what foundries need to know thomas slavin · what foundries need to know thomas slavin slavin osh...
TRANSCRIPT
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA’S SILICA RULEMAKING,
WHAT FOUNDRIES NEED TO
KNOW
Thomas Slavin
Slavin OSH Group, LLC.
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Presentation Overview
• Background on Silica Rulemaking
• Key Issues for Foundries
– Health risk
– Technological feasibility
– Economic feasibility
– Standard provisions
• What happens next
• Q&A
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Respirable Crystalline Silica
• Respirable:
4 micron
• Thoracic:
10 micron
• Inhalable:
100 micron
1 4 7 10 30 100
Size (diameter) in microns
INHALABLE
THORACIC
RESPIRABLE
3
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Respirable Crystalline Silica
(SiO2)• Crystalline
forms:– quartz– cristobalite– tridymite
• Amorphous forms: – glass, – diatomaceous
earth
• Vitreous forms– Fused silica
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Respirable Crystalline Silica
• Current PEL formula is for total respirable dust containing silica (quartz)
• New PEL is for quartz only100 µg/m3 of quartz
• Lake sand is ̴ 95% quartz
• Respirable foundry dust is ̴15% quartz
5
Dust PEL = 10 mg/m3
(%Quartz + 2)
PEL = Permissible Exposure Limit
µg/m3 = micrograms per cubic meter
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
• Lowers quartz PEL by 50% from 100 to
50µg/m3
– Cristobalite PEL already at 50µg/m3
– Construction PEL at 250µg/m3
• Contains several ancillary provisions
– Air monitoring, medical tests, training
– Regulated area
• Prohibits sweeping and compressed air
• Dismisses rotation and respirators as
suitable control methods
Proposal Highlights
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
• Health risk-
– Is there a need for a standard?
– Will a standard lower risk?
• Technological feasibility
– Is compliance possible?
– Is measurement accurate?
• Economic feasibility
– Will business impact be acceptable?
Key Regulatory Issues –
What OSHA Needs to Show
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Silicosis Deaths and Death Rates
1968-2007
This is a Great
Public Health
Success Story
8
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA Risk Estimates • Uses other silicosis plus other diseases (lung
cancer, renal disease) to establish risk
• OSHA analysis faulted for
– Study selection bias (omits many foundry studies)
– Data selection bias
– Model selection bias
– Model uncertainty bias
– Model over-fitting bias
– Confirmation bias
– Investigator bias
– Specification errors
– Threshold smoothing and shifting bias
– Uses models that cannot detect threshold
Every
Bias
in the
Book?
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Vermont Granite Shed Studies
Attfield and Costello (2004)
• 5414 workers
• 201 lung cancer cases
• Through 1994
• Excluded highest exposure
group
• Incomplete worker status
follow up
• Inaccurate exposure
assessments
• Inaccurate worker categories
• Lung cancer risk
• OSHA relies on
Vacek (2011)
• 7052 workers
• 356 cases
• Through 2004
• Included all exposure
groups
• 162 workers assumed alive
in 1994 were actually dead
• 5204 more exposure
measurements
• Job assumptions corrected
• No lung cancer risk
• OSHA rejects10
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Bottom Line on Health Risk
• Dramatic reduction in silicosis cases
• No evidence of lung cancer or other diseases
without silicosis
• Evidence of threshold above 100 mg/m3 for all effects
• Improved compliance could eliminate residual risk
• BUT… OSHA has considerable discretion in
interpreting health risk data
• Comforting to us…
...But NOT compelling to Court
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Key Foundry Issue:
Technological Feasibility• One gram of silica sand (same as artificial
sweetener packet) would generate
exposure level above proposed PEL in
space the size of football field 13 feet high
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
• Definition of compliance - What does
meeting the proposed PEL mean?
1. All employers all of the time
• (OSHA enforcement)
2. Most employers most of the time
• (OSHA proposal)
3. Any employer ever
• (OSHA feasibility analysis)
• Many foundries use definition #2
• Measure of confidence needed
Key Foundry Issue:
Technological Feasibility
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Compliance in OSHA PEA vs.
Compliance in Real World
Number
of
Samples
Proposed PEL
Compliance OSHA PEA Feasibility
AB
Exposure Levels
OSHA PEA considers control to be feasible if case study (A) was ever able to achieve proposed PEL; Compliance requires assurance that PEL is met
with some level of confidence (B).
14
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
GM Defiance Foundry
Proposed
PEL
Current
PEL
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Does this show feasibility?
Proposed
PEL
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
GM Saginaw Data
• 622 samples (1998-2013)
• Geometric mean = 22 µg/m3
• 25% exceeded proposed PEL (50 µg/m3)
• 13% exceeded current PEL (100 µg/m3)
• 5% exceeded 200 µg/m3
• What mean level must be achieved to be
in compliance?
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Foundry Exposure Data
Mean is 46 mg/m3 but 16% of samples exceed 115 mg/m3
From study by R.C. Scholz, 2014
18
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA Feasibility Analysis Uses
Wrong Target
• OSHA issues citations if any exposures exceed
PEL, even if mean is below PEL
• To be confident that proposed 50 mg/m3 PEL is met
84% of time, mean exposure must be below 20
mg/m3. (much lower to reach 95% level)
• Engineering control target must be 20 mg/m3 or
less.
• OSHA feasibility conclusion based on wrong
target.
19
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Control Capability – ISO 9 Clean Room
Could Exceed Proposed PEL
20
Silica Mass
>=0.1 µm >=0.2 µm >=0.3 µm >=0.5 µm >=1 µm >=5 µm µg/m3
ISO 1 10 2
ISO 2 100 24 10 4
ISO 3 1,000 237 102 35 8
ISO 4 10,000 2,370 1,020 352 83
ISO 5 100,000 23,700 10,200 3,520 832 29
ISO 6 1,000,000 237,000 102,000 35,200 8,320 293 0.07
ISO 7 352,000 83,200 2,930 0.66
ISO 8 3,520,000 832,000 29,300 6.63
ISO 9 35,200,000 8,320,000 293,000 66.29
ISO 14644-1 Cleanroom Standards
Class
maximum particles/m3
Exceeds Proposed PEL
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Control Capability Overstated
• OSHA assumes that engineering controls will produce
same reduction in all operations
• OSHA treats controls like recipes in a cookbook that
have never been tried before
– Foundries have extensive experience with virtually every
control discussed: ventilation, enclosure, vacuuming, non-
silica sand, etc.
– None have magic powers to reduce silica below proposed PEL
throughout the industry
– Some listed controls (e.g. pneumatic sand transfer systems)
cause more problems than they solve
– Some controls (e.g. sand substitution) cannot be used in all
operations
21
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Adding Control Capability
• Cleaning and finishing example (IV-170)
– 69% reduction in one foundry with downdraft ventilation
• No evidence of continued performance
• Assumes same reduction for all exposed operators
• Many foundries already use downdraft
• Controls chosen for optimum applicability in one situation
– 67% reduction for precleaning
• Silica due to burned in sand, not dirty castings
• Most foundries already preclean
• Additive approach (90% total reduction) not warranted
– Subtracts additional 38 µg/m3 for reduced background
• Fully effective additive assumption is not appropriate
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA Uses 18 Case Studies to
Demonstrate Technological Feasibility
• Isolated data points taken out of context
• Case studies actually show:
– Controls often unsuccessful ( for current PEL)
– Control are more difficult (several iterations)
– Implementation often takes several years
• “It is reasonable to expect that on any particular day an overexposure to silica could occur.” – OSHA area director referring to sample result used by OSHA to show feasibility of compliance with proposed PEL
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Technological Feasibility
Bottom Line1. Despite extensive, expensive and sincere
efforts, many foundries find it difficult to reliably meet the current PEL for certain operations and will not likely be able to meet the proposed lower PEL.
2. Cases used to show feasibility, show opposite
3. Real world compliance target (continuous control) is much more difficult than the test used in OSHA’s feasibility determination (occasional control)
4. Traditional control methods may not be capable of clean room level dust control
24
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Economic Feasibility
25
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA URS/
Environomics
Annualized Cost
(incremental)
$44 million $2,200 million
Annual Cost as
percent of
revenue
0.2% 9.9%
Annual Cost as
percent of profit
4.8% 276%
Widely Different Cost
Estimates
26
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Economic Feasibility –
How Can Estimates Be So Different?
1. Marginal cost error
– Assumes no greater cost to achieve 50 than 100µg/m3
– Industry estimate is 5 times more to get to 50µg/m3
– Room cooling analogy
2. First Time Success Error
– Assumes control works perfectly the first time
– Cases show lengthy multi-step process
3. Under 50 error
– No costs counted to reduce exposures below 50µg/m3
– May need to reduce exposures to 15 or 20µg/m3
27
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Economic Feasibility –
How Can Estimates Be So Different?4. Discounted cost
– No cost for 2/3 workers exposed above current PEL
– Assumes no additional cost to go from 100 to 50
– Assumes no costs under 50
5. Per worker calculation
– Divided costs of controls by number of workers (typically 4) assuming each control would protect 4 workers
– Employees may operate multiple machines
– No cost for automated operations
6. Understated costs
– Ventilation at $5/cfm
– Industry experts estimate $20 plus $7 make up
– EPA estimates $5 to $40 (2 million cfm to 5000 cfm)
28
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Economic Analysis
Unit Costs & Quantity Understated
HEPA Vacuum
OSHA Assumption:
15 gallon HEPA Vacuum
$3,495 initial cost
$1,009 annual cost
Foundry: 2 cubic yard 40 HP
system with HEPA filter ($45,000
initial cost plus $15,000 hoses and
attachments)
29
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Abrasive Blasting Case Study
• Several steps documented in case history
– Replaced shot blast machine and ventilation ($107,000
for used machine)
– Covered and ventilated conveyor from shakeout
– Added skip buckets to load shot blast
– Covered conveyor next to shot blast
– Added enclosure and ventilation to shakeout exit
– Added ventilation to skip bucket enclosure
• OSHA cost estimate for abrasive blast operator
– $1,349 per worker for improved maintenance (based on
$8,000 Norton blast cabinet)
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
7. Controls Listed but Not Costed
(Only 24 of 46 controls costed)
• Substitution of Non-Silica Sand ($2 billion)
– 3 million tons of silica sand/yr times extra $700/ton
• Pneumatic Sand Transport System ($150 million)
– If used by 25% of foundries
• Production downtime to clean before beginning maintenance ($300 million)
• Professional cleaning ($1 per sq. ft. plus $400 million downtime)
• Automated Knock-Off
• Isolate Pouring & Shakeout in Separate Room
• Precast Refractories
• Didion drum to clean scrap for furnace operators
• Non-silica core coatings
• Low silica refractory
• Automated abrasive blast to pre-clean castings before finishing
• Wet methods31
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
7. Omitted Costs • Many costs and controls are omitted entirely
• Discussed in technological section but omitted from economic section
– E.g. Precleaning castings
– $1349 for maintenance* but nothing for blast unit
– *based on $8000 blast unit
• Missing operations
– Cut off saws
– Torch cutting
– Arc Air operations
• Ductwork, engineering design, and installation
• EPA modeling and permitting
• Upgrading baghouses to BACT if required
32
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
8. Other Errors
• 10 year replacement cost assumption lowers annual capital cost (replacement needed more often in abrasive environment)
• Profit calculation used inflated profit number
– 2000-2006 years used (pre-recession)
– Industry profit based on firms reporting profit; did not include those reporting loss
• ISO/CEN sampling change lowers PEL by 20-30%
33
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA Proposed Standard
Specific Provisions
• Regulated Area
• Contaminated Clothing
• Prohibit Sweeping & Compressed Air
• Methods of Compliance
• Monitoring
• Compliance Dates
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA Proposed Standard
Regulated Area
• Demarcate areas where overexposures exist and
require respirators for anyone inside
• Unmanageable - Based on TWA exposure of employees
who may move around vs. concentration in an area
(e.g. maintenance worker who spends part of time on
office)
• Unnecessary – Adds no protection
• Costly - Increases number of people in respirator
program (engineers, quality, material handlers,
maintenance)
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA Proposed Standard
Contaminated Clothing
• “Grossly Contaminated” is not defined
• Refers to visible dust NOT respirable
• Study (Exponent) shows no appreciable
exposure from dirty clothing
• When entering or leaving regulated area:
– Clothing change
– Vacuuming clothing (8 minutes)
– Blow off booth (30 sec cycle) - $11,000 (cost not
included)
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA Proposed Standard
No Sweeping or Compressed Air
• “if Contribute to Overexposure”
– Not defined; but referred to as prohibition in question 66
• Restriction on air wands for cleaning floor, rafter blow
down may be reasonable
• Cleaning molds, patterns and core machines may be
difficult and hazardous (put worker in danger zone)
• Automated processes (e.g. molding machines) may not
be able to function
• Wet methods not suitable and may be dangerous
(rust, molten metal hazard)
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA Proposed Standard
Methods of Compliance
• Must use engineering controls, even if NOT
effective
• Employee rotation prohibited as control
(may be used for other reasons)
• Respirators are NOT allowed as a control,
but required until controls are effective
• Removed exemption for respirator as
control for exposures less than 30 days
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA Proposed Standard
Exposure Monitoring
• New monitoring requirements
– Quarterly monitoring for every employee above PEL
– Semi-annual monitoring for every employee above action level
(25mg/m3)
– Full shift samples
• If no new information is to be learned, the
requirement is unproductive & punitive
– Other sampling options (e.g. real time monitoring, area mapping)
may provide information on sources and controls
– Waste of resources that could help with controls
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
OSHA Proposed Standard
Compliance Dates
• Effective 60 days after final rule
announced
• Only 1 Year to install engineering controls
– Ignores potential permitting issues
– Ignores time to design & implement controls
– Ignores fact that accurate air sampling not
available; labs have 2 years to meet accuracy
requirements
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
WHAT HAPPENS NEXT?
41
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Background on Silica Rulemaking
– OSHA spent years developing proposal
– 2012 SBREFA panel
– 2013
• September- Published proposal (Only 152 Days to analyze
proposal and comment)
– 2014
• March - Hearings
• June - Post hearing comments (OSHA Data dump to
docket early June)
• August 18 - Post hearing briefs (docket now closed)
– Entering dark period
– OMB must review final standard
– Dr. Michaels expects final rule by 2016
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Foundry Industry Response
• Strongest industry response, according to
OSHA
• Foundry members provided valuable data,
and resources
• AFS made detailed review of OSHA
assessments and analysis of costs
• AFS economic and technological
arguments remain largely uncontradicted
in record
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
End Game Options?
1. OSHA listens
– Maintains current PEL with ancillary provisions
– Foundries still have challenge to meet current PEL (especially with change in size criteria)
– Option: redefine PEL as being mean exposure
2. OSHA ignores
– Tries to meet political target
– OMB approval hurdle
– Court challenge (overturn or force revision)
3. Compromise
– 50 µg/m3 PEL but allow respirators below 100
– More time to come into compliance
44
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
• Budget – drop funding
• Pressure to conduct updated
SBREFA review
• Ask National Academy of Sciences to
study measurement and health risk
issues
• Oversight of flawed process
Is There Room for
Congressional Action?
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Getting Ready in the
Meantime• Exposure assessment
– Variability and confidence
– Source evaluation
• Operational review – size up the
impact and need for court challenge
– Compressed air
– Sweeping
– Regulated area
Sand Casting ConferenceOctober 22, 2014 – Indianapolis, IN USA
Questions ??
Thomas Slavin
Slavin OSH Group, LLC
312-863-2350
708-937-3244
More Information: