United States.
Environmental Protection
Agency
Environmental Monitoring Systems
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA/600/S4-85/057 Jan. 1986
Project Summary
Preliminary Evaluation of a
Modified NBS PMio Sampler for
Indoor Paniculate
Measurements
J. E. Howes, Jr., R. Vijayakumar, F. S. Howard, F. Doerfler, and F. J. Burmann
The full report describes modifica-
tions to the prototype NBS portable
ambient air sampler and results of a
preliminary performance evaluation of
the modified unit. The study also in-
cluded a limited performance evaluation
of the indoor particulate sampler devel-
oped by Harvard University.
A prototype portable sampler was
designed and constructed for the U.S.
Environmental Protection Agency (EPA)
by the National Bureau of Standards
(NBS) to measure the mass concentra-
tion of inhalable particulates in indoor
environments. The prototype was eval-
uated in field applications and labora-
tory tests and found to have several
design features that needed improve-
ment. In this study, modifications were
made to overcome problems identified
in the NBS design.
A series of collocated sampling exper-
iments were performed in an indoor
environment to compare results ob-
tained with the modified NBS and the
Harvard samplers with data obtained
with a conventional ambient-type di-
chotomous sampler. These experiments
indicated that over the mass concentra-
tion range studied (~8 -14 /jg/m3). the
modified NBS and the Harvard samplers
yielded PMio mass concentration data
comparable to the dichotomous sam-
pler. However, the NBS sampler did not
accurately classify particles in the fine
(< 3 £
-------
sored development of a portable PMi0
sampler by the National Bureau of Stand-
ards (NBS). The design criteria adopted by
NBS were as follows:
• The sampler should have well defined,
stable cut points with respect to par-
ticle size.
• The samples (as with the standard EPA
area samplers) should be collected on
filters for subsequent weighing and
chemical analysis.
• The sampler should have as high a
sampling flow rate as possible since
ambient particle concentrations are
typically low and a >100 //g particle
sample per filter is necessary for a
gravimetric measurement uncertainty
of <10%.
• X-ray fluorescence, a widely used
method for elemental analysis of filter
samples, requires the sample be in an
even, homogeneous layer on the filter.
• To be acceptable to volunteers partic-
ipating in exposure studies, the sam-
pler must be light, quiet, and incon-
spicuous.
• To be able to sample at high flow rates
for long time periods and still be small
and lightweight, the sampler must
have energy-efficient components.
This means that the sampler air path
must have a low pressure drop and
that the motor-pump unit have low
internal losses.
Several prototype NBS samplers were
constructed and evaluated by Research
Triangle Institute (RTI) in previous field
studies. The work described in the full
report was undertaken to improve several
weaknesses in the design of the prototype
units identified by RTI in the earlier work.
In addition, performance data were also
obtained on the indoor sampler designed
for the Harvard Multicities Study.
Sampler Modification and
Evaluation
Modifications were made in the proto-
type NBS portable sampler to overcome
several design flaws. The modifications
included: (1) replacement of the sampler
case which was highly susceptible to
leakage, (2) redesign of the filter holder
which was inaccessible and susceptible
to leakage, and (3) addition of a motor
control circuit to stabilize the flow rate
and extend the motor lifetime.
A brief series of experiments were
performed in an indoor environment to
compare results obtained with the mod-
ified NBS samplers with those obtained
with collocated dichotomous samplers.
The comparison showed that the NBS
sampler yielded PMio mass concentra-
tion measurements that were in reason-
able agreement with the corresponding
dichotomous data. The mean difference
between the NBS and dichotomous PM10
measurements was approximately 9 per-
cent. The differences in NBS and dichot-
omous results for individual experiments
ranged from approximately 2 to -36
percent. The greatest difference was
observed during the period with the
lowest mass concentration (~ 5 /ug/m3)
and, thus, the largest weighing error. The
precision of paired measurements with
the NBS sampler was excellent; differ-
ences between duplicate measurements
was <9 percent. The mean difference
between duplicate dichotomous sampler
determinations was approximately 5 per-
cent.
The mass of the fine and coarse particle
fractions determined with the NBS sam-
plers did not agree well with measure-
ments made with collocated dichotomous
samplers. The NBS sampler which uses a
filter pack for separation of the size
fractions, yielded significantly higher
values than the dichotomous sampler for
the coarse size range (~ 3.5 to 10 //m).
The difference is attributed to the inability
of the filter pack to accurately separate
particles on the basis of aerodynamic
diameter.
The experience with the NBS sampler
in the performance tests indicated the
most serious problem with the unit is the
short pump/motor lifetime (<100 hours
for some units) and the unfavorable flow
rate—pressure drop relationship at the
operating flow rate. The premature failure
and unfavorable operating characteristics
are attributed to operation of the pump at
over twice its design flow rate in the NBS
design.
The battery pack (six 1.25V Ni-Cd D
cells) used with the modified NBS sampler
provided at least 24 hours of operation at
a flow rate of 6 liters per minute (Ipm). The
constant current motor control circuit
added to the sampler provided a very
uniform flow rate (within ±5 percent)
over a 24-hour sampling period. The
circuit also automatically decouples the
motor when the battery output drops
below a voltage that delivers a flow rate of
6 Ipm.
The performance of an indoor air
sampler developed by Harvard University
was also evaluated in this study. The
mean difference between results obtained
with collocated pairs of the Harvard and
conventional dichotomous samplers dur-
ing nine days of sampling was less than 2
percent. The range of differences in PM10
mass concentration for the nine compari-
sons was 13.7 to -14.3 percent. The
mean of differences between duplicate
measurements with the Harvard samplers
was 11 percent. The results obtained
with the Harvard sampler exhibit excellent
agreement with the dichotomous data
considering the weighing errors associ-
ated with the measurements and the
possible inhomogeneity of particles in
test atmosphere.
Conclusions
The conclusions drawn from results of
this study may be summarized as follows:
• This work and a previous field evalua-
tion have disclosed several flaws in
the design of the prototype NBS port-
able particulate sampler. The problems
that have been identified include (a)
potentially inaccurate sampling rates
due to susceptibility of the sampler
case to leakage, (b) susceptibility of the
filter cassette to leakage and accidental
disassembly if jarred or bumped, (c)
premature failure of the sampling
pump unit, and (d) possible movement
of the inlet components (funnel insert
and oil-soaked frit) thus changing the
inlet cut point. Other features which
may affect user acceptance of the NBS
prototype sampler are the difficulty in
changing filters due to inaccessibility
of the filter cassette, poor weight
distribution, and unsuitability of the
sampler configuration for personal
monitoring applications.
• A modified version of the NBS sampler
was assembled in this study with the
objective of eliminating the initial
problems in the NBS design. A limited
evaluation of the modified sampler has
indicated that the weakest component
in the system is the sampling pump.
The Sensidyne (formerly Bendix) pump
assembly used in the NBS design and
the modified sampler has an unac-
ceptably short lifetime (at least in this
application); one new assembly failed
after less than 100 hours of operation.
In this study all sampler failures were
attributed to the motor rather than the
pump. The pump also has marginally
acceptable flow characteristics for this
application. The flow rate is very
sensitive to pressure changes on the
fore and backside of the pump. This
poses a problem in performing flow
calibrations on the sampler.
-------
• Performance evaluation of the original
design by NBS and evaluation of the
modified version developed during this
study shows that the sampler yields
PMio measurements (i.e. total mass
concentration <10 fjm) which is in
reasonable agreement with collocated
dichotomous sampler data. However,
the results of this study suggest that
there is some doubt as to the ability
and usefulness of the NBS sampler to
resolve the distribution of particles
within the 0 to 10 //m aerodynamic
equivalent size range.
• The NBS and the modified systems
have demonstrated that battery life-
time will provide at least 24 hours of
sampler operation. The motor control
circuitry incorporated in the modified
sampler will provide more uniform
pump performance and longer sampler
lifetime.
• Both the original and modified version
of the NBS sampler use two filters; an
Apiezon-coated Nuclepore* filter for
collection of the ~3 to 10 //m size
fraction and a Teflon filter for the 0 to
~3 fjm size range. The errors associ-
ated with these respective measure-
ments must be assessed. The results
of this study indicate that resolution of
the fine and coarse particle fraction
may be impractical andthat the general
application of the NBS sampler should
be directed toward determination of
the mass of the particle size fraction
< 10 /um in aerodynamic equivalent
diameter.
• In this study, the average of mass
concentration measurements deter-
mined by collocated dichotomous and
Harvard samplers agreed within 2
percent based on tests conducted over
nine sampling periods (of at least 24
hours in duration).
Recommendations
The preliminary evaluation of the NBS
sampler and the modified, portable ver-
sion developed in this study suggests the
following recommendations:
• The most crucial issue in developing a
field-worthy sampler is the pump
reliability. Either the Sensidyne unit
must be upgraded to provide reliable
performance or an alternate pump
must be identified.
• The resolution of the fine and coarse
particle fractions using the Nuclepore
filter does not appear practical. It is
recommended for routine applications
that a single filter (2 or 3 fjm pore size
Teflon with polyolefin support ring) be
used. This will also simplify the use of
the sampler and permit a further
reduction in size.
• When final design criteria are set and
reliable components are found, a per-
sonal version of the sampler can be
constructed. It is recommended that
configuration of the personal sampler
should consist of an inlet that can be
clipped on an individual's lapel or shirt
pocket connected by a flexible line to a
pump unit either worn on a belt or in
soft case, e.g., camera bag with a
shoulder strap. Consideration should
be given to further size reduction of the
inlet assembly so that it is relatively
unobtrusive when clipped to a lapel or
pocket.
It may also be useful to explore sampler
modifications for alternate flow rates
rather than 6 Ipm. Then, it would be
possible to develop a personal sampler
based on the availability of pumps and
motor rather than being forced to select
hardware from a limited choice that meets
the design requirements.
•Mention of trademarks or commercial products does
not constitute endorsement or recommendation for
use by the U.S. Environmental Protection Agency.
-------
J. £". Howes, Jr., R. Vijayakumar, F. Doerfler, and F. J. Burmann are with
Environmental Monitoring and Services, Inc., Newbury Park, CA 91320; and F.
S. Howard is with Esoteric Systems, Inc., Newbury Park, CA 91320.
Robert Jungers is the EPA Project Officer (see below).
The complete report, entitled "Preliminary Evaluation of a Modified NBS PMio
Sampler for Indoor Paniculate Measurements." (Order No. PB 85-24 7 005/A S;
Cost: $9.95, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
A
*.
'
------- |