August 18, 2006

FINAL DISPOSITION OF PEER REVIEW COMMENTS FOR 

CHARACTERIZATION OF DUST LEAD LEVELS AFTER RENOVATION, REPAIR, AND
PAINTING ACTIVITIES 

GENERAL COMMENTS FROM PRIVATE AND NON-EPA FEDERAL REVIEWERS

Prepared By

BATTELLE

505 King Avenue

Columbus, Ohio  43201

EPA Contract No. EP-W-04-021

Work Assignment 2-10

Prepared For

Sineta Wooten, Project Officer

John Schwemberger, Work Assignment Manager

Samuel Brown, Deputy Work Assignment Manager

Program Assessment and Outreach Branch

National Program Chemicals Division

Office of Pollution Prevention and Toxics

U.S. Environmental Protection Agency

1200 Pennsylvania Avenue NW (7404T)

Washington, D.C.  20460

GENERAL COMMENTS ON RRP QAPP FROM PRIVATE AND NON-EPA FEDERAL REVIEWERS

TABLE OF CONTENTS

  TOC \o "1-3" \h \z \u    HYPERLINK \l "_Toc139436771"  TABLE OF
CONTENTS	  PAGEREF _Toc139436771 \h  1  

  HYPERLINK \l "_Toc139436772"  Comments from Reviewer P1	  PAGEREF
_Toc139436772 \h  2  

  HYPERLINK \l "_Toc139436773"  Comments from Reviewer P2	  PAGEREF
_Toc139436773 \h  12  

  HYPERLINK \l "_Toc139436774"  Comments from Reviewer P4	  PAGEREF
_Toc139436774 \h  18  

  HYPERLINK \l "_Toc139436775"  Comments from Reviewer P6	  PAGEREF
_Toc139436775 \h  19  

  HYPERLINK \l "_Toc139436776"  Comments from Reviewer P7	  PAGEREF
_Toc139436776 \h  20  

  HYPERLINK \l "_Toc139436777"  Comments from Reviewer P8	  PAGEREF
_Toc139436777 \h  38  

  HYPERLINK \l "_Toc139436778"  Comments from Reviewer FR1	  PAGEREF
_Toc139436778 \h  45  

  HYPERLINK \l "_Toc139436779"  Comments from Reviewer FR2	  PAGEREF
_Toc139436779 \h  48  

  HYPERLINK \l "_Toc139436780"  Comments from Reviewer FR3	  PAGEREF
_Toc139436780 \h  49  

  HYPERLINK \l "_Toc139436781"  Comments from Reviewer FR4	  PAGEREF
_Toc139436781 \h  50  

  HYPERLINK \l "_Toc139436782"  Comments from Reviewer FR5	  PAGEREF
_Toc139436782 \h  51  

 Comments from Reviewer P1

Comments on Risk Assessment: Exposure Assessment Description 

Paragraph 2

I know of no way to assess dermal exposure to leaded dust except to say
that it is near zero body burden.  From what resource(s) will contrary
information be obtained as input into the exposure assessment?

Paragraph 3

a. The study as proposed does not sample dust in ducts, nor in/on soft
and/or removable surfaces (e.g., furniture), and lead in food.  From
what resource(s) will such information be obtained as input into the
exposure assessment?

b. Collecting only 2 floor dust samples per stage seems light for
determination of dust loading a multiple distances from the job.  I do
not believe that the study proposes to take enough samples from
available useful sites and may miss the opportunity to significantly add
to the knowledge of what is taking place during renovation activities. 
This study presents a grand opportunity to study actual renovation
activities and as such should be milked for all possible data.  For
example, I would expect that job sites involve and the rooms containing
job sites will be at least 36 ft2 (6 x 6, or so).  This would allow the
opportunity to collect 36 floor samples; samples close to the job
activity and samples at some distance from the job activity.  Given 4
sampling opportunities per stage and allowing for no duplication of
sampling site per sampling opportunity, why not randomly collect samples
from half of the available floor space (4 ft2 of 8 ft2 available) per
sampling opportunity.

Response:

	The study is not being designed to specifically address the routes of
human exposure.  Rather, the study, as designed, concentrates on
providing information related to settled dust-lead loadings and air-lead
concentrations attributable to RRP activities and associated containment
and clean-up efforts.  Routes of exposure resulting from leaded dust
(such as the amount that gets integrated into food) are outside the
scope of this study.

It is the goal of the study to use vacant properties, whenever possible.
 All furnishings that happen to be in study-impacted areas will be
removed and placed in an area of the house that will be securely sealed
off.  In addition, all air vents in the study-impacted areas will be
securely covered with plastic sheeting, eliminating the need for air
vent dust wipe samples.    

We agree that more samples need to be taken per job activity and stage. 
As such, more samples are now included in the study protocol.

Comments to Quality Assurance Project Plan 

Comments made refer to the section cited.

2.2 Suggest use ASTM methods:

E1729-05 for collection of paint samples;

E2119-00 for QC when using an XRF to test paint;

E1792-03 for specification of wipe materials;

E1728-03 for wipe sampling of settled dust;

D5337-04 for pump calibration, and D4185-96(2001) or especially D6785-02
for airborne dust sampling; and,

E1727-05 for collection of soil samples.

2.3 Suggest use ASTM E2239-04 for data handling.

2.4 Suggest use ASTM methods for conduct of analyses.

2.6 Most strongly recommend that only labs currently recognized by EPA
under the NLLAP should be used for analysis services (40 CFR 745.227(f))
even though these experiments are not lead-based paint activities as
defined in Title X.

2.7 

Why use ICP-MS for paint?  If the intent is to be able to determine what
paint contributed what to the dust loading, then use it.  Otherwise,
it’s too expensive and gives no new information versus use of FAA
and/or ICP-AES.  Even use of GFAA is cheaper.  In general, more lead in
the sample means cheaper and easier analytical quantification (from very
cheap to very expensive: from FAA or ICP-AES, to GFAA, to ICP-MS).  A
good compromise between assuring a bunch of lead in the sample and
conducting jobs as in the real world is to analytically screen each
sample using FAA or ICP-AES (from $6 to $8 per sample).  If they are not
quantifiable by FAA, proceed to GFAA (may be as much as $35 per sample).
 If not quantifiable by GFAA, proceed progressively to ICP-MS analyses
(may be as much as $140 per sample).

Most strongly recommend that only labs currently recognized by EPA under
the NLLAP should be used for analysis services (40 CFR 745.227(f)) even
though these experiments are not lead-based paint activities as defined
in Title X.

2.8 See ASTM E1792-03 for specification of wipe materials.

Response:

The sample collection protocols will follow ASTM E1727-05 for soil
samples, ASTM E1728-03 for dust wipe samples and ASTM E1729-05 for paint
chip samples.  The laboratory selected for sample analysis is Schneider
Laboratories, Inc., a NLLAP- recognized lab.  Sample analysis will be
done using Atomic Absorption methods (NIOSH 7082 for air filters and EPA
Method 7420 for paint, soil, and dust wipes).  The lab will be providing
the dust wipe materials for the study in compliance with the current
standards.

Comments to Draft Study Design 

Comments made refer to the section cited.

2.1.5.1

See ASTM E1729-05 for collection of paint samples and ASTM E2119-00 for
QC when using an XRF to test paint.

Why use ICP-MS for paint?  If the intent is to be able to determine what
paint contributed what to the dust loading, then use it.  Otherwise,
it’s too expensive and gives no new information versus use of FAA
and/or ICP-AES.  Even use of GFAA is cheaper.  In general, more lead in
the sample means cheaper and easier analytical quantification (from very
cheap to very expensive: from FAA or ICP-AES, to GFAA, to ICP-MS).  A
good compromise between assuring a bunch of lead in the sample and
conducting jobs as in the real world is to analytically screen each
sample using FAA or ICP-AES (from $6 to $8 per sample).  If they are not
quantifiable by FAA, proceed to GFAA (may be as much as $35 per sample).
 If not quantifiable by GFAA, proceed progressively to ICP-MS analyses
(may be as much as $140 per sample).  In any case, only labs currently
recognized by EPA under the NLLAP should be used for analysis services
(40 CFR 745.227(f)) even though these experiments are not lead-based
paint activities as defined in Title X.

Response:

The sample collection protocols will follow ASTM E1727-05 for soil
samples, ASTM E1728-03 for dust wipe samples and ASTM E1729-05 for paint
chip samples.  The laboratory selected for sample analysis is Schneider
Laboratories, Inc., a NLLAP-recognized lab.  Sample analysis will be
done using Atomic Absorption methods (NIOSH 7082 for air filters and EPA
Method 7420 for paint, soil, and dust wipes).  

2.1.5.2

See ASTM E2255-04 and E2271-05a for conduct of clearance.

Why not sample troughs?  Suggest sample window troughs, especially
following conduct of exterior work.  The shortest route inside is
through the windows.

Response:

	The pre- and post-experiment cleaning will be performed by a firm
certified to clean after lead abatement activities and the clearance
samples will be collected by certified risk assessors or clearance
technicians, both of whom will follow all appropriate standards.

2.1.5.3

Note that area sampling is usually used to define the OSHA work area and
should be collected when work is being done (sampling during no-work
times bias exposure results low).  Area sampling and personal sampling
should be conducted using ASTM D5337-04 and D4522-97(2003) as
simultaneously as possible to facilitate comparison.

Why not sample troughs?  Suggest sample window troughs, especially
following conduct of exterior work.  The shortest route inside is
through the windows.

2.1.5.4 Note that area sampling is usually used to define the OSHA work
area and should be collected when work is being done (sampling during
no-work times bias exposure results low).  Area sampling and personal
sampling should be conducted using ASTM D5337-04 and D4522-97(2003) as
simultaneously as possible to facilitate comparison.

2.1.5.5

Shut down the area samplers when the work ends.

Randomly-selected, multiple samples of sites within replicate work areas
could/would quantitatively determine the dust loading as a result of a
replicate job phase.  Sampling site-selection bias is controlled with
random selection of sites within the work area to be sampled.  Those
sites previously sampled would be eliminated from the pool of possible
sampling sites during each sampling opportunity in a stage.  Collection
of multiple samples per opportunity increases the chance that the actual
leaded dust loading caused by the job will be measured.  Multiple
iterations of the job allows for comparison of sampling results.  See
the method for random selection of sampling sites in ASTM E2271-05a.

Restart the area samplers when work begins.

Response:

	All sampling locations will be randomly selected and identified on a
site-specific sampling plan for all experiments.  The indoor and
personal air samplers will be turned off at the same time at each stage
of the experiment for consistency purposes.

2.1.5.6

Strongly suggest conduct all cleaning activities in disposable personal
protective clothing (PPC) and respirators.

Strongly suggest conduct personal air sampling (i.e., man exposure
monitoring) of at least all experimental activities involving dust
generation and cleanup using ASTM D5337-04 and D4522-97(2003).

Note that area sampling is usually used to define the OSHA work area and
should be collected when work is being done (sampling during no-work
times bias exposure results low).  Area sampling and personal sampling
should be conducted using ASTM D5337-04 and D4522-97(2003) as
simultaneously as possible to facilitate comparison.

Response:

	All study personnel will don personal protective clothing and
respirators throughout each stage of the experiment.  Personal air
samples will be collected for both interior and exterior experiments at
each sampling stage to assess worker exposure.  

2.1.5.7 

Strongly suggest conduct all cleaning activities in disposable personal
protective clothing (PPC) and respirators.

Note that area sampling is usually used to define the OSHA work area and
should be collected when work is being done (sampling during no-work
times bias exposure results low).  Area sampling and personal sampling
should be conducted using ASTM D5337-04 and D4522-97(2003) as
simultaneously as possible to facilitate comparison.

Response:

	All study personnel will don personal protective clothing and
respirators throughout each stage of the experiment.  

See the method for random selection of sampling sites in ASTM E2271-05a.

Response:

	All sampling locations will be randomly selected and identified on the
site-specific sampling plan for all experiments.  

2.1.5.10

Note that area sampling is usually used to define the OSHA work area and
should be collected when work is being done (sampling during no-work
times bias exposure results low).  Area sampling and personal sampling
should be conducted using ASTM D5337-04 and D4522-97(2003) as
simultaneously as possible to facilitate comparison.

See the method for random selection of sampling sites in ASTM E2271-05a.

Response:

	Personal air samples will be collected for both interior and exterior
experiments at each sampling stage to assess worker exposure.  

All sampling locations will be randomly selected and identified on the
site-specific sampling plan for all experiments.  

2.1.5.12

See the method for random selection of sampling sites in ASTM E2271-05a.

Clearance requires visual examination and sampling of floors, (interior)
window sills, and window troughs.

HUD at 24 CFR 35.1330 requires clearance to include examination of
interior floors and sills, and exterior surfaces (troughs) after even
other-than-abatement activities.  Further, HUD requires clearance of the
exterior and interior following exterior jobs.  I believe it would be
quite useful to know what dust loadings are caused by each job (interior
and exterior).

Response:

	All sampling locations will be randomly selected and identified on the
site-specific sampling plan for all experiments.  

The pre- and post-experiment cleaning will be performed by a firm
certified to clean after lead abatement activities and the clearance
samples will be collected by certified risk assessors or clearance
technicians, both of whom will follow all appropriate standards.

2.2.2

See ASTM E1728-03 for wipe sampling of settled dust.

See ASTM E1792-03 for specifications of wipe materials.

Most strongly suggest against use of templates on sills.

2.2.3 Strongly suggest perform personal air sampling (man monitoring)
and area sampling at 2 L/min since that’s what everyone else uses. 
See ASTM D5337-04 for pump calibration, and D4185-96(2001) or especially
D6785-02 for airborne dust sampling.

2.2.4

See ASTM E1727-05 for collection of soil samples.

Will soil samples be collected both before and after conduct of an
experiment at a site?  If so, how are the before and after soil sampling
results to be evaluated?  Some states (e.g., Virginia) call for use of a
paired Student T-test for pre- and post-activity soil samples.

Response:

Current ASTM standards for sampling collection will be followed.  They
include ASTM E1727-05 for soil samples, ASTM E1728-03 for dust wipe
samples, and ASTM E1729-05 for paint chip samples.

The air samplers will both be set at 3L/min, following the direction of
the field operations coordinator. 

2.2.6 As presented, this paragraph is incomplete.

Response:

	Resolved.

2.4

Why use ICP-MS for paint?  If the intent is to be able to determine what
paint contributed what to the dust loading, then use it.  Otherwise,
it’s too expensive and gives no new information versus use of FAA
and/or ICP-AES.  Even use of GFAA is cheaper.  In general, more lead in
the sample means cheaper and easier analytical quantification (from very
cheap to very expensive: from FAA or ICP-AES, to GFAA, to ICP-MS).  A
good compromise between assuring a bunch of lead in the sample and
conducting jobs as in the real world is to analytically screen each
sample using FAA or ICP-AES (from $6 to $8 per sample).  If they are not
quantifiable by FAA, proceed to GFAA (may be as much as $35 per sample).
 If not quantifiable by GFAA, proceed progressively to ICP-MS analyses
(may be as much as $140 per sample). 

Most strongly recommend that only labs currently recognized by EPA under
the NLLAP should be used for analysis services (40 CFR 745.227(f)) even
though these experiments are not lead-based paint activities as defined
in Title X.

Suggest use of ASTM methods for all analyses.

Suggest use at least ICP-AES for analysis of wipe samples since FAA does
not have the MDL/MQL to achieve the performance stated (25% of 40
g/ft2).  Commonly, FAA gets a MDL of about 10 g/ft2 with an
allowed MQL of about 20 g/ft2 (about 50% of 40 g/ft2).  If want to
keep everything on-scale, suggest screen each sample using ICP-AES (from
$6 to $8 per sample).  If they are not quantifiable by ICP-AES, proceed
to GFAA (may be as much as $35 per sample).  If not quantifiable by
GFAA, proceed progressively to ICP-MS analyses (may be as much as $140
per sample).

2.8 See ASTM E1792-03 for specifications of wipe materials.

Response:

The laboratory selected for sample analysis is an NLLAP-recognized
laboratory.  Sample analysis will be done using Atomic Absorption
methods (NIOSH 7082 for air filters and EPA Method 7420 for paint, soil
and dust wipes).  The lab will provide the dust wipe materials in
accordance with current standards.

Comments to Characterization of Dust Lead Levels Associated with
Renovation, Repair, and Painting Jobs:  Summary of Study Design

Comments are shown by section title.

Outline of Design

General Comment

Suggest that the study follow the following outline per site:

Site selection; do not select carpeted work rooms;

Conduct a risk assessment (40 CFR 745.227(d)) of the entire site using
ASTM E2115-06 and E2255-04, for baseline documentation of site hazards;

Secure and seal HVAC, and build containment(s);

Clean work areas to background as presented in the design;

Conduct the site experiment(s) as presented in the design, using ASTM
standards where applicable, for example ASTM E1728-03 for wipe sampling
of settled dust;

Breakdown containment and power-up HVAC;

Perform post-abatement-like cleaning; and,

Conduct clearance (40 CFR 745.227(e)) of the entire site using ASTM
E2255-04 and E2271-05a.

Response:

The suggested outline coincides in many ways with the study design.  We
will not perform a risk assessment prior to work, instead we will have
each unit tested for lead based paint via XRF, and a cleaning crew will
clean the entire unit to clearance.  In general, the units that will be
used in this study will be vacant or abandoned, and we do not see
necessity in powering-up the HVAC after the work is completed.  The
vents will remain closed throughout the study process.

Interior Jobs, Phases, Cleaning and Plastic Sheeting

Para 2: a. Recall “ceiling to floor and out the door” for conduct of
cleaning.

b. Conduct all cleaning activities in disposable personal protective
clothing (PPC) and respirators.

c. Conduct personal air sampling (i.e., man exposure monitoring) of at
least all experimental activities involving dust generation and
cleaning.

Para 3:	a. Conduct all cleaning activities in disposable personal
protective clothing (PPC) and respirators.

Conduct personal air sampling (i.e., man exposure monitoring) of at
least all experimental activities involving dust generation and cleanup
using ASTM D5337-04 and D4522-97(2003).

Response:

All work phases will be conducted in disposable personal protective
clothing, including booties or similar footwear and full- or half-mask
respirators. 

Indoor and personal air sampling will be conducted throughout each study
phase. All study personnel will wear personal protective gear, including
respirators, and receive pre- and post-study blood tests.  The Health
and Safety Plan drafted for this study contains all the relevant
information regarding worker and site protection.  

Interior Jobs, Order of Sample Collection

Para 1:	Note that area sampling is usually used to define the OSHA work
area and should be collected when work is being done (sampling during
no-work times bias exposure results low).  Area sampling and personal
sampling should be conducted using ASTM D5337-04 and D4522-97(2003) as
simultaneously as possible to facilitate comparison.

Para 2:	a. More samples are always better than less.  I do not believe
that the study proposes to take enough samples from available useful
sites and may miss the opportunity to significantly add to the knowledge
of what is taking place during renovation activities.  This study
presents a grand opportunity to study actual renovation activities and
as such should be milked for all possible data.  For example, I would
expect that job sites involve and the rooms containing job sites will be
at least 36 ft2 (6 x 6, or so).  This would allow the opportunity to
collect 36 floor samples; samples close to the job activity and samples
at some distance from the job activity.  Given 4 sampling opportunities
per stage and allowing for no duplication of sampling site per sampling
opportunity, why not randomly collect samples from half of the available
floor space (4 ft2 of 8 ft2 available) per sampling opportunity.

	b. Randomly-selected, multiple samples of sites within replicate work
areas could/would quantitatively determine the dust loading as a result
of a replicate job phase.  Sampling site-selection bias is controlled
with random selection of sites within the work area to be sampled. 
Those sites previously sampled would be eliminated from the pool of
possible sampling sites during each sampling opportunity in a stage. 
Collection of multiple samples per opportunity increases the chance that
the actual leaded dust loading caused by the job will be measured. 
Multiple iterations of the job allows for comparison of sampling
results.

c. See the method for random selection of sampling sites in ASTM
E2271-05a.

Para 4: a. Suggest sample window troughs, especially following conduct
of exterior work.  The shortest route inside is through the windows.

More samples are better than less.  See Para 2 comments.

Response:

Indoor and personal air sampling will be conducted throughout each stage
of each experiment.

More floor samples are now being included in the work room for each
experiment. Sampling locations will be randomly selected and depicted in
a site-specific sampling plan for each of the study rooms.  

Exterior Jobs

Para 7:	Suggest sample window troughs, especially following conduct of
exterior work.  The shortest route inside is through the windows.

Response:

All of the windows will be closed during exterior experiments. 

Lead Dust Protective Measures

Para 2: a. Conduct all cleaning activities in disposable personal
protective clothing (PPC) and respirators.

b. Conduct personal air sampling (i.e., man exposure monitoring) of at
least all experimental activities involving dust generation and cleanup
using ASTM D5337-04 and D4522-97(2003).

Para 7: Suggest that the study follow the following outline per site:

Site selection; do not select carpeted work rooms;

Conduct a risk assessment (40 CFR 745.227(d)) of the entire site using
ASTM E2115-06 and E2255-04, for baseline documentation of site hazards;

Secure and seal HVAC, and build containment(s);

Clean work areas to background as presented in the design;

Conduct the site experiment(s) as presented in the design, using ASTM
standards where applicable, for example ASTM E1728-03 for wipe sampling
of settled dust;

Breakdown containment and power-up HVAC;

Perform post-abatement-like cleaning; and,

Conduct clearance (40 CFR 745.227(e)) of the entire site using ASTM
E2255-04 and E2271-05a.

Response:

All work phases will be conducted in disposable personal protective
clothing, including a full protective suit made of Tyvek or another
suitable material, booties or similar footwear, and full- or half-mask
respirators. 

Indoor and personal air sampling will be conducted throughout each
experiment. All study personnel will receive pre- and post-study blood
tests.  The Health and Safety Plan drafted for this study contains all
the relevant information regarding worker and site protection.  

The suggested outline coincides in many ways with the study design.  We
will not perform a risk assessment prior to work, instead we will have
each unit tested for lead based paint via XRF, and a cleaning crew will
clean the entire unit to clearance.  In general, the units that will be
used in this study will be vacant or abandoned, and we do not see
necessity in powering-up the HVAC after the work is completed.  The
vents will remain closed throughout the study process.

Comments from Reviewer P2

Summary of “Quality Assurance Project Plan for Selecting Sites for
Characterization of Dust Lead Levels After Renovation, Repair, and
Painting Activities”

Comments and Suggestions

In order to complete a risk assessment for the final rule, a field study
to characterize dust lead levels at appropriate stages of Renovation,
Repair, and Painting (RRP) activities is needed.  The first part of this
study requires identifying, screening, and inspection of possible sites
where the field study could be conducted. Sites targeted for inclusion
in the study are habitable, vacant housing units and vacant facilities
that could be used by children under age 6, such as daycare centers or
early year kindergartens.    

To be considered, candidate housing units and facilities that could be
used by children must meet the following criteria:

They must be built prior to 1978.

Selected units should be pre 1960, without significant history of
subsequent remodeling or rehabilitation.

They must contain lead-based paint (LBP) on 4-8 square feet (ft2) or
more on at least one component within a single room that will undergo
RRP activity, with LBP defined as dried paint film that has a lead
content at or exceeding 1.0 mg/cm2 or 0.5 percent by weight;

For housing units, they must contain three sequentially adjacent rooms
to allow sampling in a work room, a tool storage room, and an
observation room;

They must be vacant; 

Selected units should be vacant units returning to the market, rather
than abandoned units. Initial EPA testing was confounded by existing
condition of units to be cleaned. Units that have been vacated in
preparation for demolition will generally lack essential facilities will
have atypical hazard levels and additional deterioration beyond the norm
for target housing. Only units that are habitable, but vacant, should be
considered. 

Some costs will be higher for some work in vacant units. If vacant units
are necessarily included, the job should be carefully considered for
inclusion based on logistics of study and job mobilization (lights,
power, water, waste, support, security) and equivalency of job, to job
in a vacant but habitable building. The reviewer recommends any vacant
buildings chosen are fully habitable to avoid confounding the data
gathered.

The study logistics are a secondary concern but will potential affect
both cost and quality.

Response:

To the extent possible, we prefer to use a mix of housing which spans
the rule period. XRF testing will be conducted as part of the site
selection process to insure sufficient levels of LBP. 

We agree that vacant units may be more representative of target housing
than abandoned units; however, we must balance a desire for locating
housing characteristic of the target housing with availability of units
and safety of eventual occupants. EPA has an ethical and legal
responsibility to ensure that units are safe for re-occupancy. As such,
using a mix of abandoned housing and vacant housing may allow for a
wider range of jobs to be performed.

They must be accessible during the data collection period;

They must be cleanable before work begins, between study phases, and at
the completion of all study activities;

If window-related work is required, they must have an adequate number of
windows; and

If being considered for exterior work, they must contain an adequate
amount of yard space to allow for exterior data collection.

Selected units for exterior jobs should have yards free of abutting
shrubbery or other landscaping or external features in order to allow
placement of drops and collection of samples in an ideal environment.
Generally, plantings will interfere with protective barriers placed to
catch debris, but for purposes of the study, best case scenarios should
be selected if possible.

Response:

We agree that adjacent shrubbery or landscaping could complicate the
study. In cases where avoidance is not possible, we will cover the
shrubbery with plastic, as per the rule, and proceed with the study
protocol as indicated.

Permission to inspect and test units or buildings will be obtained from
site owners or managers prior to any inspection or testing taking place.
 The permission will include agreement as to whether and how the results
of the inspection and testing should be transmitted to the site owner or
manager.  Each prospective housing unit or facility will first undergo a
visual inspection to ensure that it meets the study requirements other
than the presence of LBP – three sequential rooms, vacant, cleanable,
etc.  The visual assessment will involve (1) gathering information on a
prospective housing unit or facility by viewing the property,
interviewing the property owner or manager, and checking county records,
and (2) sketching the property and labeling rooms appropriately to
support potential subsequent planning for XRF measurements, paint chip
samples, and RRP activities.  (3) gather the history of construction,
rehabilitation, remodeling, and repair from observation, anecdotal
sources, and property records.

Response:

While we agree that access to property history would be of value, the
prior work will not be included in the final analysis and thus will not
be collected for study-related purposes.

If the visual inspection confirms that a prospective housing unit or
facility meets all the study requirements other than presence of LBP, a
risk assessment will be performed. 

If there are other substantial identified risks, including high positive
dust or other sampling results, and the results of the risk assessment
fall within (or without) TBD parameters, a full lead-based paint
inspection will occur with a portable XRF device to measure lead content
of various painted components in the house – walls, window sills,
trim, doors, etc. 

[Note... some units that will be identified as containing large amounts
of intact lead based paint, may not be desirable for the study if
prework cleaning cannot be successfully completed economically. This
would be revealed by dust sampling evaluation prior to selection. The
study may want to quickly screen potential units before investing in a
full lead based paint inspection] 

Inspectors will obtain a large number of XRF measurements throughout a
house, which will be recorded by a study representative on the
appropriate data collection form. A portable XRF device that has a
HUD/EPA-issued or equivalent XRF Performance Characteristic Sheet will
be used.  The inspector will obtain paint lead content measurements from
the components of various rooms in the housing unit and record
measurements.  XRF measurements of lead in paint will follow the 16-step
procedure from Chapter 7 of the HUD Guidelines.  A certified lead-based
paint inspector or risk assessor will be utilized for the lead-based
paint inspections.  Lead paint results can usually be obtained without
damaging the painted surface and will be reported in mg/cm².  (In some
cases, an XRF reading may not be feasible and a sample for laboratory
analysis may have to be collected.)  Components that meet the PCS
criteria for lead-based paint will be considered as preliminarily having
lead based paint.  The area of each component that has lead-based paint
as indicated by XRF will be measured.

Following review of the XRF and corresponding area measurements at a
prospective housing unit or facility, study planners will determine
whether there is at least one room that could serve as the work room for
an interior job or whether there is at least one exterior side of the
house that could serve as work area for an exterior job.  If there is a
candidate work room or exterior work area, a preliminary matching of job
to housing unit or building will be made.   A determination will be made
as to which units and buildings are the best qualified for the study. 
Those units and buildings will be pursued as possible sites for the
study.   (Paint chip samples will subsequently be collected on
lead-based paint components at the units and facilities in the study to
obtain a laboratory value for the level of lead in the paint.)

Response:

The initial inspection of a unit by a site supervisor and XRF testing by
a certified technician will replace a full risk-assessment. The
evaluation proposed is sufficient for achieving the study objectives.

Detailed records will be kept of all field data collection activities. 
Information to be recorded includes:

Addresses of prospective housing units and buildings visited with
associated information including year built, property owner, vacancy
status, floor type, and likely ages of children that could or would be
at a building;

A signed permission form for each property at which XRF measurements are
obtained;

Paint lead levels on various components throughout each prospective
property as read via the portable XRF device;

A sketch of each tested property’s room layout identifying rooms in
which XRF measurements are obtained;

Location of XRF measurements and any paint chip samples identified by
level of property (0,1,2,3,…), room (BR1, DR, KI,….), and wall
(1,2,3,4);

Appropriate measurements of the surface area of components with
lead-based paint;

Condition of paint and substrate;

Lead levels found in any paint chip samples, if collected; and

Dust test results;

Letters or documents transmitting results to property owners or
managers.

All clearance documentation.

Response:

These items are not relevant for the site selection document, which
deals with paint testing and characteristics of the housing unit or COF.
 Dust test results and clearance documentation are discussed in detail
in the full study QAPP.

2.2 Sampling Methods

The study does not seem to fully consider lead dust wipe sampling or
soil sampling in the selection of sites. This information would normally
be part of a Risk Assessment. Risk Assessments are better indictors of
the overall suitability of sites for this study of the effects of
Lead-based Dust Hazards on RRP jobs. Pre and post (clearance) Risk
Assessments should be completed for each property selected. This Risk
Assessment (HUD) will form an important basis for further evaluation and
closeout.

Response:

For the purposes of this study, a complete Risk Assessment is not
required for site selection.  XRF testing will provide information on
the lead content of painted components, which will be confirmed by paint
chip samples if the unit is selected for the study.  Pre-experiment
cleaning and clearance will allow the experiment to begin with baseline
levels of dust.  Pre- and post-experiment clearance will be established
by a certified Risk Assessor or abatement firm at the conclusion of each
experiment.

Visual Assessment 2.2.1

Form C should specifically gather additional information on the
viability of future occupancy (unit is to be offered to the market this
date/unit is vacant and abandoned)

Form C should contain specific history of rehabilitation, remodeling,
and repairs:

Windows 		original/replaced this date

Floor covering 	original/ replaced this date

Cabinets		original/replaced this date

Etc.

Risk Assessment 2.2.1.1

A Risk Assessment (HUD) step should be inserted for those units that are
selected based on Visual Assessment. Risk Assessments include dust
testing and supplement the information obtained through the Visual
Assessment at a fraction of the cost of the XRF inspection. Full Risk
Assessment will rule out problematic units containing lead based paint.

Response:

The initial inspection of a unit by a site supervisor and XRF testing by
a certified technician will replace a full risk-assessment.  The
evaluation proposed is sufficient for achieving the study objectives.

2.2.2 XRF [Lead Paint Inspection] 

This section should be referred to as Lead Paint Inspection, and carried
out as such with certified individual available. The use of XRF is a
component of a Lead-based Paint Inspection. 

Response:

XRF testing is being used as the primary site-screening tool in the
inspection.  Subsequent testing will occur at other study stages.
Therefore, the current title is appropriate.

2.2.2.1 XRF

2.2.2.2 Paint Chip Samples

2.2.2.3 Dust Wipe Sampling

The study should use existing clearance protocols (appendix G). This
appears to be partially extracted in Appendix E. Additional samples
above the minimum are important to the study, but all samples required
for clearance should be included at the 4 stages, and before
commencement (which maybe accomplished in the initial proposed Risk
Assessment, and at close, six (6) sets altogether. 

Reviewer recommends the entire unit be tested at all six (6) stages,
within and without containment, for consistency.

Response:

Pre-experiment clearance samples will be collected from the
study-impacted areas that are cleaned prior to work commencing. 
Post-experiment clearance samples are collected from the entire unit, as
every room/area is cleaned after all study activities have been
completed at a property.  Sampling throughout the entire unit at each
stage will not provide valuable data and will significantly increase the
time and money required to complete the study as there are 60 interior
experiments to be conducted.

2.2.2.4 Soil Sampling

Additional soil sampling should be included for both interior and
exterior jobs as standard protocol. This post job soil sampling would
substitute for the sampling of debris and dust below the exterior
plastic sheeting. This is standard in evaluating risk and clearing
units. Maybe required for units under federal management as part of
clearance. Composite samples would be acceptable.

Response:

Pre- and post- experiment soil sampling is included in the study
protocol for both interior and exterior experiments.

Use existing clearance protocols (appendix G) Clearance

2.2.2.5 Water sampling (see power washing discussion)

Response:

Power washing is not included in the study and, therefore, water
sampling will not be part of the study design.

Comments from Reviewer P4

I, [P4], am responding as one of the peer reviewers of the EPA’s study
of lead levels associated with RR&P activities.  My review is formatted
as direct responses to the questions posed.

In general I find that the study seems well thought out and considered. 
I don’t recognize any obvious weak links, although I have made several
general comments in regard to areas I believe pose potential risks. 
Mostly, these comments amount to carefully supervising the contractors
performing the tasks at hand to make certain quality control is being
instituted.  As a remodeling contractor and owner of multi-unit housing,
I assure you that this is an area of risk in any project.

Thank you for the opportunity to share my ideas with you; I truly hope
that solid data results from the performance of this study.  If there is
any way I may be of further assistance please don’t hesitate to
contact me.

Response:

We agree that attention to quality control is a vital part of this
study.

Comments from Reviewer P6

General Comments

I have reviewed a number of studies of this type in the past and have
been involved in several from an analysis perspective. My comments have
been brief because this study is well designed, the analysis is
carefully thought out and there is high likelihood that it will achieve
the desired result.

If possible, the study should include as many distinct cities as is
financially possible to improve the prospects of generalization.

Response:

We agree that the study would benefit from being conducted in as many
different cities/locations as possible in order to improve
generalization to the nation as a whole.  In attempting to balance this
objective with time, resource, and feasibility constraints, it has been
determined that the study will include 2-3 distinct geographic areas.
Potential sites have been located in Columbus, Ohio; Pittsburgh,
Pennsylvania; and Arlington, Virginia. 

Comments from Reviewer P7

Attached are the documents generated by my review.  Only the first two
are essential; the others are raw notes, but I attached them because I
thought they would be useful in interpreting what I did say.

There are eight documents attached to this cover letter

	1	Responses to Charge Questions

	2	Some Overall Concerns

	3	Notes: Study Design Summary

	4	Notes: Study Design Detail

	5	Notes: Site Selection QA Summary

	6	Notes: Site Selection QA Detail

	7	Work Site Size and Dust Wipes, description of model and results

	8	Work Site Size and Dust Wipes, spreadsheet model

  SEQ CHAPTER \h \r 1 I.	Summary of Major Concerns

My major concerns overall fall into three categories.  While these are
reflected in my Responses to the Charge Questions, I thought I would
summarize them here.

A. 	Lack of a Dispersion Model

The clearest example of this is that there is very little attention to
sample locations as a variable in the final results.   On a full
abatement, it is anticipated that the entire area will be contaminated
because 1) the activity occurs over a wide area and 2) the activity
occurs over a long time, allowing spread of contamination; thus the
relationship between the work area and the sample location is a very
minor part of the training of a clearance examiner.   But even on full
abatement this is an untested and possibly false pair of assumptions.  

On an RRP job it is almost surely false.  Because of the dense nature
and “spherical” shape of the particles, sanding dust probably falls
away from the wall in a log normal fashion.  It will be spread by foot
traffic along the path of the work area progress and toward the
entrance/exit of the area and possibly toward equipment/supplies holding
in the work area.  Thus, there will be a fairly predictable
distribution.  On a sill sanding operation, for example, there will be a
fanshaped pattern densest near the center of the sill and falling away
relatively rapidly from the edges of the work.  To locate a sample by a
single linear measurement (center of sample area to nearest edge of work
area) means that a sample taken to the extreme right away from the
window is treated the same as a sample taken at the same distance
perpendicular from the center of the sill.  

While I admit that the effect is difficult to deal with, especially
where repeated sampling will occur in the same area, to ignore it during
the sample site selection and data evaluation phases of the work is to
invite conundrums during data analysis.  

Because of their flat shape, generated paint chips are even more
difficult to model.   Thus the placement of but a single tray on the
plastic during exterior work (where drop distances are generally greater
than indoors) seems foolish indeed.  

Response:

We have added collection of more floor samples to the work room during
each experiment for interior jobs.  The site-specific sampling plan is
designed such that over-sampling occurs close to the work area and the
exact location (x,y) of the sample is recorded for modeling purposes. 
For exterior jobs, six collection trays on or near the plastic will be
used for sample collection, in addition to three underneath.  

B.	Failure to Match Testing to Fundamental Questions

I detail many of these elsewhere.  If this design is used we will have
detailed information about the lead content of paint in the dwelling
which will remain intact throughout the study but we will have no idea
of the dust levels in the dwelling at the start of the pre-cleaning
operations (and therefore the possibility of re-injection of dust from
outside the work area).  In addition, while we will have all this
detailed data on lead levels (including confirmation by paint chip
analysis!), that information is never used in subsequent evaluations of
results.  No attention is paid to the   SEQ CHAPTER \h \r 1 methodology
for determining the actual Level of Dust Generation by any particular
RRP activity, yet that is critical to an apples-to-apples comparison
during the the multiple regression data analyses, as is a reduction of
lead dust levels to a set of values normalized by the actual lead levels
on the painted surfaces during the renovation work.  Comparison to EPA
lead hazards is irrelevant, since that will be performed at a later
stage in the risk evaluation, but comparison to the maximum possible
lead is highly relevant and ignored.  

Response:

By randomly sampling the floor and windowsills in the three study rooms
at the various stages of dust collection, we believe that the study will
yield sufficient information to determine the spatial dust-lead
distribution generated by each RRP activity – as well as the
efficiency of containment and cleanup efforts in reducing the dust-lead
loadings. 

The level of dust generation associated with particular RRP activities
will be empirically measured as part of this study.

C.	Lack of Attention to Consistent Performance During Dust Wipe Sampling

Probably the greatest single opportunity for seriously corrupted data is
inconsistent or improper application of the dust wipe sampling.   But
absolutely no attention is paid to how to solve this problem.  Should
all clearances be done by the same person?  Should videos of all
sampling be made and retained so that later anomalies can be checked? 
(Location concerns were discussed above.)  Should two technicians be
required to take “side by side” wipes in the same area?  By my own
observation of our trainees, I am sure that there will be significant
inter-examiner differences in the % dust recovery from surfaces; it
should be at least considered at some point.  

Response:

Technicians will be trained in the use of dust-wipe samples, and
Battelle will conduct quality assurance inspections to verify consistent
performance.  All clearance samples will be collected by a certified
Risk Assessor or clearance technician.

It is anticipated that the same technicians will perform all of the dust
wipe sampling during the experiment, except for the pre- and post-
clearance samples, as noted.  The data collection forms will keep track
of the technician that collects the dust-wipe samples for evaluation of
inconsistency concerns.  

II.	A Niggling Concern:  Terminology

As I began reading the study design summary documents I found myself
bogged down in confusing terminology.  It has been my experience that
poorly used language (whether on my part or that of others) indicates
poorly formed concepts.  Furthermore, they lead critics to misfocus
their attention and study designers to overlook key distinctions.  Some
problematic terminology in the documents are:

“Phases.”	The most egregious and confusing is the use of
“phases” for “job performance scenarios.”  What genius decided
to call something phases when there is no implication of time or
sequencing in the underlying data being described?  Even if we assume
that RRR projects at the same facility will be conducted using more than
one job performance scenario, the sequence of scenarios will be varied,
unlike the phases of the moon or the phases of a sine wave.   Lose the
term.  “Job Performance Scenarios” (or “JPSs”) is descriptive
[that was the term first used in the summary] and short enough to be
useable.  In the document itself the very apt term “P/CU routines”
is used for this.   After the P/CU routines for a given site are
selected, the sequence of the particular P/CU routines at that site
are properly spoken of as phases, but since that may only be two
routines at a site, speaking as the full set of P/CU routines as phases
is incorrect and misleading.   

  SEQ CHAPTER \h \r 1 “Stages”	This is acceptable, but somewhat
confusing.  The job is staged, and this is the way that we determine
residual hazard following each job stage.  These are referred to (once)
in the main document as project sampling “timepoints” (I prefer
“checkpoints”), but at least the main document terminology captures
the central idea.  Presumably there is no staging in the care taken by
the sampling technician nor in the techniques to be used in conducting a
dust wipe, only various points in the work progress for checking the
results..  

“RRP”		The problem is not “P” (painting) it is “R”
(repainting).  RRR (or R3 to lose a syllable) is the logical term.  I
know that U.S. EPA has used the term (although I more often here it
called the R&R rule in the leadnet world), that does not hallow it. 
Again, sloppy terminology implies sloppy conceptualization.  

“Levels of Generation” Generation of What?  The correct term is
“Levels of Dust Generation.”  

		As described below, there is a significant difference in the
generation of visible chips and of invisible dust.  Chips are too dense
to spread widely, too large to inhale, and to simple to clean up
compared to dust.  Since the hazard is due to airborne and settled dust,
any term that implies equality should be avoided.  

Response:

 A glossary is included to aid the reader in understanding study
terminology.  The rule title and associated acronym are outside our
scope of responsibility.

  SEQ CHAPTER \h \r 1 Thoughts on Summary of Study Design

Note: Throughout the designers use the phrase “children under 6.” 
This is appropriate to the HUD standards, but 40 CFR 745 usually uses
wording such as “children 6 years of age or under.” [See, for
example, 40 CFR 745.223 where “child-occupied facility” includes the
phrase “ child, 6 years of age or under.”]

RRP(renovation, repair, painting)[=RRR(renovation, repair, repainting)]

“Phases” (Scenarios) for job performance requirements

“Phase”	Use Plastic Sheeting	Perform Specialized Cleaning

1	 + 	+ 

2	+ 	-

3	-	+ 

4	-	-



What genius decided to identify time-independent entities as
“phases???”  Why not just call them “performance scenarios”
since they are describing alternative scenarios for performingn the
task?

“Stages” are the individual points in time for sampling activities
[for Interior jobs]  (Sampling Schedule)

	-1	pre-RRR cleanup completion		Floor and sill dust wipes only, work and
observation rooms and all other unsealed rooms

	0	cleanup correction samples		Retest after recleaning spaces that
failed to fall below EPA lead dust hazard levels after initial cleaning.

	1	RRR task completion			Air, then floor and sill dust wipes, work and
observation areas

	2	cleaning completion			Floor and sill dust wipes only, work and
observation areas

	3	wet cloth cleaning verification	Floor and sill dust wipes only, work
area only

	4	dry cloth cleaning verification	Floor and sill dust wipes only, work
and observation areas

	5	job shutdown				Floor dust wipes only, work area and exit

	6	Facility clearance			Since several Performance Scenarios may be
conducted in the same facility, a final set of completion clearance
samples will be conducted.

Exterior Jobs:

	0	Pre-job				Soil samples only

	1	RRR task completion			Dust wipe samples, trays on plastic only

	2	job shutdown				Dust wipe samples, trays below and adjacent to
plastic only.  

	(No description of steps to ensure low soil levels or to verify cleanup
prior to reoccupation.)

Outline of Analysis (there is none: What is shown is a listing of the
data summary to be prepared.   Why Averages?  How will final cleanup
levels be related back to LBP analytical data and pre-job dust wipe
results?  Will raw data be available to researchers and stakeholders or
only the summaries?  What protocols are in place (QC not QA) to ensure
consistency (ASTM dust wipes, personnel use, etc.)?  Are “phases”
randomized rather than sequential effects?  How will “learning”
effects be recognized and dealt with in the analysis?)  

Response:

The sampling protocol has been revised and the notes presented by the
reviewer are no longer representative of the current protocol.

Quality Control measures are explained in section 2.5 of the Full Study
QAPP.  They  include all QC measures taken by the lab in accordance with
the Laboratory Quality System Requirements (LQSR) developed by EPA,
along with field blanks and laboratory spike samples included in the
sample batch, without identification as such.

The phases of a job will be randomly assigned and will not follow a
sequential order.  

Levels of [dust] Generation:   High, Medium, Low

Six RRP jobs to be done (two at each Level of Generation) and two by
each Performance Scenario

Target Housing:  Interior Study

(2 RRP Job Types x 3 Levels of Generation x 4 Performance Scenarios) x 2
(for significance) =  48 Jobs

Target Housing: Exterior Study

(2 RRP Job Types x 3 Levels of Generation) x 2 (for significance) = 12
Jobs

Child-Occupied Facilities: Interior Study

(1 RRP Job Type x 3 Levels of Generation x 4 Performance Scenarios) x 1
(no significance measure) = 12 Jobs

Child-Occupied Facilities: Exterior Study

(1 RRP Job Type x 3 Levels of Generation) x 1 (no significance measure)
= 3 Jobs

Total: 48 + 12+ 12+ 3 = 75 jobs

[Query: What is learned by doing Target Housing separate from
Child-Occupied housing.  Why does the important HUD
Guidelines distinction between single- and multi-family housing not
show up at all? Also, will occupied v. unoccupied impacts be studied??]

Presence of lead-based paint: significantly greater than 2SF I hope!  

Professional Workers: What curriculum will be used for the training? 
Will workers be trained in one class for all 75 jobs or will the
possibility of disparate training arise??

Paint measurements: Why no mention of measuring paint sample size?  What
will be done with the comparison of results from #2 and #3.  Will
residual paint lead levels be measured after the job is completed?

Response:

All study personnel will be trained on the use of a respirator and in
Hazard Communications.  At least one representative of the RRP
contractor and the site suportvisors will be trained in Lead Safe
Renovation.  In addition, hands on training will be provided for the
contractors and field technicians on the detailed study protocols.  

The paint chip samples are requested to be 2 inches by 2 inches, and the
technician is required to record the exact size of the paint chip on the
data collection form.

Prestudy Cleaning (p. 3): What certification exists for abatement
clean-up as opposed to abatement?  Will the same lead contractor and
supervisor be used on all jobs?

Response:

	As the study is conducted across multiple states, the same certified
Risk Assessment or clearance firm and site supervisor will not be used
on all jobs.  Wherever possible, the same firm and site supervisor will
be used for jobs within the same city.   

Rule (presumably “Specialized”) Cleaning v. Normal Cleaning

  SEQ CHAPTER \h \r 1 Detailed Study Design and QAP for Characterization
of Dust Lead Levels.

Six basic questions.

	-	Effect of activity Level of Dust Generation on final dust levels

	-	Effect of plastic sheeting of final dust levels

	-	Effect of cleaning/verification on final dust levels

	-	Effect on adjacent areas of plastic sheeting and/or
cleaning/verification practices.

	-	Effect of second cleaning on final dust levels.

Note: Neither the word “phase” nor the word “scenario” appears
in the description on p. 1-4; “protection/cleanup routines” (P/CU
routines) is the phrase of choice.  

Response:

	The text has been changed to reflect consistent use of language.

Note: The word “stages” is not used in the description on p. 1-5;
instead they are referred to as “timepoints.”

Page 1-6: Critical point: “100 percent of work activities should be
accurately assigned to a work activity level.”  “Jobs will be
pre-assigned to an activity level based on anticipated dust lead levels
to be given off by that job.  After post-work dust lead levels are
analyzed, however, an activity can be assigned to a different actiity
level if average dust lead levels across all phases [sic] conducted for
that activity were found to be significantly lower or higher than
expected.  Variability and spread of the dust levels on each job will be
considered in evaluating whether the job was assigned to the proper
intensity level.  This would impact analyses comparing differences
betwee work activity levels or specific jobs.  

page 1-7: Airlock only on jobs using floor poly?

Response:

	Yes, airlocks will only be used in conjunction with the phases of the
job where rule containment, and thus floor covering, is used.

Page 1-8.  What are the air samples evaluating during cleaning?

Response:

	The air samplers are evaluating the difference in the level of lead
dust in the air between the two cleaning routines - rule cleaning and
baseline cleaning.  

Page 1-9: Why are the trays wiped rather than simply washed and the
washings digested and assayed?  Why is there no comparison with and
without removal of larger chips (say by use of screens)?

Response:

	The samples will be taken from the dust collection trays, as is,
without the removal of larger chips to get a better representation of
the actual dust and debris generated from the work.

Page 1-10: Why two cleanings with a wet cloth?  

Response:

The wet cloths are not used for cleaning purposes, rather, they are used
for verification of the cleaning (either rule or baseline).  The
proposed rule specifies that two wet cloth verifications be performed
and, in the case of failure, re-cleaned before proceeding to the next
verification.  Note that the floors are not re-cleaned between dry cloth
verifications.

page 1-11: 

	-	Use of window air conditioner should be documented, including rated
flow of unit.  

	-	Since RRP work is not a geometric point, what is meant by “distance
of sample locations from work activity.”  Does this imply that only
work on vertical surfaces will be performed; that is, no ceiling work?  

Response:

	If a window air conditioner is installed, it will be located in a
non-study impacted room, and will be noted in the site supervisor’s
assessment of the experiment.

	The data collection forms are set up so that the exact location of the
samples can be identified as the distance from two walls, such that all
samples have the same location specification.

p. 2-1:

	-	What specifically are the expected dust lead levels for each of these
jobs and what are the levels cutoffs for different levels of dust
generation?  

p. 2-3	The term “phase” which had been carefully avoided in Section
No. 1 has now crept in, but it is being used appropriately.  That is,
the job will be segmented into up to 4 separate jobs. Note that no
mention of randomization of job segment P/CU routine sequences.  \

Response:

	Phase is used consistently throughout the document to represent the
protection and clean-up routines used.  The experiments will be randomly
assigned a phase, as indicated in the full study QAPP. 

p. 2-6	“Confirmatory paint samples will be collected . . . , as
necessary?”  Why would it ever be necessary?  Wasn’t this tested in
the site selection protocol??

“Within the work room, the dust wipe sampling locations should be
within five feet of the work area.”  Why not “as close as possible
to the work area, but in no case further than five feet.   When testing
the P/CU routines at a given level of dust generation, care shall be
taken that all samples are at a comparable distance (horizontally and
vertically) from the work area.”  

Response:

Paint chip samples will only be collected once.  This section is a
carry-over from the site-selection QAPP.  The sample collection protocol
has been revised and is included in the full study QAPP.  It now
includes a site-selection sampling plan which randomly assigns the
locations of the dust wipe samples to be collected at each stage.

Within the adjacent rooms, the sampling locations will be located around
the center of the room.”  Why?  The logical place would be at a site
as close as possible to the work area, since it is spread we are
concerned about.  

Response:

	A site-specific sampling plan has been created for the adjacent rooms
such that the locations of all the samples are random and
pre-identified.

What common pattern will be used to position the trays so that 1) the
positioning is comparable across various P/CU routines tested; 2) the
trays do not interfere with the RRP job being performed; and 3) a
worst-case dust release scenario can be realistically developed.  

Response:

	The phases are only evaluated for interior jobs, where dust collection
trays are not used to collect dust and debris.  Interior dust wipe
samples are taken from either the plastic floor covering or the floor
itself.  Dust collection trays will be used for exterior experiments
only.  

p. 2-9	“confirmatory paint chip samples” Why paint chip samples? 
Why ICP-MS rather than flame AAS?

Response:

	Flame AAS will be the method used by the laboratory to analyze the
environmental samples.

p. 2-10 “use a detergent with high phosphate content.”  Why? 
Several studies have indicated that phosphate content alone is
insufficient to guarantee superior cleanup of LBP-contaminated sites.  A
more important point is that all cleanups will employ the same
detergent, phosphate or not.

Response:

TSP will not be used for the cleaning in this study.  After consulting
with firms familiar with abatement cleaning, it has been decided that
Simple Green will be used as the cleaning detergent for all experiments.

p. 2-12 Is the location of the post-work dust wipe samples in the
adjacent rooms of no consequence?  Not guidance is provided regarding
where to sample.  

Response:

	A site-specific sampling plan has been created for the adjacent rooms
such that the locations of all the samples are random and
pre-identified.

p. 2-14 Will a standardized “wet disposable cloth” be used?  Will
blanks be collected?

The protocol here (not the same as described above) appears to be one
wet cloth (plus a pass/fail determination) followed as many cleanings
followed by dry cloth wipes as are needed to get clearance by agreement
with the photograph.  The earlier procedure allowed for 2 wet cloth
verifications before going to a dry cloth verification.  

Response:

	All of the wet disposable cleaning cloths used for cleaning
verification will be the same across all experiments.  The proposed rule
calls for two wet cloth verifications and re-cleanings before moving on
to dry cloth verification, and that is the protocol that the study will
follow.

p. 2-15 While windowsills need only one wet wipe test, floors apparently
need two before shifting to dry cloth verification.   What must be done
after the fourth wet cloth fails?

Response:

	Windowsills, similar to floor surfaces, undergo two wet cloth
verifications and re-cleanings before moving on to dry cloth
verification.  Due to the inherent nature of some surfaces, it may be
difficult to achieve white glove (pass verification), therefore we are
limiting the number of dry cloth verifications to four.  Note that the
entire house will be cleaned and cleared before the study is complete.

p. 2-19 Earlier it was indicated that a wind sock and a wind gauge would
be used; here the gauge is omitted and the technician apparently
estimates wind speed (sentence incomplete).  Use the wind gauge!

Response:

	A wind speed measurement device will be used at all exterior experiment
sites.

p. 2-21 Will composite wipe samples be tested (NO!)?  EPA allows
composites.

Response:

	The study objectives seek to characterize the spread of dust resulting
from RRP work.  Composite dust wipe samples will result in more
generalized concentrations instead of location specific concentrations.

Notes on Site Selection QA Summary

  SEQ CHAPTER \h \r 1 QA procedure fails to note interaction between
analytical methodology, project size and type, and significance of
results.   

Why is total area of LBP important? (As opposed to QA in disturbance
area?)

Why is a a total inspection being performed rather than an examination
of the lead content of materials in the study areas (work area,
observation area, plus access routes to the work area)?  Is there a
feeling that collateral damage will be an important source of lead dust?
 And, even so, since those areas will be sealed off, how will they
impact the study?   

Response:

	A total inspection of the property using an XRF instrument is conducted
in order to determine if the unit is a candidate for the study, and
which rooms would best serve as the study rooms.  Prior to the XRF
testing, the study rooms are not identified.

Why are no dust wipe samples being collected.  Since one criterion of a
suitable site is the ability to clean it to 25% of the EPA dust-lead
hazard level, why not determine how difficult that will be in a given
property so that subsequent cleaning activity has a proper baseline? 
Again, focus on the potential study, observation and access areas.   

Response:

	We believe we can cost-effectively assess ability to clean based on a
visual inspection of the site. The cleaning will be performed by people
who have the credentials and experience in cleaning after lead abatement
activities.  They, along with the clearance results, will ultimately
determine how difficult it will be to clean a property.

Why is XRF done before suitability for RRR job?  Given the expense of an
inspection, it would seem more cost-effective to send someone into a
series of houses with an XRF and if they find suitable areas for the
projects, they then check the lead level of paint to be disturbed in the
potential study area.  If that is satisfactory, then an inspection can
be done (if desired).   

Response:

The site selection protocol calls for an initial inspection of a
property for important study related characteristics before the XRF
testing occurs 

Notes on Detailed Site Selection QA

  SEQ CHAPTER \h \r 1 p. 3	Table 1.2	-“Low Level Work: Interior
Jobs”: door removal, planing is duplicated.

			-“Medium Level Work: Interior Jobs”:  In what way is removing a
window comparable in the level of dust generation to removing 100SF of
drywall?  According to standard protocols, the paint adjacent to the
frame is cut, the framing removed and the new window inserted.  Properly
performed this should be low-level work unless the window’s paint is
in terrible condition.  On the other hand, drywall removal always
creates significant dust levels, including paint (probably primarily
chips).

–“Medium Level Work: Exterior Jobs”: Why is a carport ceiling
safer to safer to repaint than porch components or 100 SF of exterior
wall (which are listed as High Level Work”

- “High Level Work: Exterior Jobs”: “Strip paint from brick” Is
this being done by needle gun and HEPA vac (Medium level at worst, low
if properly conducted)?  By wet stripping (Ditto for needle gun, and low
for dust generation as opposed to dust and chip generation)?  By
chiseling or some other agressive hand technique?  By torch burning?  By
abrasive blasting?

This table of tasks looks like the result of a team brainstorming rather
than a summary of background studies with objective outcomes.  The
comparability of data is sorely strained if nothing better than this can
be given as a categorization of Levels of Dust Generation.

Response:

	The list of jobs has been revised and is more specific in the
explanation of the work to be performed.  EPA has determined that the
jobs are in the appropriate categories, but, as stated, analysis of the
dust generation after the work is completed may result in a
restructuring of the job list.

Section 1.5.1.   For comparability it is very important to use vacant
housing, particularly since some relatively unsafe methods will be used.
 However, it is important to keep in mind that this means that dust
dispersal by occupants (sometimes a serious problem) will not be tested
for.  

Response:

	We will use vacant housing so that no occupants or other non-study
persons are exposed to the potentially high levels of leaded dust
generated during the RRP work.  There will be contractors, field
technicians, and other study personnel in the property at various times
throughout the experiment process, which should provide some insight
into the spread of dust by tracking.

Section 1.5.2 – Cleanability doesn’t remove the risk of
cross-contamination (that is dispersal from work area to observation,
storage, and access areas).  It allows us to clean and therefore be able
to distinguish residual site contamination from work-related
contamination (whether cross- or in situ contamination)

Section 1.5.3 – Unlike the summary, this emphasizes the use of the XRF
only in the potential work rooms (the correct procedure).  

Response:

	The XRF testing will be performed in all components in potential work
rooms, which may include the entire unit after the visual inspection.  

Section 2.1.1	p. 2-1; see discussion of “4-8 SF” appended. 

Section 2.1.2   p. 2-2.  Unlike Section 1.5.3, this section includes the
notion of a full inspection again.  This is unnecessary and costly.  Why
is it even suggested?  

Also, no justification or rationale is given for when and where paint
samples will be collected for ICP-MS sampling.  Why is ICP-MS being used
anyway?  At the level of precision required for this analysis, AAS would
be cheaper and sufficiently precise?  Again, what is the rationale for
this expensive add-on?

Response:

	In this case, full inspection refers to a visual inspection by study
personnel and XRF testing of potential work areas.  

Flame AAS will be the method used by the laboratory to analyze the
environmental samples.

Section 2.2.2.  “A full XRF lead screening inspection will occur in
prospective sites” This should make it clear which parts of the site
are to actually be tested; full inspection would mean all areas.  

Response:

	The XRF testing will be performed in all components in potential work
rooms, which may include the entire unit after the visual inspection.  

Section 2.2.3.  “conduct laboratory analysis of paint-chip samples to
confirm the presence of lead based paint measured by the portable XRF
device. . . “ Does Battelle know something about the XRF readings that
is not in the EPA Performance Characteristic Sheets?  Confirming XRF
readings seems needlessly complex and costly.   It is certainly true
that peeling paint in area with no sound paint may require the use of
paint scrapes.   If sound paint is being tested, how is the efffect of
substrate on the analytical result being compensated for?  If only
damaged paint is tested, how is it determined that all layers are being
sampled?   

Response:

	The terms “confirmatory” and “confirm” have been removed as
they did not accurately describe the utility of the paint chip samples. 
The paint chip samples will be used to get a precise measure of the
paint lead levels in the exact location on the component that work will
be performed on.

	

The paint chip collection protocol calls for collecting all layers of
the paint down to the substrate, minimizing the amount of substrate
actually collected.  The technicians collecting the paint chip samples
will have undergone training in the proper technique for sampling both
sound paint and damaged paint.  The paint chip samples will only be
taken from those components that will be affected by the RRP work for
that unit.  

Section 2.5.1.   These are field blanks for paint chip samples, I
assume.  That being the case, 10% of sample levels is far too high.  Any
result of lead above detection should be suspected and the sample vials
checked from a stored, unopened container.   Appropriate statistical
tests should be run to determine that the contamination is definitely an
outlier.   (Airborne dust in some sites may be a problem, but I can
hardly imagine that the amount of lead found in such empty vials should
exceed even 1% of the amount of lead found in the actual paint samples. 
If there is no lead in the paint, there should certainly be no airborne
lead in the area.)  

Response:

	The quality control measures have been revised and are included in the
full study QAPP.

Section 2.5.4.   Why is a laboratory duplicate rather than a field
duplicate being called for?

Response:

	Duplicate samples will be performed by the laboratory in accordance
with the LQSR.  Control spike samples with a known amount of lead will
be submitted blindly to the laboratory to assess the accuracy of their
analyses.

Section 2.9.  	For prior inspections, couldn’t a spot check be used to
verify the accuracy of the earlier, but not documented by current study
standards, rather than repeating the full inspection?  (However, if only
the potential work areas are to be inspected and those by XRF only, then
I do not see the rationale for accepting any previous inspections.  The
time for checking QA/QC of earlier studies is probably comparable to the
study time, and LBP may have been removed from previously inspected
areas leading to obvious problems.  

Response:

	The results of prior inspections will be considered when determining
potential units and potential work areas; however, to maintain the
integrity of the study, we will perform our own inspection of the unit.

  SEQ CHAPTER \h \r 1 Effect of Work Area Size and Paint Lead Content on
Precision of Final Sill Wipe Results

  SEQ CHAPTER \h \r 1 While working through the summaries, I found that
the designers had never projected the actual results expected from the
wipe analyses.  As a result of my analysis, it appears that if the site
selection criteria are rigorously followed, some significant data will
remain indeterminant.  

To explain briefly.  

The RRP Contractor will go into an area of at least 1 mg/cm^2 lead-based
paint and "disturb" an area of at least 4 SF (QA for site selection
summary.)  At the end of the job (in stages) the clearance examiner will
collect floor and sill wipes and compare them to the EPA dust lead
hazard levels (40 ug/SF on floors, 250 ug/SF on sills; troughs will not
be tested).  A significant result would result if level exceeded the
hazard level OR failed to exceed the level when it could have.  

Response:

	The samples collected between each stage of the experiment will not be
compared to the EPA clearance levels.  After the experiment is finished,
the entire unit will undergo a thorough cleaning and tested for
clearance.  We do not foresee the clearance samples exceeding the
clearance levels after the thorough cleaning is conducted.

It is this last which is problematic.  On the attached spreadsheet I
have taken a typical allowed job under this QA procedure, namely a 120
SF room with 1 mg Pb/cm^2 lead levels in which the RRP contractor will
"disturb" 4 SF of painted surface.   

I have put "disturbed" in quotes because it is important to realized
that most RRP work does NOT reduce LBP to invisible dust on the areas
disturbed.  Much of it remains on the surface or comes off as visible
particles.  These will mostly be dealt with during the routine cleaning
phase.  Thus, I would consider 40% reduction to dust to be a very high
level of dust generation.   

After the job completion the job site will undergo cleaning.  During
this phase (wet clean and shop vacuuming)the site will have some
fraction of the dust removed.  Assuming that in a non-plastic-covered
area, up to 95% removal using standard cleaning is not unreasonable to
expect, then a "typical" high level of generation project followed by
normal cleaning should be 40% dust generation followed by 95% dust
removal from floors and sills.  Using this scenario, then, you can see
from the spreadsheets that for a 4 SF RRP job, there would be 619 ug
Pb/SF of surface, which is about 15.5 times the floor limit and 2.4
times the sill limit.  But that assumes that all the dust would settle
equally on the floor and the sill, which is NOT a sustainable argument. 
Since LBP paint dust is dense and compact, it will drop quickly.  Thus,
all work done below the sill level will probably never reach it and work
done above the sill level will reach it only on a straight drop!  If we
assume that on average 1% of the dust will actually reach the sill, we
can see that  even at 100% release and 80% cleanup, the sills will
probably pass.   

This seems to be a serious problem in experimental design.  I have also
done a scenario for 20 SF and here we see that at 100% release the sill
wipes should always be significant.   Note that if we had a 4 SF site
with 5 mg Pb/cm^2 in the paint, the results would be the same as this 20
SF, 1 mg/cm^2 site.

  SEQ CHAPTER \h \r 1 The basic message I read from this is that the
selection of sites (and the data analysis)has to emphasize that

1.	Acceptable sill readings must be due to proper practice or cleanup,
not simply to the site selection.

2.	To achieve this requires carefully considering four factors

  SEQ CHAPTER \h \r 1 a.  	Amount of lead in the paint

	b.    Area of paint being "disturbed"

	c.    Nature of disturbance activity and actual level of DUST
generation. 

		(Chips will behave differently and will be cleaned up more effectively
because they are visible.)

d.	Positional relation (height and horizontal separation) of sill and
site of dust generation.

Attachment: Quattro Pro Spreadsheet model of the expected dust wipe
samples for various amounts of paint disturbed and different
effectiveness levels for cleanup.  

Relationship between lead on surface and suitability of room









	Assume:







	Paint lead level	1	mg Pb/cm^2





Room Size	120	SF(10x12)





Job Size	20	SF





Constants:







LBP	1	mg Pb/cm2





unit conversions	929	cm^2/SF







1000	ug Pb/mg Pb





EPA dust-lead hazard levels







floor	40	ug Pb/SF





	sill	250	ug Pb/SF























Scenario







	A renovation contractor disturbs 4 square feet of 1 mg/cm^2 LBP while
conducting a repainting job. (Per QA for site selection.)



















Possible Outcomes:







Actual released area of paint (as % of total)





	Release Levels

Multiples of dust-lead hazard











	ug Pb total	ug Pb/SF

sill	floor	sill (1%)

	100%	18,580,000	154,833

3,871	619	6

	80%	14,864,000	123,867

3,097	495	5

	60%	11,148,000	92,900

2,323	372	4

	40%	7,432,000	61,933

1,548	248	2

	20%	3,716,000	30,967

774	124	1

	0%	0	0

0	0	0











% Released paint cleaned up







ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	92,900	774

19	3	0

	99%	185,800	1,548

39	6	0

100%	95%	929,000	7,742

194	31	0

	90%	1,858,000	15,483

387	62	1

	80%	3,716,000	30,967

774	124	1











	ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	74,320	619

15	2	0

	99%	148,640	1,239

31	5	0

80%	95%	743,200	6,193

155	25	0

	90%	1,486,400	12,387

310	50	0

	80%	2,972,800	24,773

619	99	1











	ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	55,740	465

12	2	0

	99%	111,480	929

23	4	0

60%	95%	557,400	4,645

116	19	0

	90%	1,114,800	9,290

232	37	0

	80%	2,229,600	18,580

465	74	1











	ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	37,160	310

8	1	0

	99%	74,320	619

15	2	0

40%	95%	371,600	3,097

77	12	0

	90%	743,200	6,193

155	25	0

	80%	1,486,400	12,387

310	50	0











	ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	18,580	155

4	1	0

	99%	37,160	310

8	1	0

20%	95%	185,800	1,548

39	6	0

	90%	371,600	3,097

77	12	0

	80%	743,200	6,193

155	25	0











	ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	0	0

0	0	0

	99%	0	0

0	0	0

0%	95%	0	0

0	0	0

	90%	0	0

0	0	0

	80%	0	0

0	0	0



Relationship between lead on surface and suitability of room









	Assume:







	Paint lead level	1	mg Pb/cm^2





Room Size	120	SF(10x12)





Job Size	4	SF





Constants:







LBP	1	mg Pb/cm2





unit conversions	929	cm^2/SF







1000	ug Pb/mg Pb





EPA dust-lead hazard levels







floor	40	ug Pb/SF





	sill	250	ug Pb/SF























Scenario







	A renovation contractor disturbs 4 square feet of 1 mg/cm^2 LBP while
conducting a repainting job. (Per QA for site selection.)



















Possible Outcomes:







	Actual released area of paint (as % of total)





	Release Levels

Multiples of dust-lead hazard











	ug Pb total	ug Pb/SF

sill	floor



100%	3,716,000	30,967

774	124



80%	2,972,800	24,773

619	99



60%	2,229,600	18,580

465	74



40%	1,486,400	12,387

310	50



20%	743,200	6,193

155	25



0%	0	0

0	0











	% Released paint cleaned up







ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	18,580	155

4	1	0

	99%	37,160	310

8	1	0

100%	95%	185,800	1,548

39	6	0

	90%	371,600	3,097

77	12	0

	80%	743,200	6,193

155	25	0











	ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	14,864	124

3	0	0

	99%	29,728	248

6	1	0

80%	95%	148,640	1,239

31	5	0

	90%	297,280	2,477

62	10	0

	80%	594,560	4,955

124	20	0











	ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	11,148	93

2	0	0

	99%	22,296	186

5	1	0

60%	95%	111,480	929

23	4	0

	90%	222,960	1,858

46	7	0

	80%	445,920	3,716

93	15	0











	ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	7,432	62

2	0	0

	99%	14,864	124

3	0	0

40%	95%	74,320	619

15	2	0

	90%	148,640	1,239

31	5	0

	80%	297,280	2,477

62	10	0











	ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	3,716	31

1	0	0

	99%	7,432	62

2	0	0

20%	95%	37,160	310

8	1	0

	90%	74,320	619

15	2	0

	80%	148,640	1,239

31	5	0











	ug Pb total	ug Pb/SF

floor	sill	sill (1%)

	99.5%	0	0

0	0	0

	99%	0	0

0	0	0

0%	95%	0	0

0	0	0

	90%	0	0

0	0	0

	80%	0	0

0	0	0



Comments from Reviewer P8

Peer Review Comments

The document titled “Quality Assurance Project Plan For Selecting
Sites For Characterization of Dust Lead Levels After RRP Activities”
will hereafter be referred to as “Doc. 1.”  The document titled
“Draft Study Design and Q. A. Project Plan for Characterization of
Dust Lead Levels after RRP Activities” will hereafter be referred to
as “Doc. 2.”

I.	Doc. 1 - Sect. 1.4:

a)	How will a study area be confirmed to be brought back to a baseline
after each phase is replicated?  A full TSP/HEPA vacuum cleanup must be
employed after each phase regardless of the type of work and cleanup
that was performed during each phase.  Pre-phase wipe tests must be
compared to each post-phase set of wipe tests. (Testing addressed in
Doc. 2 – Sect. 2.1.2., but “thorough cleaning” not defined.)

Response:

	A thorough cleaning will be performed by people who have the
credentials and experience in cleaning after lead abatement activities
following each experiment, with no exceptions. If the unit is to be used
for another experiment, then the post-experiment cleaning samples will
be used as the baseline pre-experiment values for the subsequent
experiment.

b) 	If a housing unit is serving as both an interior and exterior site,
to best certify that no cross-contamination will occur, both phases
should not be run concurrently.

Response:

Due to the reasons cited, and personnel requirements, two experiments on
the same unit will not be run concurrently.

II.	Doc. 2 - Table 1-2 and Doc. II – Table 2-1:

Assuming that the two referenced tables contain the sole definitions of
possible RRP jobs, they require further clarification to ensure
commonality and consistency in the various phases and replications. 
Inadequately defining and quantifying the actual tasks being studied
would result in too many variables that would unnecessarily complicate
the comparison of results.  Will a definitive site-specific
specification be written for each work site to clarify these variables? 

Response:

	Detailed information about each experiment (work performed, containment
use, clean-up routine, contractor information, etc.) will be documented
on experiment-specific information forms.  

1.	Low Level Work:

a)	Installation of recessed lighting must also include any necessary
fishing and running of wires.  It is unusual to have lead paint
(“LP”) in ceiling paint.  A more likely situation is that an
electrician will be punching several holes in LP plaster walls to run
wires for the recessed lighting.  An installation of a dummy recessed
light without fully wiring it would not be a valid test.

b)	“Repair window sill etc.” should further define what work is
being performed.  Does the “repair” involve simply scraping loose
paint to prepare for spackling and repainting, or is it an epoxy repair
due to rot or damage?

c)	Removing, planing and reinstalling the door needs to be clarified if
it is to be considered Low Level Work.  This would only be the case if
the top or bottom of the door was being planed.  If the vertical edges
of the door were being planed, a continuous line of lead paint would be
disturbed creating a high level of dust, thereby qualifying as a High
Level job.

d)	“Cut out 2 ft section of wall” is listed only in Table 2-1 (Doc.
II). It needs to be inserted into Table 1-2 (Doc. I). It should be
further clarified to state “LP Plaster.”

e)	“Scrape and repaint railing on stairs” could be broadened to
include “or porch.”  Also, a small quantity (<5 linear ft) should be
defined to differentiate this job from the High Level Work relating to
“Scrape and repaint porch components.”	

f)	Sanding and scraping of 2 sq ft on the exterior is not a sufficient
area to derive any useful data.  Assuming this task is to perform
painting prep only and not the total removal of all paint in the
specified area, depending upon the condition of the paint, an
insignificant amount of paint would be disturbed.  This is a highly
unlikely real world job scenario.  It is recommended to include a larger
area of at least 10 sq ft at a minimum.  This would also move this job
to Medium Level Work.  Scraping of loose exterior paint with no sanding
on an area of at least 10 sq ft may qualify as Low Level work.

2.	Medium Level Work:

a)	 “Replace a window from inside house” needs further
clarification.  It should be stated that the job is sash replacement
involving removing lead painted side stops, cutting ropes and pulleys,
remove parting beads and installing new box-frame replacement windows
into the existing frame.  The housing unit must have rope and pulley
windows and not merely lead painted sashes with aluminum tracks to
derive any meaningful data.  

b)	“Replace 100 sq ft of drywall” should specify to “remove
existing LP plaster walls and install new finished drywall.”

c)	“Replace clapboard siding with Vinyl” is an unrealistic scenario.
 Perhaps just mention removal of clapboards or shingles in a 100 sq ft
area.  Wood siding is basically never removed in order to install vinyl
siding, so the vinyl siding aspect of this job is superfluous.  The lead
risk is in the act of removing the wood siding.

d)	“Scrape and repaint carport ceiling” should clarify if the
carport ceiling is to be wood or metal.  Also the quantity (sq ft)
should be defined. 

3.	High Level Work:

a)	“Gut out a kitchen/bathroom”:  When gutting a kitchen or bath,
great care must be taken to seal off any structural, pipe and wiring
breaches into adjoining (either above, below, or adjacent)	areas,
especially cellars.  In practice, these breaches are hidden behind
finished plaster walls, and unless the perimeter is investigated for
breaches prior to any demolition work, and any breaches sealed with
spray foam, duct tape, etc. significant cross contamination will result.


Also, in order to accomplish 4 phases of the job in the same room, it
should be stated that the demolition will be broken up in quarters (one
wall at a time) with a full cleanup, wipe tests, etc. between each
phase.  That will result in the most valid set of test results since the
same materials would be involved using the four different phase
procedures.  If the Study desires to measure a “full kitchen gut”
for each of the 4 phases, then the four kitchens to be gutted should be
chosen carefully to have similar size and lead content (area and
concentration).

b)	“Sand painted wooden floor” should specify if floor is being
sanded to remove all paint and prepare for application of polyurethane,
or just prepare it for repainting. 

c)	“Sand crown molding, etc.” should specify if molding is being
sanded to degloss and prepare for repainting.

d)	“Repaint fascia boards etc.” should read “Scrape and repaint
fascia boards etc.”

Response:

The list of jobs has been revised to better span the spectrum of jobs
under the scope of the proposed rule.  A definitive site-specific
specification will be written for each work site and job. Where jobs
must be performed across multiple units, care will be taken to select
units that have similar characteristics (lead content, size, etc.)  

III.	Doc. 1 - Sect. 1.5.2:

a)	When analyzing potential sites, care should be taken to find houses
that are not only built in the appropriate time frame, but houses that
contain original building details so they are truly representative of
that era.  For example, many 1920 to 1950 houses with shingles have had
new shingles installed over the old ones.  Often the newer shingles are
LP, but the readings will be much lower than the original shingles. 
Also, many older homes have had newer sashes installed 40 to 50 years
ago with steel tube or aluminum tracks.  These windows will still have
LP on them, but are in no way representative of that era of housing.

Response:

The type of housing in the study will necessarily reflect what is
available at this time.  If we can find housing with original building
details, we will be glad to include such housing in the study.  It is
likely, however, that many older housing units will have already
undergone renovation and replacement of certain components.

b)	It is understood that this Study is not designed to include areas
with carpet, and is primarily to analyze dust control and dispersion in
the various phases of each job.  However, problems and issues specific
to carpeted areas which would result in altering the definition of
“Specialized Cleaning,” must at some point be addressed since a
large percentage of RRP projects will take place in carpeted areas. 
(This issue is brought up in Doc. 2 Pg B-7, however it is significant
enough of an issue to be mentioned here.)

Response:

We recognize the importance of issues related to carpets Due to time and
budget constraints, this study will not include carpeted surfaces, as it
would add an entire dimension to the study design.  

IV.	Doc. 2 – Sect. 1.4:

Determining what qualifies as a statistically significant amount of work
sites and test results is best left to experts in statistics and
sampling.  From an abatement contractor’s perspective, it seems as if
sufficient data should be collected from this study, as long as
variables such as lead concentration, lead quantity, paint and substrate
condition, weather, work practices, etc. are monitored and taken into
account.  Great care should be taken to attempt to work on as
homogeneous a set of housing from the various eras as possible.  

It also is advisable to have the same trained workers performing the
various tasks so the variable of differing work practices can be
eliminated.  Due to time constraints, this may not be practical. 
However, it is important at least to have the same workers do all of the
work on a particular job at a given site for all 4 phases.  Variations
in work practices between individual workers would negate any valid data
if this is not employed.

Response:

In an effort to minimize worker-to-worker variability, it is preferred
that the same RRP contractors perform each of the phases for a
particular job.  

V.	Doc. 2 – Sect. 1.5:

The terms “plastic sheeting is to be placed on the floor…and cover
furniture and other objects in a room” (1.5.2) and “plastic sheeting
placed on the ground” (1.5.5) needs to be clarified.  If the plastic
sheeting is not duct taped or otherwise continuously fastened to the
floor or foundation of the building, in practice a large percentage of
LP debris will reach the floor, furniture, or ground.  Even if the
sheeting is “placed” under optimal conditions to be butting up
against the foundation at the start of the job, it will almost
immediately be pulled away, folded back, or otherwise moved due to
normal motion of the workers walking, moving equipment, ladders, etc. 
It should be noted in the work spec that the sheeting must be
continuously fastened to the foundation, and any breaches in that seal
must immediately be repaired.  Doc. 2 – Sect. 2.1.5.3 partially
addresses the taping down of interior plastic sheeting, but does not
sufficiently define the mil thickness or the method of taping (spot
taping vs. continuous taping) to the perimeter of the floor).

Response:

The plastic must be securely fastened to the ground or wall.  The
detailed protocol for technicians and contractors addresses this issue
and provides guidance on properly securing the plastic.  All plastic
used in the study will be 6 mil, and that is specified in the full study
QAPP.

Also, in the phases where no plastic covering is used, there is no way
to determine how a room’s belongings are affected, since the study
will include only vacant units.  In practice, RRP projects are very
often occurring in fully occupied, fully furnished homes.  The results
of this study should be either applicable to only vacant housing, or
further testing of furnished housing units would need to be undertaken. 
Without this data, a valid option may be to mandate the covering and
sealing of all furnishings regardless of other work practices on RRP
jobs.

Response:

We agree that valuable data could be generated by including furnished
housing in the study. This is impractical for a number of reasons.
First, including furnished housing would greatly increase the sampling
costs, cleaning and clearance requirements and sources of variability in
the study. More importantly, EPA has an ethical and legal responsibility
to protect future occupants from lead exposure due to the study. Using
vacant units provides the greatest opportunity to insure that cleaning
and clearance is adequate to protect human health.  

VI.	Doc. 2 – Sect. 2.1.5.5:

The Post-Work Environmental Samples must include interior window wells
since the work practices in all four phases do not include sealing off
of the exterior of the windows in the work area.  These windows would
certainly be contaminated by Medium and High Level work, and likely also
by Low Level work.  It is recommended that the sealing of the windows
with duct tape and plastic sheeting be mandated for any exterior work.

Response:

During exterior experiments, all windows will be closed and remain
closed throughout the experiment.

VII.	Doc. 2 – Sect. 2.1.5.9:

The procedure outlined in the third paragraph in the “Floors”
section is ambiguous and seems contradictory.  Please clarify.  The
entire procedure seems to be more time consuming, unnecessarily tedious
and prone to error and therefore more costly than a standard
post-abatement cleanup followed by lead dust wipe tests. 

Response:

The cleaning verification procedure outlined is the procedure described
in the proposed rule.  One of the study objectives is to analyze the
effects of such cleaning verifications by testing after wet and dry
cleaning cloth verification steps.  The purpose of such a process is to
provide guidance for property owners and contractors who perform LBP
renovations in a housing unit on how to determine if the unit is free of
lead hazards, without the added (and not required) expense of clearance
samples. 

VIII.	Doc. 2 – Sect. 2.1.5.12: 

Regarding clearance wipe tests at the completion of the study, it is
recommended that window well wipe tests be added to the floor and sill
sampling.  It is common for wells to be contaminated during exterior
work and they could easily be overlooked in the final cleanup, resulting
in an unsafe environment once the housing unit is turned back over to
occupants.

Response:

The current protocol does not involve sampling window wells after
exterior work is performed.  If necessary in certain situations, the
study will clean and clear potentially impacted interior areas following
completion of all exterior work at a site.  For interior work, the
entire unit, including wells will be cleaned and cleared.  

IX.	Doc. 2 – Sect. 1.5.6, Sect. 2.1.5 and Appendix B, Section E
“Cleaning Verification”: 

In practice, post-abatement cleaning of lead dust does not necessarily
result in all grit and soil being removed from the floor, be it wood,
vinyl, or tile.  The contamination resulting from the abatement or
renovation project does not penetrate to the substrate, especially if
the housing unit had a significant layer of soil and grit on the floors
prior to the project. It is not unusual to see a wipe test sample with
gray or brown staining from soot or other non-hazardous contaminants,
yet still easily pass the EPA dust protocols.  Conversely, wipe test
samples that show no obvious discoloration have failed on many
occasions, especially if the LP being disturbed was white and a few
simple particles or flecks of paint adhered to the test wipe.  

Much additional labor might be expended to bring floors into a “white
glove” level of cleanliness.  Also, the workers’ attention may
easily be drawn to cleaning non-hazardous soot and stains which would
cause a “white glove” failure, while overlooking a proper lead
decontamination procedure which would require concentration on particle
elimination under radiators, in corners and crevices, etc.  A standard
lead dust wipe test to confirm EPA clearance levels may be less costly
both in time and money to confirm safe post-cleanup levels.  It could
easily surpass 2 or 3 hours (costing between $100 - $200) of additional
labor per room to bring the floors and sills to a “white glove”
standard that would result in no net increase (and possibly a decrease)
in environmental safety

Relying solely on a visual assessment to determine safe clearance levels
seems to be a less desirable method for two main reasons:  The potential
significant added cost of unnecessary cleaning to attain a “white
glove” pass, and the potential serious conflict of interest in having
a RRP contractor perform this visual inspection on his own work.  Third
party involvement, while appearing to be an added cost and extra burden,
in practice streamlines the process, in most cases saves overall cost,
and insures a better project.  (Note:  this is written by an abatement
contractor who has no vested interest in increasing business prospects
for inspectors.)

Response:

Prior to the conduct of the study at any particular housing unit or COF,
a thorough cleaning will occur to ensure that the settled dust sampling
results on floors and sills are attributable to the RRP activity (and
not prior contamination).  We agree that in practice, the rule will
apply to situations in which the RRP contractor is conducting cleaning
activities that reduce dust-lead levels that were generated by both the
RRP activity and prior contamination – however, addressing this
concern is not within scope for this particular study.

A secondary objective of the study is to assess the white glove method
in the proposed rule.  The accuracy of the method will be assessed, and
information will be collected as to how long it takes to complete it.  

 Comments from Reviewer FR1

General Comments to the Draft Study Design:

Section 2.1.5.3, page 2-11:	Why are general area air samples being
collected for airborne lead in lieu of personal breathing-zone (PBZ) air
samples? The latter would be more reflective of the workers’ actual
exposures and would still provide an indication of the potential lead
dust generated by the specific renovation/repair/painting (RRP) job
being evaluated.

Response:

	Both personal and indoor air samples will be collected for all interior
experiments, and personal air samples will be collected for exterior
experiments.  The personal air samples will not be analyzed with respect
to the direct study objectives; however they will be collected to
address health and safety concerns of the workers.  The indoor air
samples will provide valuable data on the spread of dust from the work
room to the adjacent rooms.

Section 2.1.5.3, page 2-11:	For exterior jobs, how far will the plastic
sheeting extend beyond the exterior of the building to protect the
ground? Federal and state guidelines for renovation of homes with
lead-based paint (LBP) have recommended that plastic sheeting be used to
protect the ground and shrubbery form the base of the walls to a
distance of at least 10 feet. We have suggested that this sheeting may
need to be extended up to 25 feet beyond the base of the walls.

Response:

The rule states that exterior ground covering must extend out “a
sufficient distance” from the structure to maximize the containment of
dust and debris.  The containment needs will be assessed according to
the language in the rule.  Study technicians will prepare both ground
and vertical containment beyond that required by the rule to protect the
property and neighboring properties.  

Section 2.1.5.5, page 2-12:	Your minimum detectable concentration (MDC)
and minimum quantifiable concentration (MQC) for your airborne lead
samples will be directly influenced by the sampling times. This could
present a problem if the airborne lead concentrations are relatively
low, and your sampling time brief, resulting in higher MDCs and MQCs.
Please note that NIOSH defines the MDC and the MQC as the analytical
method’s Limit of Detection (LOD) or Limit of Quantification (LOQ)
divided by the air volume for the sample. This same comment applies to
Section 2.1.5.8.

Response:

	We understand the concerns regarding the air sampling results, and all
results will be analyzed in a consistent manner, such that the samples
provide accurate and acceptable data. 

Section 2.1.5.5, page 2-12:	Clarify whether these will be wet or dry
wipe samples.

Response:

All of the dust wipe samples are to be taken with wet dust wipes
provided by the laboratory.

Section 2.1.5.9, page 2-14:	Although I believe you intend that the wipe
sampling on the interior windowsills will be performed by a field
technician, I would suggest clarifying (as you did for the wipe samples
for the floors).

Response:

Dust wipe samples will be taken between wet and dry cloth verification,
as necessary, by the field technician.  Note that all dust wipe samples
(excluding pre- and post-experiment clearance samples) are to be taken
by the technicians and all cleaning and verification activities are to
be performed by the contractors.

Section 2.1.5.12, page 2-17:	You state that floor and windowsill wipe
samples will be collected in all rooms of the housing unit to determine
if dust lead levels are below clearance levels in all rooms. However, in
Section 2.1.5.2, page 2-10, you indicate that you will clean only those
areas impacted by the study. My question is how the post-study wipe
sample results will be interpreted since some of these areas were not
evaluated prior to the study, so the lead contamination may have already
been present and not as a result of the study activities.

Response:  

	Prior to the initial cleaning, plastic sheeting will be used to close
off rooms and parts of the house that are not part of the study area or
needed for other purposes related to the study.  Those areas not closed
off will undergo a thorough cleaning.  The parts not closed off will be
cleaned between phases of an interior job.  When the study work is
completed at a unit, the entire unit will be cleaned. All clearance
testing will be performed by certified risk assessors/clearance
technicians according to all current standards. 

Section 2.2.2, page 2-18:	I would recommend using disposable templates
for collecting wipe samples.

Response:

Due to the number of dust wipe samples estimated to be collected
(approximately 2,000), the study will not use disposable templates.  The
template will be cleaned with a wet wipe or cloth between each use to
ensure no cross-contamination.

Section 2.2.3, page 2-19:	Putting on clean disposable gloves to mount a
sampling cassette is not necessary.

Response:  

	Field technicians will don clean, disposable gloves for all sampling
activities to avoid contamination of the samples or the sampling media. 

Section 2.7, page 2-23:	You need to include a requirement for air
sampling pump calibration, using a primary calibration standard.

Response:  

	The technicians will perform standard calibrations of the pump
according to the current standards and in accordance with NIOSH 7082.

Comments from Reviewer FR2

ADDITIONAL COMMENTS

As a general observation, clarification of the term “normal
cleaning” is needed.  The use of the word, normal, is indicative of
these practices being considered acceptable by the regulatory community.
  It is not normally acceptable to use dry sweeping, brushing and
regular vacuuming to clean areas where employees could be exposed to
lead.  It is clear that some historic preservation requirements may
limit employee’s abilities to utilize wet methods, or other
engineering controls, but we would expect that to not be a “normal”
practice.

If your use of the term “normal” in this study is indicative of
actual procedures currently being used by the renovation, repair and
painting industries, then it would be even more important to clarify
this language in order to prevent the assumption that the use of dry
methods for cleaning is the acceptable practice.  OSHA’s standard for
lead in construction, 29 CFR 1926.62(h) prohibits, for the most part,
cleaning and housekeeping practices utilizing dry methods, regular
vacuuming and the use of compressed air.  

It is understandable that the use of these types of procedures is
important for the collection of exposure data during the study, but
using dry sweeping, brushing and regular vacuuming during actual
renovation operations would greatly increase the likelihood of excessive
employee exposures and should not be allowed for use during any regular
renovation operations.  Additional precautions need to be taken to
assure that all parties participating in the study understand the
difference between procedures to be used for the purposes of research
only and those that would be considered acceptable under normal
operations (“rule cleaning”).

Response:

The two cleaning methods are now referred to as “rule cleaning” and
“baseline cleaning” to reflect that the majority of contractors are
not adhering to or are not aware of OSHA’s Lead in Construction
Standard (29 CFR 1926.62) and are instead broom sweeping or using a
regular vacuum on floors that aren't carpeted. As designed, the study is
intended to compare the use of the rule cleaning as compared to the
baseline cleaning that is common among contractors.  The workers will be
wearing full protective suits made of Tyvek or other suitable material
and half or full mask respirators to protect them from exposure. 

Comments from Reviewer FR3

ADDITIONAL COMMENTS

The study designs sometimes refer to OSHA safety and health requirements
as guidelines.  OSHA standards are requirements/mandates; employers have
to implement them when applicable.   

Response:

It is understood that OSHA safety and health requirements are just that,
requirements.  The study design reflects that, and it is clearly
communicated in the Health and Safety Plan, the full study QAPP, and the
provided training. 

Comments from Reviewer FR4

ASTM Standards

While a couple of ASTM standards (albeit outdated) are referenced in the
quality assurance plan, there are numerous others that should be put
into use in this study.  For instance, there are several sampling and
analytical activities for which relevant standards are not cited, and
this shortcoming should be addressed.  Consideration should also be
given to the use of relevant standards pertaining to screening,
clearance, inspections, hazard assessments, etc. 

Response:

The current ASTM standards for sample collection that are applicable to
the study are included in appendix E of the full study QAPP. They
include ASTM E1727-05 for soil samples, ASTM E1728-03 for dust wipe
samples, and ASTM E1729-05 for paint chip samples.

CDC/NIOSH Research

Researchers in this agency have conducted numerous investigations on
lead hazard reduction activities associated with renovation, remodeling
and similar work activities, but these studies appear to be overlooked. 


Response:

A literature search on topics related to this study was conducted prior
to the publication of the proposed rule and NIOSH was contacted at that
time regarding its relevant research.  The study is being conducted to
provide comparative data not available in the literature.

Comments from Reviewer FR5

Peer review comments on May 23, 2006, draft study design and quality
assurance project plan for characterization of dust lead levels after
renovation, repair, and painting activities.

Section: 

1.5:  Overall:  

	Although the study design is simplified by having a relatively narrow
set of construction types, a single urban area, and a small number of
housing units (HUs) and child occupied facilities (COFs), using a larger
number of widely located housing units and COFs of widely varying
construction types would make the study more robust and overcome the
limited variability of the building sample set.

	The power calculations for several of the study questions are shown for
the housing units, but not for the COFs.  The COF power calculations
should be shown, and the number of COFs adjusted as appropriate.	

Response:

The addition of housing units and COFs is not possible within our budget
constraints.   An increased number of samples are to be collected in
order to overcome some of the limited variability due to the number of
housing units.

The question of whether data from the COFs will be modeled separately
from the residential data (or combined with the residential data in a
single global model) has not yet been determined.  Although the
activities planned for the COFs overlaps with the activities planned for
housing, ability to model together will likely depend on any important
similarities or differences in the results revealed within the
exploratory statistical analyses.

When EPA initiated power studies for the COFs – it was determined that
the assumptions that were going into the power studies for COFs were
identical to those that were being used for the residential units –
which led us to the presumption that the data could be combined in
subsequent data analyses.

Ultimately, it would be desirable to combine the data across the two
building types – which would provide additional statistical power for
the study – however, EPA was conservative in providing power study
results only for the subset of data that would be collected in
residential housing.  Obviously, power analyses for data analysis
objectives that consider COFs on their own would yield significantly
less power than those provided in the appendix (as less data is being
collected in COFs than residential units).

1.5.1	Clarify in the question that the "plastic" here is to cover the
floor.

1.5.2	Clarify in the question that the "plastic sheeting" here is to
provide airlocks.  The key study question element is if airlocks are
used, not the material from which they will be constructed.

	With the number of HUs and COF so small, it appears that the units to
be evaluated for study questions #1 and #2 are being double-counted. If
this is not so, then table 2-5 would indicate work area rooms that are
enclosed by airlocks and those that are not; no distinction appears to
be made.  At the least, the confounding of the airlock factor needs to
be addressed statistically.

	The description of airlock-enclosed work rooms does not match Figure
2-1 or study question 4.

Response:

The plastic sheeting and plastic sheeting airlocks are both components
of the containment proposed in the rule.  They are not being evaluated
separately.  

1.5.3	While it is appropriate to distinguish between "standard cleaning
methods" and the proposed rule's cleaning method, the standard interim
control cleaning method should be included in the design, in order to
provide a clearer measure of inter-method variability in the context of
existing cleaning methods, especially because earlier studies have shown
significant replicate variability.  The design should be revised
accordingly.

Response:

The study is being done to provide comparative data for risk assessment
and cost-benefit analysis for the RRP rule.  The types of cleaning have
been chosen to provide requested data for those two analyses.  

1.5.4	Whether rooms are contained by airlocks of not is unstated; this
cofactor is discussed in study question 2, but not here.  The design
should be revised accordingly.

	The power calculation for the revised design should be included; it was
not for the draft design.

Response:

The plastic sheeting and plastic sheeting airlocks are both components
of the containment proposed in the rule.  Where the study design says
“use of plastic”, this includes plastic sheeting airlocks.

1.5.5	That plastic ground covering does not allow dust to penetrate is
well-known and should not be the focus of this study question.  It
should address whether rolling up plastic after the work can be done
cleanly, and the spatial distribution of the dust generated by this
exterior work (both on and off the plastic ground covering).

	Soil lead concentrations should be determined both before and after the
work, and used to determine soil contamination from the work.  The data
requirements discuss soil samples, but the study question addresses just
dust.

	The power calculation for the revised design should be included; it was
not for the draft design.  

Response:

This aspect of the study has been requested for the risk assessment, and
the study is designed to assess the lead levels after removal of the
plastic.  Soil samples will be collected both before and after RRP
activities to determine the soil contamination, if any, resulting from
the work.  The study question will evaluate the effectiveness of the
containment, which includes removal after work is complete.

1.5.6	Dust lead samples should be collected before the HU or COF is
cleaned, with the sampled analyzed only after the field work in the home
is completed (to avoid experimenters' bias), and the results included as
a cofactor in the analysis.  The HUD National Evaluation of its lead
hazard control grant program indicated that the initial conditions had a
significant relationship to post-clearance conditions, especially for
window troughs.

	The study design should compare the two "cleaning verification" steps
with the standard post-abatement (and post-HUD-assisted interim control)
cleaning steps, as well as with each other.

	The study design should address what will be done if the area does not
pass the cloth observation check after four cleanings.  The study design
does not indicate that the HUs and COFs will be demolished after this
project, so it must be presumed that they will be reoccupied.  In the
case of the pre-1978 target HUs, lead-based paint and lead-based paint
hazard information (including information on lead hazard evaluation and
control activities) should be provided to the owner so that the owner
can disclose the information to future buyers or renters.  A similar
action should be undertaken for non-target HUs and the COFs, in order to
provide occupants with valuable lead information (and reduce EPA
liability).

	The two cleaning types are to be tested, but the order is not shown as
being varied, so the order and type will be inextricably confounded. 
(The first cleaning of whatever type is expected to clean more than the
second cleaning of whatever type.)  The design should be revised
accordingly.

	The power calculation for the revised design should be included; it was
not for the draft design.

Response:    

The cleaning and clearance of the unit before RRP work begins is
performed to ensure that all experiments begin with a clean unit, for
comparison purposes.  We believe having clearance samples to verify the
cleaning that preceeds the study work will be adequate for the study
goals.  

The results of the XRF testing, paint chip, and final clearance results
will be provided to property owners for their records and for disclosure
to future occupants.  

The study will focus on comparison of the specialized cleaning in the
proposed rule in comparison to a “baseline” cleaning which we
believe is typically used by renovation contractors.   The methods are
varied in the phases of the jobs, which will be randomly assigned to the
RRP contractors.  

1.6	Individuals taking dust samples or soil samples for the main data
analyses should be experienced certified lead risk assessors, just as
those taking x-ray fluorescence (XRF) measurements should be experienced
certified lead inspectors.

	To test the effect of qualifications on cleaning verification, cleaning
verification by non-risk assessor work supervisors (to emulate the
proposed rule's procedures) should be conducted in parallel to cleaning
verification by experienced certified lead risk assessors (to obtain
nominally higher quality data).

Response:

XRF testing will be performed by certified inspectors.  Similarly, the
pre- and post-experiment clearance samples will be collected by a
certified risk assessor or clearance technician.  The study-related dust
wipes will be collected by a field technician who is trained in the dust
wipe collection process.  The proposed rule calls for the cleaning
verification to be performed by the person performing the renovation
work, in this study, that is the contractors.  The cleaning verification
performed by the contractors will be assessed by dust wipe samples taken
by technicians following both wet and dry verifications.

1.7	Qualifications and experience of persons conducting measurements,
RRP work, and/or cleaning verification should be recorded.

Response:   

This information will be recorded.

2.1.1	Tables 2-1 and 2-2 are inconsistent with the draft study siting
report's table 1-2; this may be a function of date of generation, but
should be resolved.

Response:

The job list underwent numerous changes throughout the study design
process.  The job list is now consistent between the documents.

2.1.2	As noted for section 1.5.2, the design should address the presence
or absence of airlocks.  The list of protection/clean-up routines
addresses just floor coverings.  Airlock confoundment in the design
should be resolved.

	As noted for section 1.5.3, the design should also include standard
interim control cleaning methods.

	Tables 2-3 and 2-4 should be revised based on the comments for section
1.5 and subsections.

Response:

The plastic sheeting and plastic sheeting airlocks are both components
of the containment proposed in the rule.  Where the study design says
“use of plastic”, this includes plastic sheeting airlocks.  

2.1.4	Revise based on the comments for section 1.5 and subsections.

Response:   

Subsection will be consistent with other sections and subsections.

2.1.5.1	As noted for section 1.5.6, dust lead samples should be
collected before the HU or COF is cleaned.

Response:   

We believe the clearance samples to check on cleaning are adequate to
meet study goals.

2.1.5.2	Mention the preliminary dust lead sampling of section 2.1.5.1 in
order to ensure that supervisors of the cleaning work of section 2.1.5.2
know about it before the unit is cleaned.

	Reconsider whether TSP should be used, based on experimental findings.

Response:   

	We will consult with the cleaning firms as far as what detergent will
be used for the pre- and post- experiment cleanings.  The same detergent
will be used across all experiments.

2.1.5.3	Address neighborhood residents' predominant language; use
additional languages beyond English as appropriate.

Response:

This will be taken into account and, if deemed necessary by the site
supervisor, alternate languages will be included on the warning signs.

Address XRF usage, including mandatory operation of the XRF in
accordance with its HUD XRF Performance Characteristic Sheet (PCS).

Response:   

A certified XRF technician will perform the testing, and applicable
state regulations will be followed.  This will include following the
Performance Characteristics Sheets.

Appendices (and corresponding entries in table of contents):

C:	Delete and continue to provide citation in main text.  Inclusion here
simply lengthens the document.

E:	Delete HUD Guidelines for the Evaluation and Control of Lead-Based
Paint Hazards in Housing Appendix 13.3, and continue to provide citation
in main text.  Inclusion here simply lengthens the document.

F:	Delete and continue to provide citation in main text.  Inclusion here
simply lengthens the document. 

G:	Delete and continue to provide citation in main text.  Inclusion here
simply lengthens the document.

Note that deleting the cited photocopied pages in the appendices will
shorten the design document by about half, a significant benefit.

Response:

While these sections could be replaced by a shorter protocol attuned to
the study needs – having this documentation available within the QAPP
for the field supervisor and others performing the study is often
helpful for ensuring that all of the study protocols are  followed
appropriately.

Peer review comments on April 19, 2006, draft quality assurance project
plan for selecting sites for characterization of dust lead levels after
renovation, repair, and painting activities.

Section: 

1.4	Although the study design is simplified by having a relatively
narrow set of construction types, a single urban area, and a small
number of housing units (HUs) and child occupied facilities (COFs),
using a larger number of widely located housing units and COFs of widely
varying construction types would make the study more robust.

	Table 1-2, col. 1, row 1, bullet 5:  This replicates bullet 4.

	Table 1-2 is inconsistent with the draft study design's Tables 2-1 and
2-2; this may be a function of date of generation, but should be
resolved.

Response:

Experiments will be conducted in multiple cities to increase the
robustness of the study, however adding more housing units or COFs is
not possible due to time and budget constraints. The job list underwent
numerous changes throughout the study design process.  The job list is
now consistent between the documents.

1.5.2	Restricting the housing units (HUs) and child occupied facilities
(COFs) to only those "in a reasonably cleanable condition" (in bullet 3)
restricts the relevance of the study to the renovation, repair, and
painting (RRP) rulemaking.  The proposed rule (71 Federal Register
1587-1636, January 10, 2006) does not restrict the proposed renovation,
repair, and painting (RRP) and cleaning verification activities to such
HUs.  For comparison, the HUD National Evaluation of its lead hazard
control grant program indicated that the initial conditions had a
significant relationship to post-clearance conditions, especially for
window troughs.

Response:

The purpose of this study, is to evaluate the effectiveness of the
proposed rule on contamination of a housing unit. Initially cleaning the
unit helps insure that baseline measurements for comparison are
accurate, and that the subsequent leaded dust found in the unit is the
result of the RRP work conducted for the study. 

1.5.3	Site selection should require that the surfaces to be worked on be
determined to be coated with lead-based paint (LBP).  As drafted, some
of the surfaces to be worked on may be non-LBP as long as a sufficient
area of other surfaces do.  Work on non-LBP surfaces would be of
interest as part of an additional internal control set, but should not
be part of the central design set, especially given the very small
number of HUs and COFs proposed.

Response:

The study is intended to conduct jobs on components with lead-based
paint, as determined by XRF and paint chip sample results.    

1.5.4	Adherence to the HUD XRF Performance Characteristic Sheet (PCS)
must be required for 100% of all paint measurements taken by portable
XRF.  A lesser requirement, such as a percentage under 100%, or use of
supposedly "similar specifications" will render much of the study
unnecessarily suspect.

	A limit on the extrema for batch quality control samples outside of the
control range should be considered.

Response:

There are typically deviations for practical reasons and it is felt that
95% is sufficient for this study.

1.6	Individuals taking dust samples or soil samples for the main data
analyses should be experienced certified lead risk assessors, just as
those taking x-ray fluorescence (XRF) measurements should be experienced
certified lead inspectors.

	To test the effect of qualifications on cleaning verification, cleaning
verification by non-risk assessor work supervisors (to emulate the
proposed rule's procedures) should be conducted in parallel to cleaning
verification by experienced certified lead risk assessors (to obtain
nominally higher quality data).

Response:

XRF testing will be performed by certified inspectors.  Similarly, the
pre- and post-experiment clearance samples will be collected by a
certified risk assessor or clearance technician.  The study-related dust
wipes will be collected by a field technician who is trained in the dust
wipe collection process.  The proposed rule calls for the cleaning
verification to be performed by the person performing the renovation
work, in this study, that is the contractors.   The cleaning
verification performed by the contractors will be assessed by dust wipe
samples taken by technicians following both wet and dry verifications.

2.1.1	Bullet 1: See comment on section 1.5.3, above.

	Bullet 3:  The study design does not indicate that the HUs and COFs
will be demolished after this project, so it must be presumed that they
will be reoccupied.  In the case of the pre-1978 target HUs, lead-based
paint and lead-based paint hazard information (including information on
lead hazard evaluation and control activities) should be provided to the
owner so that the owner can disclose the information to future buyers or
renters.  A similar action should be undertaken for non-target HUs and
the COFs, in order to provide occupants with valuable lead information
(and reduce EPA liability).

	Bullet 5: See comment on section 1.5.2, above.

	

2.1.2	Although the study design is simplified by having a relatively
narrow set of construction types, a single urban area, and a small
number of housing units (HUs) and child occupied facilities (COFs) — 3
and 1, respectively — using a larger number of widely located housing
units and COFs of widely varying construction types would make the study
more robust and overcome the limited variability of the building sample
set.

Response:

Experiments will be conducted in multiple cities to increase the
robustness of the study, however adding more housing units or COFs is
not possible due to time and budget constraints.  An increased number of
samples are to be collected in order to overcome some of the limited
variability due to the number of housing units.

2.2.2	Define "a full XRF lead screening inspection."

	Revise the requirement so that the inspection "shall" use the PCS, and
delete the phrase that it is to be used "as a guideline."

Response:

The XRF lead screening inspection will include a measurement on every
unique interior and exterior component of the unit, as well as duplicate
measurements for large components (floors, ceilings, walls, etc.)  We
will follow applicable state regulations, and the PCSs will be followed.

Update table 3-1, and the April 11th date.

Response:  

The table has been updated.

Appendices (and corresponding entries in table of contents):

A:	Delete and provide correct citation in main text.  The current
standard is E1729-05 Standard Practice for Field Collection of Dried
Paint Samples for Subsequent Lead Determination; the standard in the
appendix, dated 1999, has expired. Inclusion of the full text here
simply lengthens the document, and there may be a copyright problem if
copying fees have not been paid.

B:	Delete and continue to provide citation in main text.  Inclusion here
simply lengthens the document.

D:	Delete and continue to provide citation in main text.  Inclusion here
simply lengthens the document.

Note that deleting the cited photocopied pages in the appendices will
shorten the design document by about three-quarters, a significant
benefit.

Response:

While these sections could be replaced by a shorter protocol attuned to
the study needs – having this documentation available within the QAPP
for the field supervisor and others performing the study is often
helpful for ensuring that all of the study protocols are  followed
appropriately.

 HUD [1985]. Guidelines for the evaluation and control of lead-based
paint hazards in housing. Washington DC: U.S. Department of Housing and
Urban Development, Office of Lead Hazard Control.

 NIOSH [2002]. Hazard evaluation and technical assistance report: Lead
Safe Services, Inc., Neenah, Wisconsin. Cincinnati, OH: U.S. Department
of Health and Human Services, Centers for Disease Control and
Prevention, National Institute for Occupational Safety and Health, NIOSH
Report No. HETA 99-0305-2878.

Page   PAGE  59  of 59

 

 

 

.

/

0

 

 

ƒ

ß

S

°

␃愃̤摧槼÷܀␃愁Ĥ摧ᚗ´ᨀ0

1

M

N

O

P

a

b

c

}

~



€

‚

ƒ

„

…

¡

¢

£

¤

½

¾

¿

Ù

Ú

Û

Ü

Ý

Þ

ß

à

á

ý

þ

ÿࡵᘑ楨酆洀H渄H甄Ĉᔛ❨䉁ᘀ楨酆　ቊ洀H渄H甄Ĉ⌀þ

ÿ

 

 

0

1

2

L

M

N

P

Q

R

S

T

U

q

r

 r

s

t

Ž

©

ª

«



®

¯

°

±

²

Î

Ï

Ð

Ñ

ë

ì

í

j;

␃愃̤摧ᨗO

␃༃킄市킄愂̤摧䑾+

̤̀☊଀Ն愀̤摧䑾+

␃༃킄ᄂや廽킄怂や懽̤摧ᨗO

␃ഃ׆Āɶༀ킄ᄂや廽킄怂や懽̤摧᫔F

␃愃̤摧䑾+

␃༃梄币梄愁̤摧᫔F

̤̀☊଀੆愀̤摧䑾+

␃༃킄市킄愂̤摧᫔F

̤̀☊଀െ愀̤摧䑾+

␃愃̤摧䑾+

␃༃梄币梄愁̤摧䑾+

␃愃̤摧䑾+

␃༃킄市킄愂̤摧䑾+

␃༃梄币梄愁̤摧䑾+

␃愃̤摧䑾+

懿̤摧䑾+

h§

kd”

hã

*

*

+

—

␃愃̤摧䑾+ጀ*

+

,

@

A

@

䀆

␃愃̤摧䑾+᠀

䀂᠋$愂̤摧ᷜÏ

␃ԃĤ␆愁̤摧杹à

␃愃̤摧ᷜÏ

䀂᠋$༂킄市킄愂̤摧ᷜÏ଀-␃༃킄市킄愂̤摧ᷜÏ

퀂᠂$༂킄ᄂや廽킄怂や懽̤摧ᔜ©

h‚b

&

h,

 h,

h,

h,

␃愃̤摧㹹#

܀"␃愂Ȥ摧搉ð਀August 18, 2006

