'd be glad to.

		MR. ALBRECHT:  That would also be very helpful.  Thank you, Rich.  

		The laboratories that did perform these analyses, do you know if they
wiped the walls of the cassette prior to performing the analysis on the
samples?

		MR. BAILEY:  This is Kelly Bailey.  No, we do not know the answer to
that.

		MR. ALBRECHT:  And is that a current practice at R.J. Lee?

		DR. LEE:  Rich Lee.  I'd have to check.

		MR. ALBRECHT:  All right.  Thank you.  Did the laboratories that
analyzed these samples report any kind of error that was associated with
their result?  I'm assuming because they are AIHA laboratories, they do
have to report.  AIHA accreditation does, I believe, require that there
is an error associated with the result.

		MR. BAILEY:  Kelly Bailey.  We did not see a laboratory result,
because these were blinded by us also, and so the only person who really
can answer that question is SOMA.

		MR. ALBRECHT:  Okay.  I'm coming near the conclusion here.  The other
idea that I'm really trying to wrap my mind around here is that the
alcohol suspensions used to spike these samples, you guys alluded to
perhaps that it was destructive to the filter, and that's why they
weighed less post-application of silica than before.  

		Did you assume that this was a systematic process where they would
lose, you know, 50 µg with every application, or was it a random error?

		DR. LEE:  Rich Lee.  The data -- we did an independent set of
measurements on alcohol interaction with the filters alone, and it
appears that it's about 70 µg, and maybe varied from 50 to 100.

		MR. ALBRECHT:  Okay.

		DR. LEE:  Speaking very broadly.

		MR. ALBRECHT:  And then, my very last question, and then I'll turn it
over to Daniel, when -- well, two more, sorry.  They're quick.  When
applied silica or Min-U-Sil to the filters that also had, you know, one
of these interferences also applied, was that two applications of this
alcohol to the filter, or were both of those powders suspended in one
alcohol solution?

		DR. LEE:  They were both suspended in one.

		MR. ALBRECHT:  So it was just one application?  

		DR. LEE:  I'm sorry, that was Rich Lee.

		MR. ALBRECHT:  Thank you.  And did R.J. Lee analyze any filters that
had been spiked with both Min-U-Sil and interference?

		DR. LEE:  I will have to check.

		MR. ALBRECHT:  Okay, thank you.

		MR. PERRY:  Daniel Johansen from our lab.

		MR. JOHANSEN:  One question I had about the Min-U-Sil.  Do you have
the lot number and -- for that, that you can submit, as well?

		DR. LEE:  Rich Lee.  I would believe so, but we'll have to check.  And
if we have it, we'll submit it.

		MR. JOHANSEN:  Okay.  Now, I notice that you did this as a blind
study.  And I know you guys are an AIHA-accredited laboratory.  Do you
guys treat your samples differently from the AIHA PAT Program than you
do your other samples that you analyze at your lab?

		DR. LEE:  I think the answer is we'd like to say "no."  But in point
of fact, they come in, and I think everybody's on their best behavior.

		MR. JOHANSEN:  So do you say that there's some bias maybe that's
associated with the --

		DR. LEE:  I can't say that.  But, you know, I think it's just human
nature.

		MR. JOHANSEN:  Okay.  Just to clarify, too, the Min-U-Sil is also
subjected to the sedimentation process.  Is that correct?

		DR. LEE:  Rich Lee again, and yes.

		MR. JOHANSEN:  Okay.  Out of curiosity, what instrument do you use to
analyze your samples?  And what is the maximum power that you use for
your generator?

		DR. LEE:  I mean, we use -- I'll have to check.  It's not -- it's
the derivative of Phillips, but -- Rich Lee again.  I'm just the owner.

		MR. JOHANSEN:  Okay.

		DR. LEE:  But I will provide that information.

		MR. JOHANSEN:  Okay.  I think that's all the questions that I have.

		JUDGE SOLOMON:  Mr. Perry?

		MR. PERRY:  Okay.  And we have Robert Bart.  I suspect he'll have more
than two or three.

		MR. BURT:  Good afternoon.  I wanted to start by asking some questions
about the cost model.  You mentioned binominal distribution model.  I
just want to be sure I understand what you did with that. 

		For a real simple example, if somebody -- if some job category had
two people in it, and that's all there were, and 50 percent of those
were overexposed, then the model would say one person per facility is
overexposed.  I'm just -- is that what it would say, or would it say
something different?

		MR. WAGGENER:  Well, it's not that simple.  This is Jack Waggener
speaking.  It's not that simple.  I mean, every job was looked at
separately, from every industry.  And the curve -- I gave an example
curve there, but there could be 400 other curves that look completely
different.

		MR. BURT:  So I understand that.  I'm giving you one industry, one
firm size, two people, 50 percent overexposure.  I hate to ask someone
to calculate on stage, but that's a pretty simple calculation.  Would
you say that means one person per facility is overexposed in that
sector, in that job category?

		MR. WAGGENER:  It's getting late in the day.

		MR. SESSIONS:  Let me jump in.  Stuart Sessions.  

		The binominal expansion would give the probability that any given
facility would have zero, one, or two overexposed workers, and would
calculate the fraction of the plants in the -- let's say we were
looking at very small plants, and there were hypothetically 500 of them.
 It would give the fraction of such plants that would have zero
overexposed workers in that job category, the fraction that would have
one.

		MR. BURT:  So you carried out the distribution across everyone.  So in
my 50 percent example, you'd have approximately 25 percent of the firms
with zero, 50 percent with 50 percent, and 25 percent with 100 percent. 
Just carrying out a simple binomial, you would end up saying there were
people with no exposures, and there were people/places where everyone
was exposed.  Is that what you did?

		MR. WAGGENER:  Why don't we get back to you with a real example to
answer the question?

		JUDGE SOLOMON:  That's Mr. Waggener.

		MR. WAGGENER:  Yeah, I'm not trying to avoid your question, but you're
oversimplifying the binomial, yeah.

		MR. BURT:  No, I understand, but this is a very critical issue to what
the cost model produces.  I mean, this is the key to many things that
follow.

		You talk about -- well, I would also add that it's particularly
important to figure out what happens when there are fractional workers. 
If the model produced an answer like .5 workers per facility, would you
say -- what would you do with that?

		MR. WAGGENER:  It only uses -- the binomial only uses whole workers. 
It doesn't --

		MR. BURT:  So you didn't simply average?

		MR. WAGGENER:  No, no.

		MR. BURT:  Okay.

		MR. WAGGENER:  Only whole workers, yeah.

		MR. SESSIONS:  This is Stuart Sessions again.  The details of how it
works are in the spreadsheets that Jack --

		MR. BURT:  Oh, I'm sorry, I meant to say I want to thank you for
providing the spreadsheets, but we just haven't had time to get into
them during this hearing.

		MR. WAGGENER:  I understand.  And I just want to say, this is supposed
to be constructive criticism we're doing here, and I've done this
before.  And we'll be glad to work with you guys in any way to show you
what we did, how we did it, at any time.

		MR. BURT:  Okay.

		MR. WAGGENER:  Yeah, yeah.  And I'm not the statistician.

		MR. BURT:  You make a statement in your written testimony, "The
percentage of facilities in the first quartile of the graph would
receive one engineering control for each overexposed worker."  So that's
effective -- that sounds to me like it's applying an engineering
control that OSHA intended for four workers to each worker.  Is that
what the model would do?

		MR. WAGGENER:  It would -- for those -- there are very few of those,
but where that -- on that back end of the statistical curve, then that
could be the case.  Those workers are assumed to be spread out far
throughout in the facility, and we need --

		MR. BURT:  And then, if there are two workers, you would apply, once
again, the exposure controls OSHA might have said were appropriate for
four workers.  Is that correct?

		MR. WAGGENER:  We would on the other end.

		MR. BURT:  And the same with n4 would be four?

		MR. WAGGENER:  That's -- you're --

		MR. BURT:  So couldn't this be summarized as no matter how many people
are exposed, you always need all the controls?

		MR. WAGGENER:  No, that is not correct.

		MR. BURT:  Why not?

		MR. WAGGENER:  That's not what the model does.

		MR. BURT:  I'm just reading what you said it did. 

		MR. WAGGENER:  I'm sorry?  I don't remember.  I don't know what -- I
don't remember saying that.

		MR. BURT:  I was quoting your own testimony --

		MR. WAGGENER:  Okay.

		MR. BURT:  -- that says if there's one more -- if 25 percent of the
workers are exposed, you get all the controls that OSHA prescribed.  If
half of them are exposed, you get all the controls OSHA --

		MR. WAGGENER:  I don't know what we're debating here.

		MR. KING:  Excuse me, this is Neil King.  I wonder if you could
provide the specific page reference.  Maybe this could be answered --

		MR. BURT:  Pardon me, Page 8 of the URS testimony, first paragraph.

		MR. WAGGENER:  The actual written testimony?

		MR. BURT:  Yes.

		MR. WAGGENER:  Okay.

		MR. BURT:  I was quoting.

		MR. WAGGENER:  I don't think I have that up here with me.

		MR. BURT:  You go on to say, and I'm still on Page 8 of the URS --

		MR. WAGGENER:  Where are you reading?

		MR. BURT:  Page 8 of the URS testimony.

		MR. WAGGENER:  Where are you--

		MR. BURT:  The original written testimony.

		MR. WAGGENER:  Yeah, I understand that, but where are we?

		MR. BURT:  The second paragraph.

		MR. WAGGENER:  We've got the wrong one.

		MR. BURT:  The first question was about the first paragraph.

		MR. KING:  You're talking about the -- excuse me, the written
testimony that was submitted earlier on?

		MR. BURT:  Yes.

		MR. KING:  Not today's written --

		MR. BURT:  Yeah.

		MR. WAGGENER:  You're talking about the top of the page?

		MR. BURT:  Yeah, I was --

		MR. WAGGENER:  Okay.

		MR. BURT:  -- just trying to sort through what it meant when you talk
about controls being applied.

		MR. WAGGENER:  Okay.

		JUDGE SOLOMON:  Give him a chance to read it.

		MR. WAGGENER:  Thank you, Your Honor.  Okay, that's what I believe is
correct.

		JUDGE SOLOMON:  Yeah, Mr. King is sitting right next to you.  Do you
need to read that into the record, Mr. King?

		MR. KING:  I -- let's see what the answer is and respond.  I --

		JUDGE SOLOMON:  This is an abbreviation.  Normally, what I would do is
I would ask you if you're familiar with that.  And then --

		MR. WAGGENER:  I'm familiar.

		JUDGE SOLOMON:  -- I would ask you did you write it and all that
stuff.  So I'm not -- I don't think we need to do that.  So you -- do
you know what he's talking about?

		MR. WAGGENER:  Reading this, I know what it is.

		JUDGE SOLOMON:  Go ahead, Mr. --

		MR. WAGGENER:  In the binominal distribution, which everyone looks
different, they're not the simple two-people distribution, we are
basically in the first core quartile or the quarter, the curves are
done, and we use one worker per control or one control per worker, okay?

		MR. BURT:  Um-hum.

		MR. WAGGENER:  In the second one, between 25 and 50 percent, we used
two workers per control, half of what OSHA did.  And for the 50 percent
and above, we believe that means you have 50 percent of the workers
exposed that you would need.  We used what OSHA said, which is four
workers per control, if that was what was used in the OSHA model.  

		We keep talking about four, but there are a few where there six and
eight.

		MR. BURT:  Right, let's stick to the four then.

		MR. WAGGENER:  Yeah, right.

		MR. BURT:  So if there's only less than 25 percent of employees
exposed, and say that works out to one employee, you would have applied
the controls OSHA thought appropriate for four employees.  Is that
correct?

		MR. WAGGENER:  If there was more than -- if it's --

		MR. BURT:  If it's under 25.

		MR. WAGGENER:  -- from the top 50 percent, yeah, yeah.

		MR. BURT:  But even the one, you would still apply all the controls
OSHA thought was appropriate it seems?

		MR. WAGGENER:  I'm not following the conversation here.

		MR. BURT:  It says "would receive one engineering control."  Are those
the engineering controls OSHA recommended?

		MR. WAGGENER:  I -- you're going to have to read to me and show me
what you're talking about because I --

		JUDGE SOLOMON:  Maybe we should have -- let's get it into the record.
 Do you want to do that?

		MR. KING:  Well, he's now talking about a different hypothetical.  I
think Mr. Sessions may be able to enlighten on this one.

		MR. SESSIONS:  If I could.  Stuart Sessions.  Basically, we're talking
about -- instead of saying "number of controls," the package of
controls that OSHA posits as being appropriate to address an overexposed
worker in a particular industry in a particular job category.  So that
may be, you know, material handlers in structural clay.  It may be one
control.  It may be four, I don't know.  You know, it's the package.

		And the question is, how many times does that package need to be
applied for workers who find themselves in facilities when there is an
unknown number of such workers in the job category that are, in fact,
overexposed.  So let's say we're talking about very small facilities. 
And on average, there are .5 overexposed workers in this job category in
each such very small facility.

		The decision rule -- I'm afraid I've lost you there, okay.

		JUDGE SOLOMON:  Maybe Mr. Burt will ask another question.

		MR. KING:  This is Mr. King.  Mr. Burt, was your question whether we
were assuming that the same package of controls that OSHA would be using
to protect four workers in a particular job category is the same package
of controls that, in some of these situations, we would be -- we would
say would be used to protect just one worker?  Is that what you're
asking? 

		MR. BURT:  Yes.

		MR. KING:  Whether there's a different --

		MR. SESSIONS:  Yes.  The answer to that is "yes."  You know, if there
is a single overexposed worker in the facility, then in this step of the
analysis, we are agreeing that that worker needs the package of controls
that OSHA posits as appropriate for that job category.

		And the question is, how often, given the distribution of overexposed
workers among facilities, how often does that package of controls need
to be installed in situations where there's only one worker who needs
it?  And how often is it in situations where there are two workers?  And
how often are there four overexposed workers who, in fact, can all be
protected by that package of control?

		So, ultimately, the question is how often do you need the package of
controls, given the pattern of overexposed workers in different-sized
facilities?

		JUDGE SOLOMON:  Okay.  Just so we straight all this out, that was
Mr. Sessions, and his colloquy started with the word "yes."  Now,
before that, we had several speakers.  I don't remember what the actual
sequence was, so maybe you can straighten out the record.

		MR. KING:  Yeah, Mr. King was just trying to clarify the question as
being whether the same package of controls would be used for four
workers as for one worker, and then the answer from Mr. Sessions was
yes.

		MR. BURT:  Yes, your clarification was very helpful.  Thank you.

		MR. WAGGENER:  I'm sorry that I didn't follow it, sir.  Jack Waggener.
 

		MR. BURT:  Moving down to the next paragraph, another aspect of how
the model works.  "For the purpose of this cost model, URS has
determined that a majority of the workers, greater than 50 percent in
the job category, are exposed above the PEL.  It is rational to conclude
that the threshold has been exceeded and that the existing controls are
totally inadequate and must be replaced."

		I want to be sure I understand what that's saying.  Let's say you
encountered a situation in which there were four workers.  Two were
exposed at 35, two at 60.  You would scrap all of the controls and start
over again.  That's what it seems to be saying.

		JUDGE SOLOMON:  So, Mr. Waggener, are you familiar with that?

		MR. WAGGENER:  Yeah, actually, I just found it.  Yeah, yes, that they
would need to be replaced with a more adequate system, yeah.

		MR. KING:  Your Honor, can I just interject a second.  If there are no
more questions for Dr. Lee, for example, he has a train to catch.  I
wonder when he could go -- is that right?  Are you folks done with
Dr. Lee at this point, for example?

		MR. PERRY:  Yes, I believe so.  I think at this point we'll just be
asking questions about the cost and economic analysis.

		JUDGE SOLOMON:  Ms. Lindberg, are you going to have any questions for
him?

		MS. LINDBERG:  I just have one for Dr. Morfeld.

		JUDGE SOLOMON:  Okay.  Dr. -- or Mr. Lee, Mr. Bailey.

		MR. KING:  And, Your Honor, do we have an idea about how much longer
this will go this afternoon, or this evening at this point?

		JUDGE SOLOMON:  It's pretty much up to OSHA.  Mr. Stone is in the
back, held in abeyance, I assume. 

		MR. BURT:  Sorry, this is my clock.  Okay.  We've had quite a bit of
discussion about assigning people to establishments and what kinds of
controls are needed.  And here, let me go back to the binominal model.

		Would it be -- you stated at one point, I think, that we start with
an exposure profile, rather than looking at individual establishments. 
Would it be helpful to analyze the actual distribution and individual
establishments so that we were, instead of speculating about a binominal
or another model, we actually took a look at it?

		MR. WAGGENER:  Well, obviously, I mean -- excuse me, Jack Waggener --
the more time we have, obviously the more time we can do that kind of
thing.  And that would be something I would definitely encourage OSHA to
do.  And I've done that before with other rules when there's time and
there's very specific information, yeah.

		MR. BURT:  Turning to another aspect of it, you talked about the need
for professional cleaning.  And I think someone guessed, and I think
accurately, that OSHA may have inadvertently dropped that when we made a
couple of successive changes to the model.  And it -- as a result, it
might be necessary.

		I was curious about the parameters that were used for that.  For
example, you've got a 20,000-production area for a very small firm.  And
this is an operation that would typically have six to eight persons. 
What's the basis for -- that seems to be pretty large.  I mean, that's
a football field up to the 40-yard -- a college football field up to
the 40-yard -- goal line to 40-yard line.

		Is there a basis for that, 20,000 square feet?

		MR. WAGGENER:  That, again, let's -- is that in -- are you reading
here now?

		MR. BURT:  Yes.

		MR. WAGGENER:  Okay.

		MR. BURT:  I'm sorry, I can -- Page 24, last paragraph.

		MR. WAGGENER:  Jack Waggener again.  Page 24.  From my experience in
seeing small facilities, that's not an unusual square footage for a
small facility in these industries.

		MR. BURT:  So these numbers are based on your experience?

		MR. WAGGENER:  Also, some of the surveys and discussions with the
industries that we talked to.

		MR. BURT:  Surveys as results we don't have access to?

		MR. WAGGENER:  I think we've answered that question and I --

		MR. BURT:  Thank you.

		MR. WAGGENER:  Yeah.

		MR. BURT:  We talked about you argue that we underestimated the costs
per -- of cfm, the first the capital costs.  I haven't had a chance to
pour over your testimony on that.  Will there be somewhere a worksheet
that shows what the basis for that much a higher number is?

		MR. WAGGENER:  I believe there is.

		MR. BURT:  Okay.

		MR. WAGGENER:  And it's -- the devil's in the details of all those
Excel sheets.

		MR. BURT:  Yeah.  And then did you -- you said that you applied the
25 percent factor OSHA used.  Did you separately try to be sure that
that was sensible, given the changes in the capital cost worksheet? 

		MR. WAGGENER:  Yes.

		MR. BURT:  In other words, for example --

		Yes?

		MR. WAGGENER:  Yes, sir.

		MR. BURT:  Okay.  We'll want to take a look at that.  At one time, I
believe in the lead rule, OSHA argued that very small facilities might
have somewhat smaller capital costs because they would need less
ductwork.  Conversely larger facilities might have higher.

		Do you have any comment on that -- those possibilities?

		MR. WAGGENER:  We're going back -- excuse me, Jack Waggener -- we're
going back 40 years now, okay?

		MR. BURT:  I know.  But I'm saying the broad -- I'm saying it's
argued that now, let's just say, in your experience, do smaller
facilities have lower costs to install these things than larger ones?

		MR. WAGGENER:  Editorial-wise, there are no more small facilities in
the lead business because they went out of business.

		MR. BURT:  No, I mean, in this record --

		MR. WAGGENER:  Right.

		MR. BURT:  -- we're talking 17 persons is the average size of a
facility, as you yourself pointed out at one place.

		MR. WAGGENER:  Okay.  They're very small.  I think it was less than
eight people.

		MR. BURT:  Yeah.

		MR. WAGGENER:  Yeah.  And would you repeat the question?

		MR. BURT:  Just is there any -- people have argued that very small
facilities might have lower capital costs for cfm than very large ones. 


		MR. WAGGENER:  I --

		MR. BURT:  There would be less ductwork in a very small facility is
one reason.

		MR. WAGGENER:  That's a possibility.

		MR. BURT:  You mentioned today that you had adjusted your costs in
some way to account for variability and exposure data.  What did that
involve?

		MR. WAGGENER:  We adjusted it for the ISO/CEN change.

		MR. BURT:  Um-hum.

		MR. KING:  Excuse me.  This is Mr. King.  Were you referring to the
initial exposure monitoring issue?

		MR. BURT:  I was referring to page 19 of your testimony today, where
it notes that additional monitoring will frequently give different
results.  Did you make an adjustment in your model for this, or you're
simply criticizing OSHA for failing to make such an adjustment?  

		In other words, did you take account of this, or did you simply want
to point out that was something that should be accounted for?

		MR. WAGGENER:  To answer your question, I do not believe that we
actually made an adjustment.  And we're just pointing out the
conservatives -- that we were conservative in our estimates. 
Jack Waggener. 

		MR. BURT:  There has been some discussion of the costs of getting to
100.  Let's suppose -- I just want to pose a hypothetical.  Let's
suppose OSHA somehow figured out how to truly enforce the standard of
100, and in the next three years did so.  

		Would you now say that we can exclude those costs from this record?

		MR. KING:  I'm not sure -- This is Mr. King.  Your hypothetical is if
you repropose this standard three years from now, at a time when no
employees are exposed above 100 µg/m3?  Are you say at what -- is that
what you're saying?  Would that be --

		MR. BURT:  I'm asking, would you say that, as a result, would the
economic impacts be less?  

		MR. KING:  Would economic impacts be less if all exposures were below
100 µg/m3?			

		MR. BURT:  Yes, we somehow succeed in enforcing the existing
standards.

		MR. SESSIONS:  This is Stuart Sessions.  I would say that the -- an
estimate -- any estimate of compliance costs is premised upon the
exposure distribution that one assumes that prompts the estimate.  And
if OSHA's exposure distribution were such that there's nobody exposed
above 100, then the costs that would be estimated for such a situation
would be quite a bit lower than the costs that are estimated for the
current situation.

		MR. BURT:  So are you saying that the economic feasibility of this
rule depends on the success of OSHA compliance, with the existing rule?

		MR. KING:  This is Mr. King.  I think what we're saying is the
economic impact of this proposed rule should be evaluated against the
current baseline of what exposures are today and how many employers
would therefore have to incur costs to install engineering controls, to
reduce those current exposures to a level of -- below 50.

		MR. BURT:  This is a question for Mr. King.  You've argued a few
times that, I believe, in your testimony that, to achieve 50, employers
would need to go to 25.  Well, it isn't.  The current data suggests that
to achieve 100, people are not going to 50.  They aren't even getting to
100.  How do you reconcile those two statements, those two?

		MR. KING:  I don't think there's anything to reconcile there.  I think
if people are currently having exposures above 100, that presumably
means that their -- on any single day when OSHA comes in to monitor,
that presumably means that their long-term average exposure in that
workplace is probably above something like 50 because on some days,
statistically they're going to be monitoring above 100, if that's the
situation.

		JUDGE SOLOMON:  Let me just say, that's pure argument.  We're not
supposed to take argumentative --

		MR. BURT:  Thank you, Your Honor.  One more question for
Mr. Waggener.  I think you mentioned testifying in the chromium
hearing.  What was the topic of your testimony at that time, if you do
recall?

		MR. WAGGENER:  It was with regard to costs, regard to technical
feasibility based on the NIOSH studies.  There were others.  I think it
was also on the measurability issue, also.

		MR. BURT:  Okay.  Thank you.  Let me turn to Mr. Sessions for a few
questions.  Now, I noticed you objected to our averaging profits across
both revenues for both facilities that did and didn't have profits. 
That did lower the level of profits, didn't it -- lowered the estimated
profit rates, didn't it?

		MR. SESSIONS:  Yes.  This is Stuart Sessions.  Yes, averaging across
facilities that have profits, as well as facilities that do not have
profits, lowers the average profit rate.

		MR. BURT:  One troublesome issue that we wrestle with on these things
is OSHA is interested in firms that will be viable, not in marginal
firms that might disappear before this regulation.  In any case,
including all of the people with negative returns would seem to possibly
include -- I think it might include a good many marginal firms.  What
do you think?

		MR. SESSIONS:  This is Stuart Sessions.  One thing I can say from the
data is that there are clearly, as you look across the IRS data from
year to year, there are different numbers of firms each year that are in
the positive profits or the negative profits column.  And the IRS gives
breakdowns by asset class, also. 

		And you see interesting things like, in one year, there is a big firm
that's reporting negative profits in a very big asset class.  In the
next year, that firm is not reporting negative profits because all the
data in that asset class is -- they're only profitable firms in the
next year.

		So, you know, I don't -- I understand perhaps that you're trying to,
you know, there is a reason in analysis to try to leave out what could
be called baseline closures, and that's appropriate.  But I don't think
that's what's going on in the IRS data, and particularly when you're
looking at it over, you know, 6 years, 10 years, 12 years, you know,
these are -- most of those firms are around for a good while.

		So, you know, my guidance -- my recommendation continues to be to
attempt to assess the ability of the entire industry to bear costs.

		MR. BURT:  In term -- if I could go back, if you object, I have no
problem with this, the other day you answered a question that small
firms have lower profits than large firms.  Am I remembering correctly?

		MR. SESSIONS:  This is Stuart Sessions.  I may have said that, in
general, to the extent I've looked at the data, for the most part, small
firms tend to have lower profit rates than larger firms.

		MR. BURT:  But would you find, say, the IRS breakdown by asset class a
reasonable way of addressing whether that's true from year to year?

		MR. SESSIONS:  Yes.  And, you know, there are issues about breaking
down the distribution the IRS gives you by asset class into, for
example, the SBA distinction based upon employee size and --

		MR. BURT:  Yes.

		MR. SESSIONS:  -- things like that.  That's all very tricky.  And,
you know, there are other data sources that are -- that will break
small versus large down in different ways.  The RMA data breaks it down
by sales, for example, and other data sources may provide different and
perhaps better, and additional, insights into this issue beyond what the
IRS can provide.

		MR. BURT:  The additional data sites -- sources you mentioned, RMA
and Dun and Bradstreet, are these random samples of firms?

		MR. SESSIONS:  This is Stuart Sessions.  No, they definitely are not. 
And an additional data source that I may not have mentioned today but
that I put in my written comments is any number of industry-specific
benchmarking and sample survey sorts of efforts, benchmarking or best
practices assessment and that sort of thing.  

		And my recommendation would be to try to sort of access all of the
relevant potential data sources for the industry in question, in order
to handle the sort of variety of different difficult issues that you've
got facing you, including, you know, the IRS data falls short on the
representativeness question, the three- or four- digit industry versus
six-digit.

		And some of the other data sources do better than that, but the other
data sources do have, as you point out, RMA in particular, those are
firms that are healthy enough to be looking for financing and then filed
their financial statements.  And so, there's some questions about the
representativeness of that data. 

		And, in fact, I think back in 2003, OSHA did use the RMA data.  And I
don't know why OSHA switched from that to the IRS, but my counsel would
be to look at both and try to use some amalgamation of what you can
glean from each.

		MR. BURT:  That covers what I have for now for today.

		MR. MORRILL:  Your Honor, I have one quick question.  I'm wondering,
because Peter Morfeld's been sitting here for a while, and I know you
have one question for him.  I'm wondering if we could skip to that one
question and then --

		MR. PERRY:  Yes, absolutely we can.  

		MS. LINDBERG:  Sorry to keep you waiting, Dr. Morfeld.  This is an
easy question, though.  You mentioned several studies in your
presentation today, and that OSHA's not sure it has in its -- in the
docket.  We're wondering if you would mind supplying them to the docket.
 It's the Borm et al. study from 2011, the Erin (ph.) et al. study from
2011 and the Solgul (ph.) et al. from 2012.

		MR. KING:  This is Mr. King.  Actually, we are planning to submit for
the record a whole slew of studies that were referenced in
Dr. Morfeld's testimony and comments but were not otherwise included in
the docket prior to that time.

		And these will be in hard copy because they will all -- they are
copyright protected, and we are very sensitive -- Dr. Morfeld and his
organization in particular, the University of Cologne, are very
sensitive about protecting the copyright features of these documents.

		MS. LINDBERG:  Sure, that's understood.  And that will include the
three studies I mentioned?

		DR. MORFELD:  Yes, of course.

		MS. LINDBERG:  Great.  Thank you.

		JUDGE SOLOMON:  You know, maybe I should -- I don't want to be a
buttinski here, and it's not really my province.  But when we had a
medical panel before, and I asked this question, and I'm going to ask it
now.  And that is, in the year 2000, the Department of Labor did
fact-finding with respect to pneumoconiosis.  

		Silicosis is a subset of pneumoconiosis.  And in the fact-finding that
they made, they made a determination that smoking, for example, is
additive to pneumoconiosis.  And they also discussed something called
legal pneumoconiosis.  Is there -- I just want to ask whether there's
any comment, any literature or anything that we need to know about?

		DR. MORFELD:  So this is a longer discussion, whether smoking is
really increasing the risk of developing silicosis.  We try to adjust
for smoking in our analysis because we expected to see this argument
popping up.  But there was no big effect in our analysis, at least.

		But we used smoking as a covariable to control for this argument, yes.
 So they're -- in some studies, it appears to have an effect and others
not.  But you have to be careful because the argument is made that, in
particular, in those workers who have been shown to have developed
silicosis, the physician asked more precisely about the histories.  

		So this can be a bias, or you get more precise and detailed smoking
information among the silicotics.

		JUDGE SOLOMON:  The only thing I would say is we have some other
physicians who have submitted their testimony and reports, so maybe you
want to take a look at that and respond to some of that.  Sorry that I
had to butt in.

		Okay, so the doctor can leave.  And let me just say for the record,
feeling dunk.  

		MR. PERRY:  And thank you very much, Dr. Morfeld, for appearing
today, this afternoon, and spending the day with us all the way from
Germany.

		And now, I believe Robert Stone is ready with a few questions.

		MR. STONE:  The good news, this will be brief because my colleague,
Bob Burt, I think, addressed most of my questions.  However, in your
discussion of the distribution of controls, I thought it understood it. 
Now, I'd like to run through it very quickly one time.  I promise it
will be very brief.

		As I understand it, you use our exposure profile, and you modified
based -- by 20 percent based on the ISO effect.  Is that right?  And
one -- if you get to 50 percent, you then apply the controls to all
workers in that job class that's at 50 percent.  Is that right so far?

		MR. WAGGENER:  I believe that's correct.

		MR. STONE:  Okay.  Now, my question where I'm now confused is are you
then -- let's say there would normally be one control for every four
workers.  Is that what you would apply to all the workers in the job
class depending on the number in each site?

		MR. WAGGENER:  If those higher of the two, yes, over the 50.

		MR. STONE:  Okay.  Then, there was the discussion about -- but you
may have to add or completely replace the current controls?  Or is that
also in effect -- or add more controls?  That was my last question on
this subject.

		MR. WAGGENER:  I think I'll have to get back to you on that one.  I
don't recall.

		MR. STONE:  Okay.

		MR. WAGGENER:  Yeah, yeah.

		MR. KING:  But -- and just to clarify the question -- so this is Mr.
King -- so that he'll know what to get back on, the question is, if for
a particular job category in an industry, the determination is made that
more than 50 percent of the workers are exposed above -- would be
exposed above the 50, the PEL, then would they assume that, for that job
category, controls have to be installed to protect all of the workers?

		MR. STONE:  That was part of the question.  And I think the answer to
that is "yes."  

		MR. KING:  And then, the second part was --

are there -- are the controls basically replacement controls?  You're
going to have to basically -- I forgot what the term was, you're going
to have to basically start from scratch and put in new controls for
these workers?  And are you adding more controls than you otherwise
would?

		MR. WAGGENER:  Again, it is not that simple because every job has a
number of different controls that we approach.  And it's those details
that are in the spreadsheets that I talked about.  I'm not avoiding your
question; it's just not that simple.

		MR. STONE:  Okay.  But if -- let's say we're -- let's say we're 30
percent, and you have certain controls, would you typically have more
controls for the ones that are over, if you got to 50 percent of the
workers?

		MR. WAGGENER:  I didn't understand that.

		MR. STONE:  Okay.  You have a certain set of controls for the job.  If
you have 30 percent of the workers, you would use your -- you'd apply
your formula, and then you'd have them as to who gets controls.

		MR. WAGGENER:  Yeah.

		MR. STONE:  Would you -- if they got to 50 percent, would you
actually have them have more controls than the people who were at 30
percent?  That's what was throwing me off now.

		MR. WAGGENER:  We're saying, when we get to the last half on the
curve, in 400-and-something different examples, did it have all kind of
variances in it, and that we are using one -- four workers per control.

		MR. STONE:  Okay, but they're not more controls for the people -- for
the people who are getting controls at 50, they're not getting more
controls than the ones who are getting it at 30 percent overexposed, for
the workers that are getting their controls.

		MR. KING:  This is Mr. King.  Again, just on clarification, I think
what Mr. Waggener's model is doing is trying to figure out in how many
instances will we have to install the package of controls.  And he was
saying where -- for a particular job category, if they find more than
50 percent of the workers are overexposed, then you're basically saying
for that job category, they're going to need controls pretty much right
across the board because it's obvious that the controls in place for
that job category are not doing the job that they have to do.

		Now, if you're saying it's another situation where, instead of 50
percent, only 25 percent, let's say, are exposed above, you wouldn't
have to install as many packages of controls.  But the package of
controls you would install would essentially be the same one as you'd be
using when over 50 percent aren't.

		MR. STONE:  Oh, okay.  That's it.  Okay.  I'm good, thank you.  

		Now, I have a few more questions actually for Mr. Waggener.  And
actually, it's really to clarify.  I think in your slides, you sort of
went out of your way to emphasize something, and I wanted to raise it.

		In two slides actually, in slide 9 and slide 11, I'll state what you
said in bold red -- in red on those slides at the bottom.  "Note:  OSHA
did not calculate very small and small facilities separately, but only
provided combined values.  URS recalculated these separately."  

		And then, with an added bullet, "OSHA calculation obscured the
significant cost impact on very small facilities."  Is that -- that's
what you said in number 11 at the bottom of the page, slide 11?

		MR. WAGGENER:  That's correct.

		MR. STONE:  Okay.  Well, actually, I don't believe that is correct. 
And let me just raise this for you to consider.

		OSHA separately divided throughout the PEA and in several places in
the preamble separate analyses and appendices for very small industries.
 I could point you to them if you wanted me.  But in addition, those are
completely just for very small.

		In addition, in several summary tables, we break out costs for all
facilities, all establishments, small establishments and very small
establishments.  Table 5.1, which begins the chapter, has that.  There's
an appendix to Chapter 5.  There's an appendix to Chapter 6, so there's
plenty of materials.

		We've carefully tried to break it out.  I think the confusion is that
you saw us presenting results for 1 to 500, not 20 to 500.  And we did
that because we have a legal obligation to the Reg Flex Act to consider
the effect on small entities, which is starting with one to everybody.  

		But we thought we were very careful to try to break out the cost for
small.  And we tried to take separate -- I had separate unit costs for
very small entities in many cases.  You may not agree with how we dealt
with it, but we didn't disguise it, and we didn't obscure it in our
view.

		MR. WAGGENER:  I think I accept everything you just said.  I mean,
I've seen the PEA personally myself and numerous other people.  And it's
how it was all presented.  So maybe obscured is the wrong word, okay?

		MR. STONE:  Okay.  And partially, I wanted this on the transcript
because, you know, even though I don't see anyone from SBA here, they
might look at the transcript and decide that we were trying to hide
things from them.  We were not.

		The second item is also a point of clarification, if I can find it. 
Okay, sorry.  Okay.  On slide 13, again, there are two places you sort
of emphasize this.  In one place, and this is -- I've heard this
actually throughout the hearings.  I've sort of not responded because we
can deal with it later.  But it's stated that OSHA basically estimates
the total number of overexposed workers nationwide, selected controls
assumed to be sufficient for four workers.  

		And are you aware that we did not do that generally?

		MR. WAGGENER:  No, I'm not.

		MR. STONE:  Okay.  Well, we have 18 pages of tables in some detail,
describing the controls and the number of workers for -- per control. 
It's actually in ERG (2007b), that's the reference.  It's a long title;
I'm not going to repeat it.

		And it's Table 3.3, Pages 3, let's see, 13 through 3 -- 30.  And
related to that in a later table, you specifically refer to an example. 
And the example is in the enclosed cabs for front-end loaders, OSHA
assumed the capacity of four overexposed workers per cab.

		That's not accurate.  In that same table on Page 329, and the only
place I think we specify a front-end loader cabs, we have two workers. 
And so -- and yeah, okay.  So I just don't know what the question was. 
Was he aware of it?  So this is actually for asphalt pavement products.

		Anyway, now, my last question concerns when you got to ancillary
provisions, and let me see if I can get to the page for that.

		MR. BOXERMAN:  Your Honor, may I have ask that -- it's five after
6:00.

		JUDGE SOLOMON:  He's probably either the last questioner or the next
to the last questioner.

		MR. STONE:  I think this is going to be about it.  

		MR. BOXERMAN:  He said this is his last question.  We've been very
cooperative.  It's long past the close date.  He's been sitting here for
five hours now and been testifying or questioned.  I think it's unfair
for OSHA to be continuing at this hour to continue to ask additional
questions.  And I would ask the chair for OSHA to reconsider whether
it's fair and reasonable and appropriate to continue questioning.

		We are obviously being cooperative.  We are here and present, but it's
way past where anybody else has been subject to this kind of
questioning.  I ask -- I request respectfully to stop.

		JUDGE SOLOMON:  Mr. Waggener, do you have any other --

		MR. WAGGENER:  Apparently not.  I'm fine, thank you.

		MR. PERRY:  If I may just ask one easy one, okay?  I just want to make
sure I understand that you used OSHA's exposure profile, and you made an
adjustment for the change from the existing curve for particle size
sampling criteria to the new ISO/CEN curve, which you say would average
about 20 percent greater dust mass being collected.

		So you kind of shoved everything up by 20 percent; is that correct?

		MR. WAGGENER:  It's Waggener, excuse me.  That is -- those details
are in the spreadsheet.  However, it does not shove it up -- it doesn't
give you 20 percent more employees.

		MR. PERRY:  No, I understand -- I think I understand that.

		MR. SESSIONS:  This is Stuart Sessions.  I can respond directly to
that.  The answer is we assumed in the exposure interval below 50, let's
say we're dealing with the ISO/CEN adjustment increasing the number of
workers exposed above 50, what we did was look at the number that OSHA
gave that were exposed between 25 and 50, and 20 percent increase to
41.2, I think gives 50.

		So we took 8.3 of the 25 and assumed a uniform distribution across the
lower exposure.

		MR. PERRY:  Okay, I understand.  Would you do something different if
most of those exposures -- if that exposure profile was drawn mostly
from exposure data using measurements that were taken with devices that
conform to the ISO/CEN curve, such as aluminum cyclone?

		MR. WAGGENER:  If it was done the proper way as ISO/CEN, we wouldn't
be adjusting it.  Yeah.

		MR. PERRY:  Okay.

		MR. WAGGENER:  I believe, and I stand to be corrected, that most of
the data or all the data in the PEA is not based on the ISO/CEN, I don't
think.

		MR. PERRY:  Okay.  Well, I mean, we'll look at that, but I believe
that is -- I believe we are done.  On behalf of the Agency, I really
thank Mr. Morrill.  I really thank the ACC and the Silica Panel for the
afternoon and early evening you spent with us.  You have been very
patient, and we appreciate that.  I think we learned a lot from you
today.  So thank you, and we'll look forward to your post-hearing
submissions.

		JUDGE SOLOMON:  Okay.  So if there's nothing else from anybody, the
time is 6:10, and the hearing is closed. 

		(Whereupon, at 6:10 p.m., the hearing was continued, to resume the
next day, Thursday, March 27, 2014, at 9:30 a.m.)

C E R T I F I C A T E

	This is to certify that the attached proceedings in the matter of:

INFORMAL PUBLIC HEARINGS FOR THE PROPOSED RULE 

ON OCCUPATIONAL EXPOSURE TO

RESPIRABLE CRYSTALLINE SILICA

March 26, 2014

Washington, D.C. 

were held as herein appears, and that this is the original transcription
thereof for the files of the United States Department of Labor,
Occupational Safety & Health Administration.

				 							    					____________________________

				    	ED SCHWEITZER

				    	Official Reporter

_________________________

		Continued

 PAGE   2137 

  PAGE  2137 

Free State Reporting, Inc.

1378 Cape St. Claire Road

Annapolis, MD 21409

(410) 974-0947

			 

Free State Reporting, Inc.

1378 Cape Saint Claire Road

Annapolis, MD 21409

(410) 974-0947

			   PAGE  2137 

