Michael P. Wilson, Ph.D, MPH
Assistant Research Scientist
Center for Occupational and Environmental Health
School of Public Health
140 Earl Warren Hall #7360
University of California
Berkeley, CA 94720-7360
Ph: +1-510-642-5703
Fax: +1-510-642-5815
mpwilson@berkeley.edu
http://coeh.berkeley.edu
September 13, 2005
Councilmember Linda Maio
City of Berkeley
lmaio@ci.berkeley.ca.us
RE: Pacific Steel Casting
Dear Councilmember Maio,
This letter contains my recommendations regarding a strategic plan for addressing emissions from
the Pacific Steel Casting plant in west Berkeley. My recommendations are based broadly on a
pollution prevention, or Toxics Use Reduction (TUR) approach, as compared to a Health Risk
Assessment (HRA) approach. The overarching objectives of the plan are as follows:
1) Address emissions that are producing noxious odors in west and north Berkeley;
2) Address emissions of toxic metals and volatile organic compounds (VOCs) that may or may
not be related to the odors;
3) Maintain and improve worker health and safety and ensure that efforts to address
environmental emissions do not jeopardize worker health and safety;
4) Support the economic viability of the plant in its present location;
5) Support the plant collective bargaining agreement and employee representation.
My comments regarding the proposed HRA are attached for your information (Appendix A).
Background
Toxics Use Reduction (TUR) is a fundamental form of pollution prevention that focuses on
industrial activities to promote safer and cleaner production and to enhance economic viability. As
compared to risk assessment, which attempts to measure health impact and generally precludes
stakeholder participation, TUR is oriented toward identifying and solving health and environmental
problems at their source through various engineering, planning and other strategies. These include
the following:




toxic chemical substitution
production process modification
finished product reformulation
production modernization


improvements in operations and maintenance
in-process recycling of production materials
In some cases, the only available solution for a particular process may be to improve control
technology; knowing this, however, requires a thorough assessment of current practices, available
technologies and best practices in the industry.
Massachusetts implemented a statewide TUR approach in 1989 with the passage of the Toxics Use
Reduction Act (TURA). The Act defined TUR as "in-plant changes in production processes or raw
materials that reduce, avoid, or eliminate the use of toxic or hazardous substances or generation of
hazardous by-product per unit of product . . . without shifting risks between workers, consumers or
parts of the environment." (http://www.turi.org/) (http://www.sustainableproduction.org/). A
2000 study based on 35 published case studies of TUR among Massachusetts companies, together
with interviews with key personnel, reported that between 1990 and 1997, Massachusetts companies
decreased their use of toxic chemicals by 24 percent and decreased their volume of toxic by-product
by 41 percent, indexed to production. In almost 50 percent of the cases analyzed, improved worker
health and safety was cited as a benefit of the toxic use reduction projects.1 Solvents were eliminated
or reduced in 63 percent of the cases. 46 percent of the companies profiled introduced water-based
chemicals in the place of more volatile ones, and acids and caustics were reduced or eliminated in
about 20 percent of the cases. Massachusetts outperformed virtually every other manufacturing
state in the country on releases of substances under the Toxics Release Inventory (TRI).2
Pacific Steel Casting
Pacific Steel Casting substantially reduced air releases during the period 1988 to 2002, as reported
under the TRI (Table I). Releases have increased somewhat since 1999.
Table I. Toxics Release Inventory data summary for Pacific Steel Casting, 1988 – 2002.
Year
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Total
Total
Air
Water
Land Underground Environmental Total Off-Site ProductionReleases Releases Releases
Injection
Releases
Transfers
Related Waste
203,774
0
0
0
203,774
52,120
NA
169,727
0
0
0
169,727
133,246
NA
158,588
0
0
0
158,588
53,601
NA
130,764
0
0
0
130,764
21,100
596,000
112,107
0
0
0
112,107
29,859
491,100
72,558
0
0
0
72,558
32,940
444,030
86,504
0
0
0
86,504
28,240
116,350
6,510
0
0
0
6,510
33,850
557,560
1,005
0
0
0
1,005
30,150
497,960
1,005
0
0
0
1,005
105,300
711,600
270
0
0
0
270
66,550
603,000
10
0
0
0
10
25,950
319,610
270
0
0
0
270
33,750
418,470
255
0
0
0
255
68,800
397,550
515
0
0
0
515
2,800
366,890
2
The plant is also facing increased public concern over noxious odors in west and north Berkeley.
Public pressure related to the odors, and their perceived risks, is likely to increase in the future,
irrespective of the outcomes of the proposed HRA. A TUR approach is therefore appropriate in
this case. The PSC case is somewhat similar to the Dow Midland case in Michigan, which partnered
with the Natural Resources Defense Council in implementing a TUR strategy that reduced emissions
substantially while also reducing costs to Dow. The TUR approach in this case benefited the
Midland community, plant employees and the economic viability of the plant itself.3 With the
correct set of players and motivations, PSC in Berkeley could also be a success story.
Recommendations
1. Convene a meeting of stakeholders to consider a TUR approach, as compared to the Health Risk
Assessment approach.
Stakeholders would include PSC staff, union representatives, the West Berkeley Alliance, City staff,
EPA Region IX, and other members of the public. I recommend that the City propose to terminate
the Health Risk Assessment and that as a condition of doing so PSC redirect funds previously
committed to the HRA to a professional toxics use reduction evaluation. Leif Magnuson, Pollution
Prevention Coordinator for U.S. EPA Region IX, indicated to me that the Region would be
interested in participating in this process.4 A cost-sharing approach might be arranged between
Region IX and PSC, if appropriate etc. Region IX pollution prevention staff have particular
expertise in pollution prevention site assessment and treatment technologies and work regularly with
reputable contractors in this arena (http://www.epa.gov/region09/waste/p2/index.html). A
summary TUR evaluation conducted by a contractor for Region IX is attached (Appendix B).
2. Draft a TUR proposal that includes the following steps:
a.
b.
c.
d.
e.
A TUR evaluation of PSC by a contractor selected by the stakeholders;
An evaluation of the contractor’s report by stakeholders;
Drafting of a TUR Plan by PSC to address concerns identified in the report;
Establishing a calendar for implementing the TUR Plan;
Consideration of mechanisms by the City to ensure reasonable progress in adhering
to the TUR plan calendar;
f. Evaluation of the TUR Plan, calendar and mechanisms by stakeholders;
g. Submission of semi-annual progress reports to the City in implementing the TUR
Plan.
3. Draft a Berkeley TUR Plan procedure for PSC.
A TUR plan is produced by, and maintained at, the facility. It is submitted to the City and is made
publicly available. The plan document should include a management policy statement regarding
TUR, the scope of the plan, employee notification and involvement procedures, characterization of
the process that involves the use of chemicals or substances of concern, cost of these chemicals or
substances, options identification for alternatives to these chemicals or substances, options evaluation
and options implementation.
3
The plan should establish one- two- and five-year goals for reduction of each chemical or substance
of concern used or processed at PSC. The plan should be updated every year, including an
assessment of the implementation schedule and additional TUR commitments.
In addition to the plan document, PSC should submit data to the City -- normalized by production
volume -- about historic and projected changes in facility-wide use and individual process byproduct
generation for each chemical or substance of concern. Two indices can be used to characterize byproduct reduction goals, namely the By-product Reduction Index (BRI) and the Emission Reduction
Index (ERI). The BRI refers to the non-product output of a process and is therefore a measure of
materials efficiency. The ERI refers to pollution released from the facility. I would be happy to
assist the City in implementing a TUR strategy for the PSC case.
Please let me know if you have questions or would like to discuss this in greater detail.
Sincerely,
Michael P. Wilson, Ph.D, MPH
Assistance Research Scientist
References
1.
2.
3.
4.
Roelofs , M-E, Ellenbecker,. Pollution Prevention and the Work Environment: The Massachusetts
Experience. Applied Occupational and Environmental Hygiene 15(11):843-850(2000).
O'Rourke, D, Lee E. Mandatory planning for environmental innovation: Evaluating regulatory
mechanisms for toxics use reduction. Journal of Environmental Planning and Management 47
(2):181 - 198(2004).
Natural Resources Defense Council. Preventing Industrial Pollution at its Source: A Final Report of
the Michigan Source Reduction Initiative http://www.nrdc.org/water/pollution/msri/msriinx.asp
(accessed Sept 13, 2005). 2000-2002.
Leif Magnuson. Pollution Prevention Coordinator, U.S. EPA Region IX. Phone: (415) 972-3286.
Fax: (415) 947-3530. magnuson.leif@epa.gov. (Personal communication September 13, 2005).
4
Appendix A
Comments on the proposed Health Risk Assessment submitted to the Bay Area Air Quality
Management District September 13, 2005.
5
Michael P. Wilson, Ph.D, MPH
Assistant Research Scientist
Center for Occupational and Environmental
Health
School of Public Health
140 Earl Warren Hall #7360
University of California
Berkeley, CA 94720-7360
Ph: +1-510-642-5703
Fax: +1-510-642-5815
mpwilson@berkeley.edu
http://coeh.berkeley.edu
September 12, 2005
Mr. Scott Lutz
Bay Area Air Quality Management District
939 Ellis Street
San Francisco, CA 94109
Dear Mr. Lutz,
This letter represents my comment regarding the August 26, 2005 Revised Health Risk Assessment
Protocol for Pacific Steel Casting (PSC), Berkeley, California, proposed by Environmental Resources
Management (ERM). My comments include general concerns about the use of health risk
assessment (HRA) as a tool for policymaking in the PSC case, followed by comments on the
technical content of the HRA itself.
General concerns
While HRA has a clear role under certain conditions, it is my view that in the PSC case it is not the
best choice of policy tools. The reasons for this are as follows.
First, the HRA will likely have three possible interpretations pertaining to the health risk to residents
and nearby workers of PSC emissions, namely (1) the emissions pose negligible health risk; (2) they
pose a very slight health risk to sensitive populations; or (3) under certain conditions, they probably
pose a health risk, especially to sensitive populations. The difficulty, of course, is in deciding what in
fact these results mean. It is unlikely that any one of these findings will address ongoing public
concern over noxious odors in west and northwest Berkeley, and it is possible that the results of the
HRA could result in an increase in tensions between stakeholders and in calls for additional studies.
Second, as you are well aware, the uncertainty inherent in a HRA subjects the process to
extraordinary variability, which often leads to at least one party (legitimately) challenging the results.
The assumptions needed to produce inputs to the HRA, for example, are easily challenged on
numerous counts. Given that:
Health risk = function (hazard * exposure)
6
and exposure is a function of the nature of emission sources, paths and receivers, and hazard is a
function of chemical attributes and their myriad health endpoints, uncertainty enters the HRA
process in at least the following areas:
Exposure:
Source
Emission type (heavy metals, volatile organics, particulate matter etc)
Emission rate over time and under varying conditions of operation
Emission peaks and excursions
Effectiveness of control technologies
Changes over time due to aging of equipment and maintenance
Path
Emissions that combine intended, fugitive and accidental releases
Plume changes from changes in wind direction and velocity
Respiratory dose versus hand-to-mouth dose in the case of metals
Receiver
The frequency of contact with emissions
The exposure concentration (intensity) in the breathing zone
The duration of contact with emissions
The relative sensitivity of the receiver as compared to default values
The cumulative effect of multiple environmental health stressors
The normal distribution of disease among a population
Hazard:
The toxicological properties of the substances in question
The relevant health endpoints of the substances, alone and in combination:
Cancer
Respiratory effects, acute and chronic
Reproductive
Developmental
Immune
Neurological (central and peripheral)
Bioaccumulation
Combined effects from chemical mixtures
Third, even a carefully constructed and thorough HRA is not able to produce a robust assessment of
the actual health impact of plant emissions on public health. The emissions represent one of many
environmental health stressors among the population living and working in the west/north Berkeley
area. Emissions from the 880 freeway, consumer products, occupational exposures etc. all
contribute to the total environmental health impact. The relative contribution of PSC emissions to
7
this spectrum of stressors is unknown and will not be answered by the HRA. The health risk of
PSC emissions will be represented in the HRA in isolation from these other stressors. For this
reason, the HRA can be misleading if it concludes that the health effects of PSC emissions are
“negligible”, for example. While they might contribute an extremely small share of the total set of
environmental health stressors, in the overall picture these could be significant. The (necessarily)
narrow focus of the HRA weakens its usefulness for purposes of public health policy and decisionmaking in this case.
Overall, it is extremely difficult for the HRA to capture the nature of exposures (which are
lognormally distributed) and the nature of health effects (which are normally distributed). It will
likely produce a single “risk estimate” that reflects various safety factors to account for sensitive
populations etc. but which is not really a true measure of the public health burden of PSC. Even if
the HRA produces a result suggesting that PSC presents “no health risk”, the west/north Berkeley
community will continue to experience, and object to, noxious odors, which are the driving issue in
the PSC case. The HRA is therefore unlikely to satisfy the public’s legitimate concerns.
Alternatives assessment
I recommend that BAAQMD, the City of Berkeley, the union, PSC management and community
groups explore an alternatives assessment as compared to a risk assessment. The point of this
approach is to solve the underlying problem at hand, as compared to measuring the risk of that
problem. In general, an alternatives assessment would employ toxics use reduction strategies and
planning, based on evidence in other regions, to address emissions of toxic materials from PSC and
their associated odors. The state of Massachusetts adopted this approach in 1989 with the Toxics Use
Reduction Act (TURA) and has achieved significant improvements in emissions of toxic materials
since that time without negative impacts on production; in fact, Massachusetts businesses have
achieved significant savings by maintaining thorough accounting of their use of toxic materials and
implementing reductions strategies. For your information, I have attached a paper by O’Rourke and
Lee that evaluates the TURA. (Appendix A).
Implementing a toxics use reduction (alternatives assessment) strategy for PSC has the potential to
meet the needs of all stakeholders in this matter. I would be happy to work with the parties in
working toward this objective.
Thank you for giving me the opportunity to review this document.
Sincerely,
Michael P. Wilson, Ph.D, MPH
Assistant Research Scientist
8
Comments on the HRA
The data for estimating source emissions is contained in the supplemental material entitled Emission
Inventory Calculations. Data are presented for Plant #1 ( BAAQMD Plant 187), Plant #2 (BAAQMD
Plant 703) and Plant #3 (BAAQMD Plant 1603). In general, explanations regarding how data were
derived for each column are needed for understanding this spreadsheet. For example, I would like
to know how the “uncontrolled emission factor” data were derived, as these values seem to be the
basis for the emission factors described in the four columns on the right side of the spreadsheet.
Were these values estimated by ERM or were they provided by PSC? Emissions estimates form the
basis of the risk assessment, so transparency is important here.
My specific concerns regarding these data are as follows:
Plant #1.
Page 1. What is the pollutant to be measured at S2, S3 and S4?
Page 1. What is the diameter of PM that will be measured? I would be particularly
concerned about respirable particles <1 micron in diameter. Will ERM measure PM of <10
microns and <1 micron, or will ERM estimate a particle size distribution?
Page 1. On what basis does ERM assume that EAF Source Test Results for Plant #2 are
representative of the operations at S1 in Plant #1?
Page 1. On what basis does ERM assume that P703, S37 (grinder, Plant #2, tested in 1989 )
is representative of emissions for S12 (cleaning and grinding, Plant #1) today? Are these
operations similar? Have these operations changed since 1989? Documentation is needed
to support this assumption.
Page 1. What is the BAAQMD PSC Inventory and speciation?
Page 1. On what basis does ERM assume that P703, S32 (Rotoblast, Plant #2, tested in
1989 ) is representative of emissions for S15 (Pangborn table blast, Plant #1) today? Are
these operations similar? Have these operations changed since 1989? Documentation is
needed to support this assumption.
Page 2. At S-18, why has ERM elected not to sample emissions for hexane (1.38 mg/s), for
example, but is estimating emissions based on AP-42? This seems to be a relatively
important emission source. How representative is the AP-42 estimate of conditions at PSC?
Page 2. Again, ERM makes the assumption that conditions in Plant #2 in 1989 are similar
to conditions in Plant #1 for S15, S16 and S17 today. What is the basis for this assumption?
9
Plant #2
Page 1. Again, the Basis for Emissions data for formaldehyde at S7, S8, S9, S10, S11, S12 and
S28 are dated November 1989. Is ERM confident that production levels for these processes
are unchanged from that time period? If so, what evidence supports this assumption? In
the absence of such evidence, emissions testing for these processes should be conducted to
reflect current conditions.
Page 1. The Basis for Emissions data for PM and for volatile organics (with the exception of
phenol) are based on estimates from AP-42. Is ERM confident that the production
processes at PSC are similar to those provided by AP-42? If so, ERM should provide
documentation to this effect. Does PM refer to <10 microns or <1 micron?
Page 1. ERM proposes to use P6 Stack to estimate the Basis for Emissions for S13 – S20,
S22 and 23. On what basis does ERM assume that these processes and emissions are
similar? What documentation would support this assumption?
Page 1. ERM proposes a source test for S21. What is the pollutant to be tested at this site?
Page 2. The Basis for Emissions is November 1989 for all processes. Again, is ERM confident
that production levels for these processes are unchanged from that time period? If so, what
evidence supports this assumption?
Page 2. What is pollutant to be tested at S27, S29, S31 and S31?
Page 3. Again, the Basis for Emissions is November 1989 for all processes. Is ERM confident
that production levels for these processes are unchanged from that time period? While the
emission values are extremely small individually, in aggregate they could be worth evaluating.
In addition, because the health risk assessment is rooted in emission estimates, the accuracy
of these values is important.
Page 4. What is the pollutant to be measured at P10?
Page 4. At S26 and 32000, the data indicate 0 emissions, no external stack. Are there
industrial hygiene controls (such as local exhaust ventilation) in place for these volatile
organics, or are the emissions released freely inside the plant? Are the IH controls released
outside the plant? Is ERM confident that AP-42 correctly reflects conditions for these
processes at PSC? If so, what evidence is there to support this assumption?
Page 4. Is ERM confident that AP-42 represents a reasonable basis for emissions for S26
and 32000 at PSC? Explanation is needed here.
10
Plant #3.
Page 1. What is the pollutant to be measured at P1 and P3?
Page 1. The Basis for Emissions for P2 is noted by ERM as P703, S32, tested in November
1989. Is ERM confident that P2 is similar to P703, S32 in terms of emissions? If so, what is
the basis of this assumption? Is ERM confident that conditions have not changed for these
operations since 1989? If so, what is the basis of this assumption?
Other concerns.
Section 3.0 Models/modeling assumptions.
What are the factors that will be used to adjust default values to account for children and other
sensitive populations (page 4)? For example, the inhaled dose will change significantly in liters/kgday for children as compared to the BAAQMD-recommended interim breathing rate of 302
liters/kg-day.
Does a particle deposition velocity of 0.02 meters per second capture respirable particles, or does
this only capture particles that impact with the posterior oropharynx (page 4)?
Section 4.0 Land use coefficients and meteorological data
Why is a rural dispersion coefficient used here? Does this lead to a more or less conservative
estimate?
Section 5.0 Deposition
Does deposition refer to exposure pathways related to deposition of emission materials on surfaces
in homes and workplaces? If so, why is deposition excluded from the study, given that many heavy
metals are listed in the emission inventory?
Section 7.0 Risk characterization
PSC has self-reported emissions of phenol (255 pounds), manganese (250 pounds), nickel (5
pounds) and chromium (5 pounds) under the Toxics Release Inventory (Appendix C). All of these
are reproductive toxicants; phenol and nickel are developmental toxicant, as are many of the metals
listed in the inventory. Will ERM evaluate reproductive and development health effects?
In what ways will ERM describe the limitations of the results of the HRA? In what will ERM
describe the areas of uncertainty in areas of emissions, path and receiver?
11
Finally, if PSC is funding ERM to conduct the HRA, it is worth asking what portion of ERM’s
income is dependent on private clients such as PSC and the portion of HRAs performed by ERM
for these clients that have resulted in reports of significant (and insignificant) health risks. A
concern in these kinds of relationships is whether, and how often, the consultant will produce a risk
assessment that points to significant health risks that are produced by the paying client.
* * * * * * * *
12
Appendix B
Cover letter, U.S. EPA Region IX pollution prevention office summarizing the findings from
the pollution prevention (P2) assessment conducted by Tetra Tech EM, Inc.
13
Date
(Facility)
Subject: Pollution Prevention Opportunities for (Facility) Plating, Inc.
Dear (Facility Contact):
This letter summarizes the findings from the pollution prevention (P2) assessment conducted by Tetra Tech
EM Inc. (Tetra Tech) for (Facility) Plating, Inc. Tetra Tech is providing technical assistance for this project
to the U.S. Environmental Protection Agency (EPA) Region 9. The P2 assessment was based on facility and
process information provided by (Facility) in the mini-assessment questionnaire and worksheets, our April
visit to your facility, and additional process information provided to us after our facility visit.
Enclosure 1 describes four proposed P2 projects Tetra Tech identified that can be implemented at (Facility)
to reduce waste generation and improve process efficiency. Each P2 project is presented in a standard one
page (two-sided) format that describes the P2 technique or technology proposed, implementation issues,
expected impacts, estimated costs, assumptions, and estimated overall results. The four P2 projects identified
and their estimated cost, savings, and payback periods are summarized in the following table.
Proposed
Project
No.
1
2
3
4
Proposed Project
Reduction in Water Use Through Use of Flow
Restrictors and Solenoid Valves with Timers
Deoxidizer Bath Life Extension
Regenerate Electroless Nickel Bath Using
Electrodialysis
Using Ultrasonics to Improve Cleaning and
Reduce Rejects
Savings
(per
year)
$19,200
$2,920
$20,700
$14,700
Estimated
Cost
(capital;
O&M)
$1,620;
$0/yr
$0;
$75/yr
$23,890;
$3,210/yr
$27,000;
$1,300/yr
Payback
(years)
< 0.1
< 0.1
1.4
2.0
Enclosure 2 contains additional information relevant to some of the proposed P2 projects that may assist
(Facility) with implementation decisions. Enclosure 3 includes a graphical summary of facility and process
information that supported our efforts to identify P2 opportunities for (Facility). These summary graphs and
tables were generated based on information provided by (Facility).
In addition to the four projects proposed for implementation at (Facility), Tetra Tech also considered the use
of a sludge dryer. (Facility) currently generates 119,664 pounds (59.8 tons) of sludge annually and pays
$21,800 to dispose of this sludge. The sludge dryer Tetra Tech considered for implementation at (Facility)
was a unit that attaches to the filter press and is operated by either steam heat or electricity. For the purposes
of evaluation, Tetra Tech assumed the smallest unit available (5 ft3) would be sufficient. The capital cost of
this unit is $22,900 and operation and maintenance costs would be approximately $625 per year for weekly
greasing of the fittings. The cost to operate the unit by electricity or steam heat is negligible in comparison to
the current facility operating costs. It was assumed that the unit could reduce the sludge volume and sludge
disposal costs by 50 percent or $10,900 per year. The payback period for the sludge disposal unit would be
2.2 years. However, Tetra Tech identified a waste disposal company, Phillips Service Company, that could
dispose of the sludge generated at (Facility) for half the cost, based on recent sludge analyses, (Facility) is
14
currently paying. This waste disposal alternative results in an immediate reduction in costs for sludge disposal
and no capital or operation and maintenance costs. If (Facility) were to consider a sludge dryer after
switching sludge disposal contractors, for a sludge dryer capital cost of $22,900, operation and maintenance
costs of $625 per year, and a sludge disposal savings of $5,450 per year, the payback period would be 4.7
years. For this reason, Tetra Tech recommends that (Facility) first consider the alternative sludge disposal
contractor.
(Facility) should review and discuss the proposed P2 projects. Tetra Tech will contact you next week to
schedule a teleconference meeting. The purpose of this meeting will be to (1) discuss the proposed projects,
(2) identify the projects that (Facility) is interested in implementing, and (3) obtain feedback on projects that
(Facility) chooses not to pursue. If you have any immediate questions, please call me at (415) 744-2153 or
Tom Adkisson of Tetra Tech at (415) 222-8320.
Sincerely,
Leif Magnuson
EPA Region 9 Project Manager
cc:
Eve Levin, EPA
Tom Adkisson, Tetra Tech
Patrick Wooliever, Tetra Tech
Peter Ko, Tetra Tech
File
Enclosures (3) (Available on request)
15