OCR error (C:\Conversion\JobRoot\00000CD5\tiff\20013I2H.tif): Unspecified error
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are'
1. Environmental Health Effects Research
2 Environmental Protection Technology
3. Ecological Research
4 Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects, assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
AN EVALUATION OF PERSONAL SAMPLING PUMPS IN
SUB-ZERO TEMPERATURES
Carl D. Parker
Martin B. Lee
Joan C. Sharpe
Research Triangle Institute
Research Triangle Park, N.C. 27709
Contract No. 210-74-0124
RTI Project No. 43U-1271
U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
Public Health Service
Center for Disease Control
National Institute for Occupational Safety and Health
Division of Physical Sciences and Engineering
December 1977
-------
DISCLAIMER
The contents of this report are reproduced herein as received from
the contractor.
k!
The opinions, findings, and conclusions expressed herein are not
necessarily those of the National Institute for Occupational Safety
and Health, nor does mention of company names or products constitute
endorsement by the National Institute for Occupational Safety and
Health.
NIOSH Project Officer: Charles S. McCammon
Partial funding of this project was provided
by the Environmental Protection Agency under
The Energy/Environment R and D Program.
DHEW (NIOSH) Publication No. 78-117
-------
ABSTRACT
Personal sampling pumps suitable for industrial hygiene surveys were evaluated
to discover their characteristics as a function of temperature for tempera-
tures between 25° and -5U°C. The pumps evaluated were significantly influ-
enced by low temperatures. In general, most provided a sampling capability at
-10° to -20°C, but were marginal at lower temperatures. None were useful at
-50°C. Most of the pumps survived low temperature exposures to -50°C without
significant damage.
The nickel-cadmium batteries which power these pumps are concluded to be the
most suitable available. Although the energy available from these batteries
is significantly reduced at low temperature, a large percentage of the nominal
energy is available at -20°C and some energy at -40°C. These batteries are
not useful at -50°C.
m
-------
ACKNOWLEDGMENTS
The authors wish to express their gratitude to the following companies for
their cooperation in the conduct of these investigations and for allowing us
to use a complement of sampling pumps on a consignment basis.
Anatole J. Sipin Company
386 Park Avenue South
New York, New York 10066
E.I. Du Pont De Nemours and Company
Applied Technology Division
Brandywine Building
Wilmington, Delaware 19898
MDA Scientific, Inc.
(Lefco Engineering)
808 Busse Highway
Park Ridge, Illinois 60068
Mine Safety Appliances Company
400 Penn Center Boulevard
Pittsburgh, Pennsylvania 15235
Research Appliance Company
Route 8
Gibsonia, Pennsylvania 15044
Spectrex Company
3594 Haven Avenue
Redwood City, California 94063
iv
-------
CONTENTS
Abstract iii
Acknowledgments iv
Introduction 1
Background 1
Scope 1
Test Protocol 2
Stability of the Flow Rate 8
Pressure Drop Versus Flow Rate Capabilities 8
Reliability of Calibration
Continuously Sample Cold Air
Other Performances
1
8
8
9
The Evaluation Program 10
Solicitation 4 10
Pumps Evaluated
Characteristics of the Alaskan
Environmental Effects
Metals
10
Environment 10
on Materials 16
16
G
S
2392PS . . . ,
C-200 . . . ,
DuPont Model P-125 . . ,
Accuhaler Model 808 . ,
Spectrex Model PAS-1000
Si pin Model SP-1 . . . ,
Summary
Recommendations
Investigations
Rubbers 16
Plastics 17
Unj__ .*_ 1 Tf
................ \ . 17
18
18
19
19
38
54
59
73
88
99
105
108
Ill
Ill
The Working Environment in (told Climates Ill
Motors
Lubricants ....
Batteries ....
Electric Components
Experimental Results
MSA Model
MSA Model
RAC Model
MSA Model
Battery Investigations
Developments
Recommended Use Techniques ,
Standards Recommended by RTI
References
Appendix
Ill
112
112
114
116
119
-------
FIGURES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
Photograph of the Basic Test Facility .• . . .
Schematic of the Basic Test Facility
Photograph Showing One of Each Type Pump Evaluated . . .
Photograph of the MSA Model G
Flow Rate Stability with Time, Model G, No. 1
Flow Rate Stability, Corrected for Temperature, Model G,
Flow Rate Stability, Corrected for Temperature, Model G,
Flow Rate Stability, Corrected for Temperature, Model. G,
Flow Rate Stability, Corrected for Temperature, Model G,
Flow Rate Stability, Corrected for Temperature, Model G,
Flow Rate Versus Pressure Differential, Model G, No. 1 .
Flow Rate Versus Pressure Differential, Model G, No. 2 .
Flow Rate Versus Pressure Differential, Model G, No. 3 .
Flow Rate Versus Pressure Differential, Model G, No. 4 .
Flow Rate Versus Pressure Differential, Model G, No. 5 .
Flow Rate as a Function of Temperature, Model G
Pump Current as a Function of Temperature, Model G . . .
Photograph of the MSA Model S
Illustration of the Model S Bypass Valve Arrangement . .
Flow Rate Stability, Corrected for Temperature, Model S,
Flow Rate Stability, Corrected for Temperature, Model S,
Flow Rate Stability, Corrected for Temperature, Model S,
Flow Rate Stability, Corrected for Temperature, Model S,
Flow Rate Stability, Corrected for Temperature, Model S,
Flow 'Rate Versus Pressure Differential Model S, No. 1
Flow Rate Versus Pressure Differential Model S, No. 2 .
Flow Rate Versus Pressure Differential Model S, No. 3 .
Flow Rate Versus Pressure Differential Model S, No. 4 .
Flow Rate Versus Pressure Differential Model S, No. 5 .
Flow Rate as a Function of Temperature Model S
Pump/Battery Currents in the MSA Model S Units
Photograph of the R AC Unit
Flow Rate Stability, Corrected for Temperature, RAC Model
2392-PS
Flow Rate Pressure Differential, RAC Model 2392-PS . . .
Flow Rate Calibration as a Function of Temperature, RAC V
2392-PS
Photograph of the MSA Model C-200
No. 1
No. 2
No. 3
No. 4
No. 5
No. 1
No. 2
No. 3
No. 4
No. 5
lodel
. 3
. 4
. 12
. 20
. 21
. 23
. 24
. 25
. 26
. 27
. 29
. 30
. 31
. 32
. 33
. 34
. 37
. 39
. 40
. 41
. 42
. 43
. 44
. 45
. 47
. 48
. 49
. 50
. 51
. 52
. 53
. 54
. 56
. 57
. 58
. 60
VI
-------
FIGURES
37. Flow Rate Stability, Corrected for Temperature, Model C-200,
No. 1 61
38. Flow Rate Stability, Corrected for Temperature, Model C-200,
No. 2 62
39. Flow Rate Stability, Corrected for Temperature, Model C-200,
No. 3 63
40. Flow Rate Stability, Corrected for Temperature, Model C-200,
No. 4 64
41. Flow Rate Stability, Corrected for Temperature, Model C-200,
No. 5 65
42. Flow Rate Stability of the C-200 Unit No. 1 with the Hg
Cell at Room Temperature 67
43. Flow Rate Versus Pressure Differential Model C-200, No. 1 ... 68
44. Flow Rate Versus Pressure Differential
45. Flow Rate Versus Pressure Differential
46. Flow Rate Versus Pressure Differential
47. Flow Rate Versus Pressure Differential
Model C-200, No. 2 ... 69
Model C-200, No. 3 ... 70
Model C-200, No. 4 ... 71
Model C-200, No. 5 ... 72
48. Photograph of the Du Pont Model P-125 74
49. Flow Rate Stability, Corrected for Temperature, Model P-125
No. 1 .* 75
50. Flow Rate Stability, Corrected for Temperature, Model P-125
No. 2 76
51. Flow Rate Stability, Corrected for Temperature, Model P-125
No. 3 77
52. Flow Rate Stability, Corrected for Temperature, Model P-125
No. 4 78
53. Flow Rate Stability, Corrected for Temperature, Model P-125
No. 5 79
54. Flow Rate Versus Differential Pressure, Model P-125, No. 1 ... 81
55. Flow Rate Versus Differential Pressure, Model P-125, No. 2 ... 82
56. Flow Rate Versus.Differential Pressure, Model P-125, No. 3 ... 83
57. Flow Rate Versus Differential Pressure, Model P-125, No. 4 ... 84
58. Flow Rate Versus Differential Pressure, Model P-125, No. 5 ... 85
59. Flow Rate Calibration as a Function of Temperature, Model P-125 . 86
60. Photograph of the MDA Model 808 88
61. Test Apparatus for the Accuhaler, Model 808 89
62. Flow Rate Stability, Corrected for Temperature, Model 808,
No. 1 91
63. Flow Rate Stability, Corrected for Temperature, Model 808,
No. 2 92
64. Flow Rate Stability, Corrected for Temperature, Model 808,
No. 3 93
65. Flow Rate Stability, Corrected for Temperature, Model 808,
No. 4 94
66. Flow Rate Stability, Corrected for Temperature, Model 808,
No. 5 95
67. Characteristics of Two Model 808 Limiting Orifices 97
-------
FIGURES
68. Flow Rate Calibration as a Function of Temperatures, Model 808 . 98
69. Photograph of the Spectrex PAS-100 99
70. Flow Rate Stability, Corrected for Temperature, Model PAS-1000,
No. 1 101
71. Flow Rate Stability, Corrected for Temperature, Model PAS-1000,
No. 2 102
72. Flow Rate Versus Differential Pressure, Model PAS-1000, No. 1 .. 103
73. Flow Rate Versus Differential Pressure, Model PAS-1000, No. 2 . . 104
74. Photograph of the Si pin Model SP-1 106
75. Flow Rate Stability, Corrected for Temperature, Model SP-1 ... 107
76. Flow Rate Versus Differential Pressure, Model SP-1 109
TABLES
1. Pumps Solicited and Received for Evaluation 11
1A. Companies Receiving Requests for Literature Descriptive of
Personal Sampling Pumps 119
V11T
-------
INTRODUCTION
The Research Triangle Institute (RTI) has conducted an experimental program to
evaluate the characteristics of personal sampling pumps at temperatures as low
as -50°C. These evaluations were completed under a National Institute for
Occupational Safety and Health (NIOSH) contract, "An Evaluation of Personal
Sampling Pumps In Sub-Zero Temperatures," Contract No. 210-76-0124v The re-
sults of these evaluations are reported herein.
BACKGROUND
The development of the large oil reserves along the northern slope of Alaska's
Brooks Range and nearby Prudhoe Bay typifies the industrialization taking
place in cold environs. These developments, precipitated by America's growing
dependency on foreign oil with its inherent threat of crippling embargoes, is
taking place in a severe environment which tests the endurance of both men and
machines. The growth of industralization in these and other cold regions has
generated a growing need for air sampling instrumentation and methodology
suitable for industrial hygiene investigations in sub-zero temperatures. In
response to these needs, NIOSH has sponsored the investigations described
herein. Personal sampling pumps are significantly important industrial
hygiene tools, and RTI has investigated their suitability for use in cold
environs, i.e., to temperatures as low as -50°C. The Alaskan environment is
also briefly defined since it is a cold region currently undergoing consid-
erable industrialization and is perhaps reasonably typical of other cold
environs.
SCOPE
The pumps evaluated during these investigations were tested at temperatures
between ambient room temperature (25°C) and a low temperature of -50°C. While
many environmental factors or stresses act in synergism with temperature, most
are of significance only at high temperatures and were not incorporated into
the experimental program described in this report. One environmental factor
which does have an important synergistic relationship with temperature at low
temperatures is humidity. In cold environs humidity is low. Consequently,
air sampled during these tests was obtained from a dry air manifold.
The personal sampling pumps evaluated can be categorized into two different
groups. One group is the Coal Mine Dust Personal Sampling Units (CMDPSU) for
which a certification program has been defined and is in effect (Ref. 1). In
order to be certified, the CMDPSU pumps must be capable of operating from
their internal battery packs at a nominal flow of 2 liters per minute against
a resistance of 4 inches of water for not less than 8 hours, presumably at
room temperature. Thus, a performance requirement has been defined which is
1
-------
useful in establishing test conditions for the evaluation tests described
herein.
The other class of pumps are used with a variety of sample collection devices
under a variety of conditions. This suggests, and it is true, that these
pumps, in the aggregate, provide industrial hygienists a wide range of
sampling capability. A given pump, in contrast, may only function over a
narrow region of the aggregate range. Thus, it is impractical to define a
single set of test conditions which will be meaningful in terms of the capa-
bility of all of these pumps. The practice adopted for these evaluations was
to test each pump under a set of conditions which was somewhat representative
of each pump's capabilities as stated by the manufacturer. CONSEQUENTLY, THE
DATA REPORTED HEREIN ARE NOT INTENDED TO BE USED TO COMPARE ONE PUMP AGAINST
ANOTHER. Instead, these data are intended only to demonstrate how each pump's
operation is affected by a low temperature environment.
TEST PROTOCOL
Those aspects of the evaluation program which required experimental tests at
low temperatures were generally conducted over a range of temperatures between
+25° and -50°C. In some instances, performances at intermediate temperatures
were such that no tests were run at lower temperatures. However, all pumps
were placed in an operational mode for at least one period of 8 hours at
-50°C. In general, all tests were completed at temperatures of -20°C and
above before tests at lower temperatures were conducted, so that any cata-
strophic failures at very low temperatures would not circumvent a completion
of planned testing at intermediate temperatures, i.e., -20°C and above. Test
apparatus and procedures for some of the more frequent tests are described in
following paragraphs.
During the experimental program, pump batteries were used very carefully to
enhance their capacity and cycle-life. With few exceptions, batteries were
stored fully charged, used for a full cycle of operation, and recharged
according to the manufacturers instructions using the charger supplied with
the pump. It is doubtful that many pumps will be used as carefully in the
field environment. Careful use was probably responsible for the reliable
performance of batteries observed during those investigations.
Basically, the test apparatus consisted of a temperature test chamber which
contained the pumps under test and a heat exchanger for each pump to equili-
brate the air sample with the test chamber ambient. The test chamber was
cooled with liquid nitrogen (LNp). When the LN2 is turned on, the
chamber begins to modulate a valve that releases gas from the LN2 source
into the chamber. When the chamber reaches the set temperature, the modula-
tion changes to control temperature at the set point. Our observation is that
for a temperature change of 50°C, e.g., +20°C to -30°C, a thermocouple in the
center of the chamber will indicate that 55 percent of the transition has
occurred in 5 minutes, 75 percent in 10 minutes, and 90 percent in 25 minutes.
The transition is essentially complete in about 50 minutes. A thermocouple
located in a 2£/min. effluent from the shortest («7 ft.) heat exchanger
essentially duplicates the chamber air temperature. The chamber pressure is
-------
slightly above ambient when gas from the I_N£ source is modulated into the
chamber; thus, condensation does not occur on the pumps under tests. The
actual pressure difference largely reflects the tightness of the seals at the
access port through the chamber walls. Our observation is that this differ-
ence is less than 1/4" HoO, and it was neglected as an experimental
factor. Pressure in the dry air manifold was negligibly different from the
room ambient. The sampled air was taken from a dry air manifold to more
closely simulate use conditions in cold environs and circumvent condensation
in the heat exchanger and pump. The instrumentation was external to the test
chamber. Flow rates were measured with rotometers that were calibrated
against bubble flow meters at the beginning of the test program. Pressure
drops were measured with a U-tube water manometer and temperatures with either
an alcohol-in-glass thermometer or a thermocouple. Figure lisa photograph
of the basic test facility, and Figure 2 is a schematic diagram of the basic
test apparatus configuration. Specific procedures and apparatus are described
in subsequent paragraphs.
Figure 1 Photograph of the Basic Test Facility
-------
-------
In the basic test configuration of Figure 2, a pump under test pulls sample
air from a room temperature, room pressure manifold, and the flow rate is
measured at room temperature. The air sample passes into the test facility
and through a heat exchanger before entering the pump and, therefore, is at
the test chamber temperature when it enters the pump. Thus, the sample air is
exhausted at the test temperature and at room pressure. According to the
general gas law (PV=NRT) the volume of air which passes through the pump is
less than the volume indicated by the room temperature rotometer. To
elaborate, one can write that for the sampled air,
TR Tc '
where PR.VR, and TR are the pressure, volume and absolute
temperature of a given sample at room conditions and Pc,Vg, Tc
are similar quantities at the test chamber conditions. Since PR and
Pc are the same,
VR - Vc
Thus a volume VR at room condition will occupy a volume of
V = VD Tc
C R y-
'R
when passing through the pump and exhausting into the test chanber at test
chamber conditions. Since the volume is related to the volume flow rate by
Qt - V ,
where Q is volumetric flow rate and t is time, it is also true that
Qc = QR T^ '
C R TR
In subsequent presentations of data, references to corrected data mean that
flow rates measured at room temperature have been multiplied by the
TQ/TR ratio to obtain a value corrected to the test chamber condi-
tions.
It is also significant that the rotometer on the input of several of the
pumps, e.g. the CMDPSUs, will also be influenced by the sample temperature and
the fact that it is at a different pressure. Nelson [Ref 2] gives correction
factors for rotometers for variations of pressure and temperature which differ
from the initial rotometer calibration pressure and temperature. For a tube
-------
of linear range (which is characteristic of the tubes used in these evalua
tions and the tubes on most of the pumps), the correction is given by
QT P
'2*2 =
where
2^2
rate at test temperature and pressure,
Q
T,P. = flow rate at calibration temperature and pressure.
= test temperature and pressure, and
I,, P. = calibration temperature and pressure.
Rotometers on the pumps presumably were calibrated at or near room temperature
and a pressure near 1 atmosphere. Most of these tests were conducted at 0.997
atmosphere (-4 inches of water gauge) or less. Thus, corrections due to
pressure differences are small, i.e., less than 0.25 percent, and are not
considered further. The correction for temperature, however, can be
significant. According to Nelson [Ref.2], the flow rate indicated by a pump
rotometer at test chamber conditions will differ from the flow rate indicated
at calibration conditions (and also indicated by the room temperature
rotometer in Figure 2) by a factor of
where
= flow rate corresponding to calibration conditions
(and which would be read on the room temperature
rotometer),
= flow rate indicated on the pump rotometer at test
chamber conditions,
= test chamber temperature, and
= room temperature (calibration temperature).
6
-------
Thus, if a CMDPSU unit is pumping 2£ /min at room temperature (e.g. 21°£ or
294°K), the pump rotometer and the room temperature rotometer should read
2a /min. If the test chamber temperature is lowered to 0°C (273°K), for
example, and the room temperature rotometer continues to read 2£ /min, the
pump is actually pumping at a volumetric flow rate of
273
= 1.86A/min,
and the pump rotometer would indicate a flow of
Qcp = 2 A/mi n
= 1.93£/rnin,
using the room temperature calibration curve.
The significant point is that although the mass flow is constant throughout
the system, the room temperature rotometer shows a different volumetric flow
rate than is actually pumped at low temperatures, i.e., the room temperature
rotometer reads high. A pump rotometer at the same temperature as the pump
will also read differently. If its reading is interpreted from a room
temperature calibration, it will indicate a higher flow rate than is actually
being pumped (by the square root of the temperature ratio), and a lower flow
rate than the room temperature unit (again by the square root of the
temperature ratio). In the preceding example, the actual volumetric flow rate
through the pump is 7 percent less than is indicated by the room temperature
rotometer and 3.5 percent less than indicated by the pump rotometer using the
room temperature calibration. The two rotometers will differ by only 3.6
percent.
Under worst case conditions, i.e., assuming a room temperature of 25°C, a pump
rotometer calibrated at 25°C, and a test temperature of -50°C, a 2&/min flow
on the room temperature rotometer would appear as a flow of 2 (223/298) =
1.73& /min (a difference of 13.5 percent) on the pump rotometer, and would
correspond to a volumetric flow rate of 2 (223/298) = 1.5s, /min (a difference
of 25 percent) through the pump.
In the discussions of experimental data in subsequent paragraphs, references
to corrected flow rates mean that flow rates measured on the room temperature
rotometer have been multiplied by the ratio of the absolute test chamber
temperature to the absolute room temperature. These corrected data may or may
not be more useful. When an air sample volume is cooled, a particulate or
gaseous contaminant concentration will also become more dense, i.e., the
contaminant concentration expressed in mg/m^, for example, will also
increase. (The concentration in parts per million (ppm) will remain
unchanged.)
7
-------
Stability of The Flow Rate
Flow rate stability with time was measured at test temperatures as follows.
Pumps under test were placed in the test chamber and adjusted, at room
temperature, to a specified flow rate at a specified inlet pressure.
(Presumably, in a cold region environment, pumps would be stored, recharged,
and obtained by the wearer/operator in a protected environment.) The pumps
were turned on and the test chamber temperature was set to the desired
temperature at time zero. Flow rates, pressure drops, and battery voltages
were monitored and recorded at regular intervals over an 8 hour period. Pumps
that provided the wearer/operator with a flow-readout device, e.g., the
rotometers on the CMDPSU pumps, were re-adjusted during tests to a higher flow
rate whenever the flow dropped to an unacceptable value. (It will be seen
that an adjustment to maximum was usually required.) At low temperatures,
when acceptable flows could not be maintained, the pumps were switched during
the tests from their internal battery packs to a laboratory power supply.
Pumps that did not provide a flow readout or a convenient flow adjustment were
not readjusted during tests. A wearer/operator in a field environment would
not have any way of knowing that the flow rates were inadequate or how to
adjust the flow rate to a known value. When the flow rate decreased to an
excessively low value, however, a laboratory power supply was substituted for
the internal batteries.
Pressure Drop Versus Flow Rate Capabilities
Pumps under test were instrumented to measure flow rate and pressure drop as
illustrated in Figure 1. The test chamber temperature was established and the
pump temperature allowed to equilibrate for one hour. The pump was not
operated during the equilibration period so that the batteries would remain in
an approximately fully charged state for all test temperatures. The pumps
were operated at each test temperature. The pressure drop value was used to
establish a test pressure and the flow rate determined. After completing a
data set at a given temperature, a new test temperature was established, the
pump temperature equilibrated for an hour with the pump not operating, and a
new data set determined at the test temperature. For each data set, the pump
was initially calibrated to a "standard" flow rate and pressure drop at room
temperature.
Reliability of Calibration
These data are inherently available in the data descriptive of each pump's
flow-rate stability with time. It was obtained from the flow rate stability
data corresponding to the first-hour data point at each test temperature.
Continuously Sample Cold Air
It became evident during these evaluations that most of the pumps would
function reasonably well at -20°C, and that few would function satisfactorily
at -30°C. Consequently, tests to demonstrate an ability to continuously
sample cold air were conducted at -20°C. These tests consisted of
establishing a manifold inside the test chamber at approximately the desired
pressure drop and operating a complement of pumps from the manifold at -20°C
8
-------
for 8-hour periods. The only data tabulated was to document each
pump's operation at the end of the 8-hour period.
Other Performances
Other performances evaluated consisted of subjective evaluations of different
pump features or characteristics over the period of the experimental program.
These included such features as ease of recalibration, battery performance,
diaphragm reliability, bearing design, lubricants, and case ruggedness.
-------
THE EVALUATION PROGRAM
SOLICITATION
It was the intent of RTI to identify as many personal sampling pumps as
possible to be included in this evaluation program. Various directories and
guides were searched to identify manufacturers and distributors of personal
sampling pumps. The directories searched were the Thomas Register, MacRae's
Bluebook, U. S. Industrial Directory (Vol. Ill), and Guide to Scientific
Instruments (published by the American Association for the Advancement of
Science).Other sources of information were various trade publications and
RTI and NIOSH personnel. Each distributor or manufacturer identified received
a letter describing the planned evaluation program and requesting brochures,
specifications, and other information descriptive of personal sampling pumps.
From responses, personal sampling pumps were selected for evaluation and
solicited on consignment from the manufacturer or distributor. The selection
was reviewed and approved by the NIOSH Project Officer. Five of each model
of pump was requested to enhance the validity of the experimental results. An
Appendix includes a list of the organizations that received a request for
descriptive literature, a copy of the request letter, and a copy of the pump
solicitation letter without the company-specific details.
PUMPS EVALUATED
A total of 85 pumps were requested for evaluation, i.e., 5 each of 17
different pumps. Twenty-nine were eventually received. Some manufacturers
chose not to participate in the program, and others were still somewhat in a
development period and their pumps were not ready for an evaluation. The
pumps requested and those actually received for the evaluation are identified
in Table I. Figure 3 is a photograph showing one of each type pump evaluated.
CHARACTERISTICS OF THE ALASKAN ENVIRONMENT
Separating the interior of Alaska from its territory to the north is a
600-mile stretch of low, but rugged mountains known as the Brooks Range. This
expanse of mountains forms a natural partition between the climatic divisions
designated (by the U. S. Weather Bureau) as the Interior Basin to the south
and the Arctic Area to the north. The harsh, subzero environs of the Arctic
10
-------
TABLE 1
PUMPS SOLICITED AND RECEIVED FOR EVALUATION
Manufacturer/Address
ANATOLE J. SIPIN COMPANY
386 Park Avenue South
New York, NY 10066
Pumps
Solicited
(5 of each)
Model SPI
Pumps
Received
1 Model
BENDIX ENVIRONMENTAL SCIENCE DIVISION
1400 Taylor Avenue
Baltimore, MD 212U4
Micronair II
Model Cll
CENTURY SYSTEMS CORPORATION
P. 0. Box 133
Arkansas City, KS 67005
Portable Air
Sampling Pump
E. I. DU PONT DE NEMOURS AND COMPANY
Applied Technology Division
Brandywine Building
Wilmington, DE 19898
Model P200
Model P4000
5 Model P125
MDA SCIENTIFIC, INC.
(Lefco Engineering)
808 Busse Highway
Park Ridge, IL 60068
Model 808
5 Model 808
MICROCHEMICAL SPECIALTIES CO.
1825 Eastshore Highway
Berkeley, CA 94710
Model 4000
MINE SAFETY APPLIANCES CO.
400 Penn Center Boulevard
Pittsburgh, PA 15235
Model C-200
Model G
Model S
5 Model C-200
5 Model G
5 Model S
RESEARCH APPLIANCE CO.
Route 8
Gibsonia, PA 15044
Cat. No. 2392-K
1 Cat. No.
2392-PS
11
-------
TABLE 1
PUMPS SOLICITED AND RECEIVED FOR EVALUATION
(Continued)
Manufacturer/Address
SKC INC.
P. 0. Box 8538
Pittsburgh, PA 15220
Attn.: Mr. Lloyd Guild
SPECTREX COMPANY
3594 Haven Avenue
Redwood City, CA 94063
Attn.: Mr. John M. Hoyte
Pumps
Solicited
(5 of each)
Model No. 222-3P
(50-200 ml/min-
with changer)
Model No. PAS-2000
Model No. PAS-1000
Pumps
Recei ved
2 Model No,
PAS-1000
WILLSON PRODUCTS DIVISION
P. 0. Box 622
Reading, PA 19603
Model BC
Figure 3 Photograph Showing One of Each Type Pump Evaluated
12
-------
Area, particularly the Alaskan oil fields located along the north slope of the
Brooks Range, are our primary concern here.
Although recognized as a definite topographical barrier, it is difficult to
determine just how much the Brooks Range influences the climate of the Arctic
Area. In similar topographies, around Ft. Yukon in the Interior, for example,
there is a significant downslope drainage of cold air off the northern slopes
of the surrounding range which drops temperatures in the area to notable lower
levels. However, this does not seem to occur in the Arctic Area. In fact,
temperatures within the Arctic Area generally fall within a much narrower
limit than those of the Interior Basin and, on the average, tend to be higher.
In the Arctic Area, temperatures in January average out at -24°C and in July,
8°C. The mean annual temperature logs in at -12°C.
Temperatures never reach extreme highs in the Arctic Area. As the Sun moves
further above the horizon during the summer months, prolonged periods of
continued daylight with a greater amount of possible sunshine occur and
temperatures rise. However, even then, highs seldom go above 27°C.
This is partially due to the fact that in summer the Sun's rays reach the
earth at such relatively low angles as to cause little surface warming.
During the winter months extreme lows from the Weather Bureau's five reporting
stations in the Arctic Area range between -43°C and -53°C, with the station at
Umiat holding the record low. It is seldom that Arctic Area stations have
recorded temperatures lower than -46°C. By comparison, 9 out of 10 reporting
stations in the Interior Basin have recorded lows of less than -46°C.
The Arctic experiences much more wind during the winter months than any of the
other climatic divisions of Alaska. As a result, prolonged periods of
drifting and blowing snow occur, making winter life especially tenuous. This
is particularly true along the coast from Cape Lisburne to Barter Island,
which encompasses the oil fields at Prudhoe Bay. Winds along this area may
occur frequently at 50 to 60 mph. At Barter Island it is not unusual to have
winds of 70 mph during the months of February, March, and November, and the
station there recorded a wind speed of 86 mph in February one year.
Because of these high winds and subsequent drifting snow, measurements of
precipitation have been difficult to obtain. At any rate, precipitation is
light. Maritime factors play some role here, but due to the prevalence of
cold air and its low moisture carrying capacity, precipitation is much lighter
than one would normally expect for maritime characteristics.
The mean yearly precipitation (all forms converted to rain equivalent) for the
entire Arctic Area is recorded at 6 inches. Cape Lisburne records the highest
13
-------
annual precipitation at 16 inches. Other stations average from 4 to 10
inches. Kotzebue, which is sheltered from moisture-carrying southwestern
winds, averages 8 inches yearly.
Snowfall makes up a large portion of the precipitation reading, and Cape
Lisburne receives the greatest amount annually at 65 inches. Other areas
along the Brooks Range average about 40 inches. Kotzebue receives a smaller
proportion of snow to total precipitation than any other reporting station.
The same winter winds which whip the snow about tend to pack it down into
firm, thick layers along the northern coastal sections. From Cape Lisburne
eastward, these ice layers begin to appear along the shoreline around
September or October and remain there until warming temperatures bring about
their breakup around June or July. Once these layers have adhered and there
is no open water surrounding the coastline, the climate demonstrates fewer
maritime characteristics and more continental characteristics.
Throughout the year much of the Arctic Area remains in a continuous state of
permafrost. Vegetation exists as tundra, a mat-like cover over the frozen
ground composed of lichens, mosses, wild flowers and tiny willow trees. In
spring when the surface begins to warm, the top layers of the permafrost will
melt making the entire area soggy and creating numerous glacial lakes. Near
the shore, the tundra is largely covered by standing water. (Refs. 3-10)
Arctic Area weather stations in the general area of the Brooks Range and
Prudhoe Bay include Point Barrow, Barter Island, and Kotzebue. Point Barrow,
centrally located to the north of the Brooks Range oil reserves, is the most
northerly of the U. S. Weather Bureau's stations, and usually records one of
the lowest mean air temperatures for Alaska during the winter months.
(However, it does not hold the lowest temperature on record. Tanana, in the
Interior Basin, dropped to -60°C in January 1886.) Temperatures here
generally fall below freezing all year round. A daily maxima of 0°C or higher
may be recorded for only 109 days out of the year, while a daily minima of
less than 0°C may be recorded for 323 days. In addition, lows of 0°C or less
are usually recorded every month of the year.
February is the coldest month with a mean temperature of -28°C. March is only
slightly warmer, and it is not until April that temperatures really begin an
upward trend. May marks a definite swing from winter to summer, and
temperatures become consistently warmer with July recording the highest mean
temperature for the year at about 4°C. It is at this point, in late July or
early August, that the surrounding Arctic Ocean becomes ice-free for the first
time in summer. September usually marks the end of summer, and by November
temperatures have dropped such that half the daily means for the month are
zero or below.
Airline pilots find that March is the most favorable month at Point Barrow as
far as average cloudiness, fog, precipitation and wind are concerned. In
fact, March and December are both considered calm months with average hourly
wind speeds of less than 11 mph. However, this does not mean high winds are
not experienced during these months. Indeed, they are, and the highest winds
14
-------
for any month of the year are often recorded at 40 to 50+ mph. The highest
monthly mean wind speed occurs in October at 13.7 mph.
At the end of March, when daylight increases from 9 to 14 hours, temperatures
begin to rise. With this upward trend clouds appear, and by April there are
clouds in the sky daily. By May, ceilings of 1,000 feet or less occur at
least half the time, and snow and drizzle may fall frequently.
As the Sun moves higher and sunlight extends to 24 hours/day, cloud cover
becomes so extensive that by June Point Barrow experiences 24 hours of
continuous, moderate light daily. By July, the city is often under heavy fog.
This cloudy, foggy weather continues into October and November, and it is not
until colder, shorter days occur (in December and January) that the sky begins
to clear.
Precipitation in Arctic Alaska is quite low and Point Barrow is no exception.
The city records an average precipitation of only 4 inches of rain or so for
the year. This includes 29 inches of snow converted to a rain equivalent.
(One inch of rain is considered the equivalent of 10 inches of snow.) July
records the greatest monthly average precipitation, and August the greatest
average number of days with precipitation of .01 inches or more. October
generally has the most snowfall of any month.
Barter Island, located off the northeast shore of the Alaskan mainland, and in
the general vicinity of Prudhoe Bay, has no real topographic features of its
own to affect temperature and precipitation. The island terrain and that of
the nearby mainland is low, marshy, flat tundra with numerous lakes, and there
are no elevations of consequence until the Brooks Range, 65 miles south.
Consequently, its climate is largely determined by the surrounding open Arctic
waters.
In fact, the Arctic Ocean (though often frozen) is one key element which
prevents extremely low temperatures in the area during the long Arctic night
period (November-January), and during summer months moderates the effects of a
continuous period of possible sunshine from mid-May to late July.
Freezing temperatures are reached as a general rule during all months of the
year. Daily temperatures reach a monthly maxima of -8°C in April and a minima
of -13°C in June.
Snow, which falls practically every month, covers the ground about 8 months of
the year. Accurate measurements of snowfall and precipitation, however, are
difficult to obtain due to the very strong winds experienced from October to
February. These winds contribute to generally unpleasant winter conditions.
Kotzebue is a long, thin peninsula located in northwestern Alaska, 26 miles
inside the Arctic circle. Water bodies surrounding Kotzebue produce a
maritime type of climate from mid-May to October. During this period cloudy
skies and fog prevail and temperatures are fairly uniform.
15
-------
The Kotzebue area generally experiences less severe low temperatures than the
area north of the Brooks Range. In fact, temperatures here have never dropped
as low as -46°C and there are long periods in the year (70-90 day stretches)
when temperatures reach 0°C or higher. Maximum temperatures around 2°C occur
even during the coldest months of the year and minimum readings remain at
-18°C at least half of the days during the coldest winter months.
Annual precipitation is light at 8 inches. Over half that occurs during the
months of July, August, and September. Snowfall averages about 41 inches a
year and falls nearly every month in the year. Cyclonic storms, accompanied by
high winds may occur during the winter months. (Refs. 11 - 13)
ENVIRONMENTAL EFFECTS ON MATERIALS
Metal s
One of the most significant limitations to the use of metals at low
temperatures is their increased tendency to brittle failure. While yield
strength and tensile strength may actually increase as temperatures decrease,
most metals concurrently develop a lowered resistance to impact or shock
loading, and brittleness occurs. In some metals there is also a distinct
decrease in ductility with lowered temperatures occurring over a relatively
narrow temperature range. In addition, notch defects (scratches, nicks,
holes, threads, machining marks, etc.) increase a metal's vulnerability to low
temperature at which a metal will begin to behave in a brittle fashion.
[Refs. 14-17]
Austenitic stainless steels are generally well recommended for low temperature
environments because they retain their ductility and tensile strength, and are
brittle-safe. Their most notable application is in cryogenics, where they are
effective down to extremely low temperatures (-253°C). In addition, simple
single phase structure of austenitic stainless steels allows them to maintain
their low temperature properties, even after welding.
Metals themse'lves may not be suitable for subzero temperature applications,
but metal alloys have proved to be more promising. Although aluminum alloys
have lower densities than stainless steels, their yield'strength-to-density
ratios qualify some for low temperature use. Welding, however, reduces their
effectiveness. Titanium alloys possess relatively high strength and low
density, and are appealing in that they are lighter structures than steel or
aluminum alloys, but with equal load-carrying capacity. Although relatively
expensive to use at present, they are viewed as likely candidates for future
applications. Nickel alloys possess high ductility at low temperatures in
both solution-annealed and age hardened conditions. They are quite useful in
wide temperature extremes. [Ref. 18]
Rubbers
Low temperatures often have adverse effects upon rubbers. Temperature
decreases may create changes in brittleness, flexibility, and compression set
characteristics, resulting in lost resilience, increased stiffness and
16
-------
increased hardness of the rubber. Further drops in temperature may eventually
result in crystallization and vitrification, but the material usually becomes
unserviceable well before this point. Adding selected plasticizers tends to
improve flexibility at low temperatures, but often this is done at cost to
tear, resistance to abrasion, and bondability.
There are suggested low temperature types of rubber available, such as
silicone rubber. However, some of these have low chemical resistance, making
them susceptible to material(s) passing through the rubber tends to extract
those compounds giving the rubber its low temperature characteristics, thus
reducing the rubber's efficiency. [Ref. 17]
Plastics
Although the strength of plastics increases with lowered temperatures, the
durability of nearly all plastics subject to shock decreases as temperature
decreases. Many plastics can be used successfully at temperatures as low as
-40°C, provided that they are not subject to shock loading. Fluorocarbon
plastics such as Teflon and Kel-F retain useful ductibility at temperatures
even lower.
Organic plastics may undergo reversible or irreversible changes at low
temperatures. Reversible effects include:
1. Dimensional changes due to thermal contraction and loss
of moisture.
2. Increased yield and ultimate strength,
3. Decreased ductility,
and 4. Decreased resistance to impact.
Irreversible effects include:
1. Dimensional changes due to change of state,
2. Physical failure,
and 3. Crystallization.
Motors
Motors will start and operate satisfactorily at -50°C if specially developed
low temperature lubricants are used. In general, as temperatures are lowered
to this point, final operating speed is lowered and input power demand
increases. [Ref. 17]
Lubricants
Many lubricants prove useless at low temperatures because of their tendency to
freeze or harden. However, there are commercial engine and gear lubricants
available which are useful down to -53.3°C, and some instrument greases useful
at -73.3°C. Solid-film lubricants, used in cryogenics, are effective at even
lower temperatures. [Ref. 17]
17
-------
Batteries
Storage batteries lose a large percentage of their power capacity at low
temperatures. For most applications in severe cold weather, the
nickel-cadmium appears to be the most reliable. A more complete discussion of
batteries at low temperatures follows in a subsequent section of this report.
Electronic Components
Component parts, such as resistors, capacitors and transistors, may develop
problems at temperatures as low as -40°C. Their values may vary enough to
necessitate readjustment when accurate situations are needed. However, in
most cases, current products are manufactured which have been specifically
designed for military usage, and are thus guaranteed for satisfactory
operation at temperatures as low as -55°C. [Refs. 14, 2U]
Resistors--
Wirewound resistors perform satisfactorily at low temperatures with very
little change in resistance. Even as low as -65°C, resistance may vary less
than 1 percent of its normal value. Copper-nickel wirewounds, for example,
exhibit only a 0.5 percent change in resistance at -65°C. Temporary
electrical discontinuity in variable wirewound resistors (due to ice formation
or hardening of the lubricant on the resistance element) has been reported for
operations at -55°C. [Refs. 17, 21]
Where carbon composition resistors are concerned, change of resistance with
decreasing temperature is insignificant. A slight increase in resistance
(.5 percent) may occur at -55°C. With other types of composition resistors,
resistance may vary from 1U-50 percent of normal value as temperature ranges
from -55°C to -3.9°C. In addition, cracks in plastic insulating tubes may
occur, shortening the resistor's life. Torque and electrical discontinuity
problems may also occur for this type of resistor at low temperatures. [Refs.
17, 22]
Capacitors--
Capacitors are affected to varying extents by subzero temperatures. For most
applications, temperature compensating capacitors are readily available from
manufacturers. The capacitance of electrolytic capacitors drops off rapidly
at low temperatures, although manufacturers can control this to some extent in
determining the nature of the chemical makeup of the electrolyte. Aluminum
electrolytes become ineffective at -40°C and lose all their capacitance at
-55°C. (The electrolyte freezes, creating a high power factor and low
capacitance.)
Tantalum capacitors may retain 80 percent of their normal value at -55°C, but
only if used at low frequencies. If high frequencies are used, values drop
off quickly. In general, tantalum capacitors are more acceptable for low
temperatures than aluminum capacitors. [Refs. 17, 12, 24] Ceramic capacitors
are designed to withstand and be operational at temperatures as low as
-55°C, with little capacitance change (less than 0.5 percent of normal
capacitance). [Ref. 25]
18
-------
Paper and mica capacitors function at -55°C, although large changes (20
percent below normal variation) may pccur with paper capacitors as the
temperature drops to -50°C. Glass dielectrics operate at temperatures from
-55°C to +125°C with only -1 percent change in normal capacitance in the lower
levels.
Plastic-film capacitors appear promising as far as low temperatures are
concerned. Polystyrene, polyparaxylene and TFE-fluorocarbon in this category
generally rate at only 1 percent below normal capacitance at -50°C while even
other plastic-film types rated no lower than 5 percent below normal at -50°C.
[Refs. 26-29] Wax impregnated capacitors are subject to extensive cracking
below -40°C, resulting in permanent changes in capacitance, insulation
resistance and ac losses. [Ref. 17]
Transistors--
Transistors (and integrated circuits), like capacitors and resistors, can be
designed to compensate for low temperature environments, and manufacturers
usually produce a grade of components which fall into the military
applications class (providing for guaranteed operation as low as -55°C).
Detailed information on..their performance at low temperatures is not readily
available. However, temperature ranges and other information regarding
performance at low temperatures is usually available from the manufacturer's
catalog or specifications sheet. [Refs. 30-32]
Switches, Controls, Jacks and Plugs--
Difficulties with many of these components at subzero temperatures often occur
because greases used with them freeze. In addition, differences in
contraction of closely allied plastic and metal parts may cause cracking,
which in turn may result in poor connections and short circuits. [Refs.17,20]
EXPERIMENTAL RESULTS
MSA Model G
Five MSA Model G CMDPSU units were evaluated. Figure 4 is a photograph of two
of these units. (A thermocouple wire added for test purposes is evident on
one of these pumps.)
Flow Rate Stability with Time--
The flow rate stability of the MSA Model G units was measured in the test
apparatus of Figures 1 and 2. The pumps were calibrated at room temperature
for 2.1 liters per minute at 4 inches of water. The pumps under test were
turned on and the test chamber turned on to the desired temperature somewhat
19
-------
Figure 4 Photograph of the MSA Model G
concurrently, and the flow rate and pressure drop were measured over an 8-hour
period. (As discussed previously, it can be assumed that the pump is
approximately equilibrated at the test temperature after one-half hour and
completely equilibrated in one hour.) Since the Model G has a rotometer to
indicate flow, the flow rate was adjusted upward whenever it dropped to an
unacceptable value. The pressure drop across each pump and each pump's
rotometer reading were also monitored. These data are presented in this
section.
Figure 5 is a plot of the Model G unit No. 1 flow rate stability with time.
As shown, the room temperature performance of this unit was excellent. The
-1° and -11°C data also show acceptable performance. In the -11°C data, for
example, the pump's rotometer never dropped below the "yellow" band, which is
indicative of satisfactory performance. At -22°C, the flow dropped below 1.6&
/min after 4 hours. A power supply was substituted for the battery pack and
the test continued. The pump's performance was maintained above the 1.6«,/min
flow rate, but a dramatic improvement was not achieved. This suggests that
the degradation in performance was only partly due to battery degradation.
At -30°C, the flow rate dropped after one hour to 1.37&/min. It was then
adjusted for maximum flow which was 1.76ji/min. After 4 hours, it had dropped
again to less than 1.6£/min. A power supply was substituted for the battery
pack and acceptable performance was maintained for the remainder of the 8-hour
period.
20
-------
o 0 0 o
— oo o o °
£j ° o N 2
o M _ r CM •*>
Jg *• 1 III
• 0 o
CM CM ~
—
>-
a.
a. _
3
CO
ac
ui
1
Z o
z H
x >•
< "
1 £
0 H
2 <
r- to
* _
o i
H F
U. ll.
CO 0
3 H
•9 —
o >
< co 0
CO «
U _ (M $ -
C *- ** i
S J
1
+
i
§
1
+ -
•
1
1
1
1
+ -
1
1
i
+
/
~ f -*
3 x I* -n3
^ w*^ ~
i
CO
N-
i-H
,
^?
UI
z ^
— J2
H 5
1/1
01
» «
o;
o
r—
I i
IJL.
in
CM QJ
3
cr-
•r-
U.
•^
r»
* *!
~ ~"
jiu / I ) 31VW MOld
-------
It is very significant that the battery terminal voltage at the time of
substitution was l.lV/cell (i.e., 5.56V for 5 cells). However, the pump
current had increased from 201 mA at +22°C to 282 mA. When switched to a
power supply at 6.2 volts, the pump current was 289 mA and eventually
increased to 291 mA. This is evidence that degradation in performance is due
to an increased power demand from the pump rather than a degradation in the
battery pack.
At -42°C, the flow rate was adjusted to a maximum after one hour. This did
not restore acceptable performance and a power supply was immediately
substituted for the battery pack. The battery pack voltage at that point was
1.14V/cell. However, the pump current had increased from a +23.5°C value of
196 mA to 323 mA. It eventually reached 340 mA after 7 hours.
At -50°C, pump performance immediately degraded and a flow adjustment to
maximum and a power supply substitution did not restore acceptable
performance. The battery voltage at the substitution point (i.e., at 30
minutes) was 1.15V/cell, and the pump current was 330 mA. It eventually
reached 425 mA. The pump flow remained very low ( 300ml/min) and tended to
pulsate such that accurate readings could not be made.
In Figure 6, the flow rate stability data from Figure 5 are repeated except
that the flow rates are corrected for each test temperature. This type of
format tends to spread the data out somewhat; however, it is generally a more
useful presentation.
Figure 7 is a plot of corrected flow rate stability data for the MSA Model G,
No. 2. Some significant features of these data are that at -1°C the flow rate
remained higher than at +24°C, and the -22°C flow rate remained at a
satisfactory level. (The higher flow rate at -1°C is also evident in other
data.) The -22°C performances were excellent. The pump's rotometer remained
in the "yellow" range throughout the tests, and the room temperature flow was
1.8£/min after. 8 hours. It is significant that the battery voltage was
l.lV/cell .at the end of the 8-hour test.
The -30°C results were marginal. With a maximum flow rate adjustment, the
pump supplied 1.96£/min at 3.85 in. of H20 for more than 6 hours. The
battery pack voltage was l.lV/cell after 6 hours, and it was delivering 224
mA. At room temperature, the pump required 184 mA at 6.6V or 1.32V/cell.
These results suggest that battery pack capacity and not degradation is a
significant factor in a pump's performance at low temperature. The Model G
requires significantly more current to operate at low temperatures.
The -40° and -50°C data are similar to the -30°C data. . Satisfactory
performance could not be maintained even with a power supply. The pump
required excessive current and, after one hour at -50°C, failed to pump at
all.
Figures 8, 9, and 10 are plots of temperature corrected flow rates for the MSA
Model G, Nos. 3,4, and 5, respectively. These results are very similar to the
22
-------
0
•
CO
K
Ul
1
2 P
± >
X K
^ r
K
b.
10 O
3 H
O i
< CO
gcs
i--
0 0 0
o e •
MOP
O O O
• o a
CVI
CM
O
CO
o
o
%^
O
/
o
o_
Q
/
a
I
1
"1
1°
I
I
1
a
1
a
n
a *s
< X
1 1
< x
J
T
x<
II
(J
a>
s_
o
o
re
4J
to
a>
-i->
re
o
10
O)
3
CT
-------
CJ
c>
z
o
8
o
(O
/
oo 0
^ 0 o 0 °
1 ^^ ^5 ® f^
- N ro = ^f
+ 1 II
50 x a +
Nl V
50 XQ +
i X N
o n x^
o
2|
cvj
1
< -i CO
/
<
, / =
U- ^i ^
N.
(0
in
fO
CM
QL
O
UJ
CM
O
a>
TD
o
OJ
to
i.
ai
D
E
O)
S-
o
l*-
T3
O
O)
S-
S-
o
.a
(T3
cu
-t-J
(O
o;
r---
01
(9JrUDJ9dUJ8J. JOJ.
("!•"/!) MOHd
24
-------
0 00 0
*
fO
d
CVJ CVJ
+ 11 '
• ? 9 *? ~l
: 1 // 1
UJ
<
5 1
, /
t '•
3 1
/
JK 1
* •
0 1
I
, P /
^ H •
i« 1
IP
t •
?« 1
* z
32 •
u. u. 1
S5 •
3 i
< V)
in
UJ *^ •"* 1
c33-
z
»
•
1 fl 1
U J-
1L J
11 J-
li /
o D / <
I \
1 \
on .
-M'
• ^—
to -S
fO
•M
( y™i
v/ /
o>
4->
13
ce:
o
CM ^~
u_
co
CM
O)
s-
Joj
MOld
25
-------
o
ro
CM
o
d
«
CM
CM
°o° «
o CM *
CMCM 2
1 1 1
o
(O
OJ
a
s-
o
-o
CD
+j
u
0)
S-
o
o
(O
4J
oo
0)
•4-)
03
Qi
O
CT>
0)
JQJ
(uiui/i) MOId dWfld
26
-------
in
d
z
oT
UJ
0
o
2
<
CO
^
0 0
Vo
CM ro
+ i
X
(
• x
JJ
0°
o b
CM —
| 1
on
II
o D
it
o
o
CVJ
CM
(O
oc.
I
•o
o
O)
S-
re
S-
O)
s-
o
o
O)
s-
s_
o
co
IT3
or
o
OJ
3
CT
) MOld
27
-------
results from No. 1 and No. 2. The performances are excellent to about -10°,
and good to marginal at -22°C. At temperatures below -22°C, performances tend
to be marginal.
As temperature is lowered, these pumps require an increased current from the
power source. At 250 mA, the battery pack rated capacity is adequate for an
8-hour operation. At the higher currents required at lower temperatures,
e.g., 280 mA at -30°C, the rated capacity is exhausted in about 7 hours. Thus,
the degraded performance at low temperature is largely due to pump
characteristics. The increased current demand at low temperatures is such
that the battery pack's ampere-hour capacity is exceeded, compounding the
problem.
Flow Rate/Pressure Drop Character!stics--
The flow rate of each Model G was measured as a function of the pressure drop
across the pump and temperature in the basic test apparatus of Figure 2. Each
pump was adjusted for a flow rate of 2&/min at a pressure differential of 4
in. of H£0 at 24°C, and this setting was not changed throughout these
tests. For each test, the pump was equilibrated at the test temperature for a
1.5 hour period, nonoperating. After equilibration, the pump under test was
turned on and the flow rate measured as a function of pressure differential
across the pump. The pumps under test were not operated during the
equilibration periods so that each test would be run with essentially a fully
charged battery pack.
The results of these tests are shown in the data plotted in Figures 11, 12,
13, 14, and 15 for pump Nos. 1, 2, 3, 4, and 5, respectively. In each case,
flow rates and corrected flow rates are shown for temperatures of +24°, 0° and
-20°C. (The 0° and -20°C data are corrected for temperature.) At -30°C, none
of the 5 pumps would operate after equilibrating for 1.5 hours. The pump
motors did operate in most instances, but the rotometers only pulsated and the
pump differential pressure was negligible.
These da.ta show that the Model G pumps can perform a useful function at
temperatures as low as -20°C, and that their performance is at best marginal
at -30°C. It also shows a significant variance among the 5 pumps. Pump No. 3
and Pump No. 5 yielded the poorest and best performances, respectively. These
differences are evident in other data, but perhaps are mos^ obvious here.
Reliability of Calibration over Temperature--
The plots in Figure 16 were derived from the flow rate stability data
presented earlier. Thus, for a given pump, each data point was acquired on a
separate day. The pumps were initially calibrated at 2.1£/roin at 4.0 in. of
H20 at 22°C, and the data plotted in Figure 16 are either the 1 or 1.5
hour data point for each test temperature. (The 1.5 hour data point was
preferred; the 1 hour point is plotted for those low temperature instances in
28
-------
e>
8
o
10
rsi a)
25
of
20
A
10
Differentia
OJ
Q.
O)
>
( NIW /i ) 3ivy MOU
29
-------
CM
d
o
<"
(O
,0 O
^ ^^. A —
7° //«/• A |
O'
/
-° /°
7°
°' /•
O^
oX o'
O UJ
O CC
5f o:
CM o
T a ^
o
e
O
CM
o
IO
in
CM
CM
o
CM
IO
IO
CD
r—
CU
o
to
c
0)
CU
CU
s_
13
CU
Q-
13
00
cu
>
CU
C£
C\J
0)
3
IT)
CM
CM
q
CM'
10
M01J
30
-------
ro
9
z
CO
in
10
CO
O
•o
o
O "
CM ro
/
>v v
• iJr cf
X
a.
<
4J
c
OJ
s-
O)
•I—
Q
(I)
3
10
to
-------
(9
o
Ili
o
UJ
oc
o:
o
o
o
o
*•
CM
O
e
O
CM
\//«
\* * t
/
K' /
o .«\o --/
/
in
ro
O
10
o •
10
cu
in
CU
•§'
o g
CM $:
1 £
M-
.5 t
to
cu
S-
Q-
S-
Ol
cu
4->
(T3
in
cu
S-
CT/
O
c\i
in
CNIW/ i ) 3iva
32
-------
10
d
z
18
o
10
(M
o
c
O
CVJ
If)
un
0
O)
•D
0
(D
CD
s_
1
C/l
S-
cu
TO
O
Lt-
01
q
csi
I ) 3IVH MOld
33
-------
o
e
-o
o
a)
(O
«*-
o
c:
o
u
c
to
(O
(U
4J
to
OL
O
(U
i.
3
cr.
3IVd
34
-------
which a flow rate adjustment was required at the 1 hour point.) Thus, these
data reflect between-days variances as well as variances due to temperature
changes. In view of our experience with the flow rate/pressure drop data
(inoperative pumps at low temperatures), we elected to accept these
between-days variances as an alternative to variances due to battery
degradation which would occur if the pumps were operated at low temperature so
as to acquire these data on a single day.
Figure 16 also shows a considerable variance between pumps. At -22°C, for
example, the measured changes in flow rate vary from 8 to 30 percent. At
-40°C these changes vary from 25 to 60 percent. At -10°C the observed maximum
changes are comparable to the changes permitted by 30 CFR 74.
Continuously Sample Cold Air—
The experimental evidence presented in preceding sections of this report
suggest that the Model G can perform satisfactorily at -20°C, but is marginal
in performance at lower temperatures. Thus, -20°C was selected as a
temperature to test its ability to pump cold air continuously. After all
other testing was completed, all five units were coupled to a manifold at 4
in. of H20 and adjusted to pump 2n/min. The test chamber was set to -20°C
and the pumps operated for 8 hours. This procedure was repeated on 16
separate days after the completion of all other testing. There were no pump
failures during these tests and flow rates remained satisfactory, i.e.,
characteristic of performances observed at about -20°C during all testing.
Ease of Recalibration—•
The flow rate of the Model G was readily adjustable at temperatures as low as
-30°C. The adjustment mechanism permitted precise flow adjustments that were
commensurate with the resolution of the pump's rotometer, e.g., approximately
O.U/min. At -40° and -50°C, the flow tends to pulsate; thus, precise
settings are not possible.
Battery Performance--
All of the personal sampling pumps evaluated were powered by secondary (i.e.,
rechargeable) sealed, sintered plate nickel cadmium (Ni-Cd) batteries. These
batteries have very low internal resistance and can deliver high currents with
little loss of voltage. A very significant feature is that when delivering
moderate currents they will perform satisfactorily at very low temperatures.
These cells can supply useful but reduced energy at temperatures as low as
-40°C. [Ref.33] In this as well as in other respects, Ni-Cd batteries are
superior to other battery systems.
In contrast to the low temperature performance of Ni-Cd cells (e.g., large
currents with good regulation at -40°C) other systems do not perform as well.
The capacity of carbon zinc batteries drops off sharply at low temperatures
due to the increased resistivity and viscosity of the electrolyte, the
decreased rate at which chemical reactions occur, and freezing. The net
result is that carbon zinc systems provide very little service at sub-zero
temperatures. Mercury cells are not nearly as effective at low temperatures
as at room temperature. The zinc anode structure, for example, provides
performance at -40°C equivalent to 20 percent of its room temperature
35
-------
capacity. Alkaline and silver oxide zinc batteries can supply only about 5 to
10 percent of their room temperature capacity at -40°C.
The capacity of a Ni-Cd battery depends largely upon the manner of operating
the battery. Cells are usually characterized at 25°C and at a stated
discharge rate, e.g., C/5 or C/10, where C is the rated ampere-hour capacity.
The actual capacity is the function of the discharge rate and tends to
decrease significantly with an increase in the discharge rate. A cell
characterized at a C/5 rate will yield about 80 percent of rated capacity at a
C/l (1 hour) rate and about 108 percent of rated capacity at the C/10 rate.
The cell capacity is also influenced by temperature. At 0°C, for example, a
Ni-Cd cell will deliver approximately 90 percent of its room temperature
capacity. [Ref.33]
There is very little data available descriptive of the reliability i.e.,
cycle-life of Ni-Cd cells at low temperatures. Data that are available show a
seriously reduced cycle-life at -18° and -34°C; however, these results are
attributable to the establishment of unsatisfactory operational conditions
rather than any failure to the cell itself. [Ref.34] The reader should be
mindful that many factors other than temperature influence the cycle-life and
performance of Ni-Cd batteries. These include use factors such as routine
depth of discharge, recharge practices, and discharge rates.
The results of these evaluations suggest that the Ni-Cd batteries used in
personal sampling pumps yield good performances. While they are inadequate
for the desired performances at low temperature, the problem is largely due to
pump characteristics. Indeed, battery capacity is reduced at low temperature;
however, the pumps demand more current. Thus the batteries must discharge at
the higher rate which further reduces their capacities. These problems are
compounded by the increased current demand of the pumps.
Figure 17 illustrates the increased current demand for the Model G pumps as
temperature is lowered. In supplying a higher current, the 2,000 mA-hr.
battery pack tends to be exhausted in less time and, additionally, the
capacity is decreased at the lower temperature. If, as stated in a preceding
paragraph, the capacity is only 90 percent of its room temperature value at
0°C the Model G battery pack should supply the required current for
approximately (0.9) (2,000)/200 = 9 hrs. At -30°C, it is estimated to last
somewhat less than (0.9) (2,000)7270 = 7.6 hrs. (We have used 0.9 again
because an estimated reduction in capacity is not available for -30°C.) At
-40°C, the estimated time would be significantly less than 5.4 hours.
It is significant that the battery pack voltage was never less than 1.1 V/cell
when replaced by a power supply. This is indicative of the cell's low
internal impedance at low temperatures. Typically, the battery packs were
furnishing currents of, for example, 300 mA at -40°C when replaced by a power
supply. (Since the replacement time was about 1 hour, one can infer that the
capacity degradation at -40°C is about 85 percent.)
It is concluded that the battery packs are inadquate to operate the Model G
pumps at -30°C and below. This is due to battery capacity degradation and the
36
-------
2
<
O
UJ
tr
(T
UJ
O £
CJ 2
i UJ
OJ
TJ
o
OJ
(O
0)
Q.
o
c
o
4->
O
CO
(O
cu
S-
3
3
D-
O
o
o
3
O
O
CM
37
-------
increased current demand with decreasing temperature. However, the pump's
performance was also unsatisfactory to marginal when supplied from a
laboratory power supply. Thus, a larger battery pack will not insure a
significant improvement in performance and is not warranted.
%
Diaphragm Reliability and Bearing Design--
It is noted that all of the Model G pumps survived the low temperature testing
without significant damage. There appears to be some crazing of the plastic
pump housing on one of the five units that may have occurred at -50°C, but the
pumps' performances are not affected. Although the diaphragm reliability
has not been measured, it is observed to be reliable in that there is no
evidence of unreliability after significant testing at low temperatures.
Similarly, there is no evidence of failure or performance degradation due to
bearing design.
Lubricants--
The characteristics of the lubricants have not been explicitly evaluated. It
is considered likely that the lubricant is partially the reason for the pumps'
increased current demand at low temperatures.
Case Ruggedness--
There were no instances of case damage or degradation due to exposure and
operation at the low temperatures. There was some crazing of the pump housing
in one unit (No. 2) , but it had no effect on the pump's operation.
Subsequent operation at -2U°C suggests that the pump remains a reliable,
viable structure. Other instances of crazing are very slight and are
considered insignificant.
MSA Model S
Five MSA Model S CMDPSU units were included in these evaluations. Figure 18
is a photograph of two of these units. The Model S is essentially identical
to the Model G except that the Model S incorporates a by-pass valve designed
to compensate for sample flow restrictions so that the pump can operate more
efficiently. This valve configuration is illustrated in Figure 19. The
by-pass valve allows atmospheric air to flow to the pump; thus, the
differential pressure across the pump is reduced. The valves are mechanically
coupled such that if the by-pass valve is held fixed relative to the case
while the sample valve is turned, opening the sample valve will tend to close
the by-pass valve. The pump's rotometer indicates only the sample flow. The
Model G does not include such a by-pass arrangement. These pumps are
essentially the same if the Model S by-pass valve is completely closed.
Except for this difference the MSA Models G and S appear to be identical with
interchangeable cases, pumps, motors, and battery packs. The Model S pumps
were subjected to identical testing as the Model G. The results of these
tests are discussed in the following paragraphs.
38
-------
Figure 18 Photograph of the MSA Model S
Flow Rate Stability with Time—
The flow rates of the MSA Model S units were measured in the basic test
apparatus, illustrated in Figures 1 and 2, following the procedures described
previously. These results are shown in Figures 20 through 24 for the MSA
Model S units 1 through 5, respectively. As discussed previously, the flow
rates were measured at room temperature and have been corrected to reflect the
lower volume of more dense air that actually moves through the pumps. These
data plots show stable performances at +24°, -2°, and -10°C. At -22°C the
performances tend to be erratic. Satisfactory flow rates could not be
maintained for an 8 hour period in most instances at -22°C. Power supply
substitutions were required in units 1 and 2, and maximum flows in units 3, 4,
and 5, operating from their battery packs, were down to about 1.4 /min after
7.5 to 8 hours. (This is the equivalent of (1.4) (297/251) = 1.66 /min at
+24°C.) In no instance could satisfactory performance be maintained at -30°C
without a power supply substitution.
39
-------
BYPASS
INLET
EXHAUST
SAMPLE
INLET
Figure 19 Illustration of the Model S Bypass Valve Arrangement
40
-------
o
z
CO
UJ
o
o
2E
<
CO
Q.
a.
v>
3p
c >•
0
II
0>
UJ
o
o
0
t
CM
9 ?
CM
Si
o
e
O
o D
//
• o a
I/
•
CO
0)
ro
o;
3
o
o
CM
O)
S-
cn
•r—
u_
-------
CO
tr.
CM
O
01
T3
O
(U
(O
CD
Q
o;
T3
(U
-(->
O
O)
S-
fO
•P
O)
Qi
I
OvJ
O)
s-
Lu
(U|W/|) MQld
42
-------
ro
c/
Ul
Q
O
<
CO
- oo
- N
CO
IT
UJ
CO
O
O)
T3
O
(II
S-
(T3
s.
0)
D
(U
-M
O
0)
S-
o
-Q
re
-4->
t/)
O)
4->
re
a:
o
CM
CM
a>
s-
(8jn4DJ9duiei joj.
43
-------
o
z
Ul
o
o
2
w
v>
K
Ul
M
(U
•o
o
(U
3
-(->
ro
0)
a.
cu
•o
(U
•4->
o
ai
s_
S-
o
o
_a
(O
(U
•4->
(O
Qi
o
CO
C\J
cu
a;
ev
(«jn4DJ9dUJ81 JOJ P9P9JJOO)
( NIW/I ) MOld
44
-------
10
UJ
Q
O
<
V)
2
2
UJ
I-
H
in o
= t
o ?
< •
f\
\ < 0
1 0 <
; \
0
ID
CM
to
(E
UJ
IT)
O
Z
oo
a
•o
o
cu
S-
-p
(O
o
E
s-
o
o
a>
oo
O)
C£
o
-------
Ho
th
It is significant that the Model S units were also tested in an operating mode
at -40° and -50°C for an 8 hour period. Although performances were
unsatisfactory, there is no evidence of damage due to these exposures.
Flow Rate/Pressure Drop Characteristics--
The flow rates of the Model S units were measured as a function of the
pressure drop across each unit at several test temperatures. The procedure
was the same as for the Model G. Each pump was adjusted for a flow rate of 2a
/min at a pressure differential of 4 inches of ^0 at 24°C and each pump
was equilibrated at the test temperature for a 1.5 hour period, non-operating.
The results of these tests are shown in Figures 25 through 29. In each plot
flow rates measured at 24°C, 0°C, and -20°C, and flow rates corrected for the
temperature difference between 24°C and the test temperature are included. As
was the case for the Model G pumps, none of these units would operate at -30°C
and below after equilibrating, non-operating, for 1.5 hours. The motors would
usually operate; however, the rotometers only pulsated and the pump
differential pressure was negligible. These data show 0°C performances
comparable to room temperature performances. There is significant degradation
at -20°C and considerable pump to pump variance.
Reliability of Calibration Over Tempera tu re—
The data plotted in Figure 30 were derived from the flow rate stability data
discussed earlier. Thus, for each pump, each data point was acquired on a
separate date. The pumps were initially calibrated at 2.1Vmin at 4 inches of
at room temperature and the data plotted in Figure 30 are
e 1.5 hour data points for each test temperature. (At -40°C most of the
data is the 1 hour data point because a flow rate adjustment was required
before the 1.5 hour point was reached.) Thus, these results reflect between-
day s variances as well as variances due to temperature. In view of our
experience with the flow rate/pressure drop data (inoperative pumps at low
temperatures), these between-days variances were accepted as preferable to
variances due to battery degredation which would occur if the pumps were
operated continuously while equilibrating.
Data from all five Model S pumps are plotted in Figure 30. The two curves
included are intended to be somewhat characteristic of unit No. 1 as a worst-
case example and unit No. 4 as the best example. These curves suggest that
the flow rate will decrease by about 12 to 19 percent at -20°C and about 19 to
35 percent at -30°C. Inherent in these data are contributions due to battery
degredation, but these were minimized by utilizing the 1.5 hour data.
Continuously Sample Cold Air--
The MSA Model S units were subjected to the same routine of frequent operation
at -20°C, 2£/min and 4 inches of H20 as the Model G units. These
procedures were repeated on 13 separate days. There were no failures during
the 8 hour runs and the flow rates remained characteristic of the -20°C
performances observed during other periods of testing.
Ease of Recalibration--
The flow rate of the Model S is readily adjustable at temperatures as low as
-30°C. At lower temperatures these pumps tend to be either erratic, to
pulsate, or not to pump at all, and flow adjustments become impractical.
46
-------
W
UJ
Q
O
<
CO
o „ o
10
eg
oo
CD
-o
O
O
CM
o I
CM oj
X M-
a.
<
IO
o
0}
to
01
D-
3
10
S-
OJ
(O
o:
o
in
CNJ
I
a-.
Q
CM
IO
47
-------
CM
0
z -o
0)
I
1
LJ O
Q
i ?
< 0 °
CO ft
* V
CM >
\;
o o\» o
• ///
o///
/o» o'
////
////
//// /
/•-/ / /
^ x/ /^ /<
/*//<< /<
»/*//«,« « /
1 1
O 10
CM —
•o
x»°S
Xx I
.X«/o S -
/y/ ° «
//7 )*
//7 / /-
fff / /
V / y
?°/ /
< <
/ /
' /
/
/
r
i
q
o
tf>
J Q
-J 10
CM
O
•z.
ft
oo
'oJ
•o
o
0)
OJ
Q
0)
s-
3
I/)
I/)
01
S-
Q.
S-
-------
ro
d
a.
<
IO
in
co
o
»
oo
O)
-a
o
(O
(1)
M-
ai
s-
10
t/)
a>
s-
0)
(O
ct:
I
CM
01
CD
10
MOId
49
-------
1
i
u
u
(O
cc:
o
oo
CM
s-
3
cr.
q
eg
10
(•uiuj/i) 3ivM MOHd
50
-------
q
cvi
(•Ujui/i) 3ivM M01J
51
Lf)
O
0)
-o
o
IO
d
CO
(O
2
x y
o
CM
*
•
in
0)
s_
3
10
10
ns
o;
o
CO
CJ
en
-------
10
CO
-J « X
UJ
O
— evi ro
iii
• O D
O
10
o *-.
* o
UJ
IT
UJ
o
I
O
ro
oo
03
CtL
O
CO
0)
S-
o-
< xooxa
i 1 I
IO
csi
O 10
esi —
(U|IU/|)
31Vd
52
• •
i I
Q »
Moid
o
^t
•
-------
O
e
UJ
to
(U
TJ
O
Ol
£
c
0>
s_
3
O
cu
(0
CO
Q.
E
ro
(V")
53
-------
Battery Performance--
The battery packs for the Model S units were identical and interchangeable
with the battery packs in the Model G units. The general discussion of Ni-Cd
batteries included in the preceding section on the Model G battery performance
is equally valid for the Model S units. Figure 31 shows the pump/battery
currents observed in the Model S units as a function of temperature. These
results are almost identical to the results observed and plotted in Figure 17
for the Model G units.
Diaphragm Reliability, Bearing Design, Lubricants, and Case Ruggedness--
The diaphragms, bearings, lubricants, and cases of the Model S units are
apparently identical to those of the Model G and the same conclusions apply.
There is no evidence of diaphragm or bearing unreliability. Clearly, either
could contribute to the increased current demand at low temperatures and it is
likely that the lubricant characteristics contribute significantly. These
aspects of the pumps were not explicitly measured. No significant case damage
was observed other than the very slight crazing on the pump housing and a
slight crack at one of the battery pack screws. The latter was very likely
due to over-torquing on the part of RTI.
RAC Model 2392PS
A single RAC personal sampling pump was available for the evaluation program.
This unit has a MESA approval for use in methane-air mixtures and is
configured for use as a 2&/min pump. e.g.. it has a rotometer calibrated
between 1.6 and 2.0 Vmin with a resolution of about 0.1 Vmin. Consequently,
it was tested similarly to the CMDPSU units. The RAC unit also has a similar
series arrangement of rotometer, control valve, damper, and pump as the CMDPSU
units. Figure 32 is a photograph of the RAC unit.
Figure 32 Photograph of the RAC Unit
54
-------
Flow Rate Stability with Time--
Figure 33 is a plot of the measured flow rate corrected for the temperature
difference between room temperature a.nd the test temperature for the RAC unit
as a function of both time and temperature. As illustrated, the room
temperature performance was excellent. At 0°C, the flow rate was
significantly reduced but was stable over the 8-hour period. The flow rate
was about 1.55& /min at room temperature which corresponds to 1.43& /min at
0°C. At -10°C, its performance was marginal and both a flow rate adjustment
and a power supply substitution was made in an attempt to maintain a
satisfactory flow rate. At low temperatures, flow rate adjustments have
little effect on the RAC unit and the power supply substitution was also of
marginal value. At -20°C, a flow rate adjustment at 1.5 hours had no effect
at all. The pump continued to operate from its battery pack at about
1.4A /min (measured at room temperature) for about 5.5 hours. At -30°C, the
flow rate was adjusted to a maximum and a power supply was substituted for the
battery pack after 1.5 hours; thus, a flow of about 1.2^/min measured at room
temperature was maintained. This corresponds to about l.OH/min at -30°C.
At -50°C the RAC failed to pump after 0.5 hours. An examination revealed that
a push rod-socket interface coupling the motor to the pump had separated,
probably because the flexible socket was excessively stiff at -50°C. Normal
operation was restored by simply reinserting the push rod in the socket. It
was also observed that the two halves of a rubber pulsation damper had
partially separated and leaked. There were no further tests run on the RAC
unit.
The flow adjustment on the RAC unit is best described as a pinch or clamp
mechanism that clamps and restricts the sample flow tubing. At low
temperatures the tubing tends to remain "set" and releasing the clamp
mechanism does not provide positive control over the flow.
Flow Rate/Pressure Drop Character!stics--
Figure 34 is a plot of the flow rate of the RAC unit as a function of the
differential pressure across the pump with temperature as a parameter. The
pump was initially adjusted for a flow of 2ji/min at 4 inches of water at 21°C.
It was equilibrated, non-operating, at each test temperature for about 1.5
hours. After equilibrating, the pump was started and a set of data acquired
corresponding to the test temperature. This procedure was repeated for other
test temperatures.
The data in Figure 34 includes the 21°, 0°, -10°, and -20°C curves. Curves
corresponding to both the measured flow rates (measured at room temperature)
and curves corrected to correspond to the test temperatures are included.
Reliability of Calibration over Temperature--
The reliability of calibration of the RAC unit as a function of temperature is
illustrated in Figure 35. These data were acquired as described previously,
i.e., these are the 1.5 hour data points from the flow rate stability
experiments. The data points show flow rate degradations from the calibration
55
-------
O
O
CM
CM
— CM
UJ
«
CO
o—n
Qa
a o
<0
CO
a:
ro
CM
CM
CM
oo
Q_
CM
CO
CSJ
OJ
,0
o
o>
3
-l->
(O
S-
0)
Q.
OJ
s-
o
-a
QJ
u
b
o
o
UJ ^
p 2
-------
ooo oo o o
• o o a «! <
•O O D I
III I
o o o Q o <
/ / J
•o o a a < 4
•6 6 a o < 4
11 11
I I
• p oa p < 5
'! U I I 4
CM
to
i
a.
CM
D_
CM
-------
If)
II
t/1
CL
ro
0)
-O
o
OJ
^ ^
£
UJ
e
oc
u
0.
u
o
T
•P
u
c
tfl
(O
o
rd
(U
•M
fO
a:
IT)
co
q
ca
(U|UI/|) 31VW MCTU
58
-------
point (t=0) of 20, 23, 28, and 42 percent at 0°, -10°, -20°, and -30°C,
respectively.
Continuously Sample Cold Air--
In addition to the .operations for the acquisition of experimental data, the
RAC unit was operated for only one 8 hour period at -20°C. It was not
operating at the end of 8 hours. The -20°C flow rate stability data shows
reasonable performance for about 5.5 hours and the motor was inoperative at 7
hours. From these and other data, we conclude the pump is not suitable for
continuously pumping -20°C air for 8-hour increments.
Ease of Recalibration--
The flow rate adjustment on the RAC unit is a pinch or clamp mechanism that
restricts the tubing that carries the sample from the rotometer to the damper
and pump. The flexible sample tubing passes through this clamp assembly which
acts to restrict the tubing. The degree of restriction is controlled by a
screwdriver setting. At low temperatures the tubing tends to take a "set" and
subsequent flow rate adjustments tend to be ineffective.
Battery Performance--
Batteries in the RAC unit are contained in a molded housing and are not
accessible. From the housing size and open circuit voltage, one can conclude
that the battery pack consists of three size C Ni-Cd cells. Pump currents
were not measured; however, when a power supply was substituted for the
battery pack at -10°C (after 5 hours), -30°C (after 1.5 hours), and -50°C
(after 0.5 hours), the terminal voltage was greater than 1.2 volts per cell
(assuming 3 cells) in every instance. It is significant that the substitution
of the power supply did not restore the desired performance. Thus, while the
battery pack is probably inadequate to provide the desired performance at low
temperatures, it is not the principal limiting factor.
Diaphragm Reliability and Bearing Design--
A subjective assessment of the diaphragm reliability suggests that exposure
and operation at low temperatures can be tolerated. There is no evidence of
diaphragm failure nor is there any evidence of problems with the motor
bearings during these evaluations.
Lubricants--
There was no evaluation of the bearing lubricant. It is a likely source of
pump loading at low temperatures.
Case Ruggedness—
Except for the separated pulsation damper, there is no evidence of damage to
any component in the RAC personal sampling pump.
MSA Model C-200
The MSA Model C-200 personal sampling pump is designed for a flow range of 25
to 200 ml/min through a flow resistance of up to 2.5 in. of HoO. Five
units were provided for the evaluation program. These were adjusted by the
manufacturer for a flow rate of 200 ml/min at 1.5 inches of h^O and most
59
-------
tests were conducted at this setting. The pump incorporates a counter which
indicates the number of pump strokes and, thus, the volume of air pumped.
Figure 36 is a photograph showing two of the C-200's. A small mercury battery
evident in Figure 36 is used in the C-200 in a voltage regulating circuit. It
was removed from the pump for testing purposes.
Flow Rate Stability
Without altering the
manufacturer, these
of 1.5 inches of H^O
tests are plotted in
excellent performanc
-10°C the performanc
unsatisfactory. In
temperatures.
with Time--
flow rate adjustment of the C-200's as received from the
pumps were operated for 8 hours against a flow restriction
at several test temperatures. The results of these
Figures 37 through 41. These curves show generally
es at room temperature and good performances at -2°C. At
es are mixed, i.e., some are marginal and some are
no instances were the performances acceptable at lower
It is important to note that the flow rate data plotted in Figures 37 through
41 are corrected for temperature, i.e., the flow rates were measured at room
temperature and multiplied by the ratio of the absolute test temperature to
the absolute room temperature before plotting. During these tests a flow rate
Figure 36 Photograph of the MSA Model C-200
60
-------
00
o
o
CM
I
o
(D
OJ
(O
•o
IO
Q.
0)
8
CM
I
o
<
CO
UJ
s-
o
TD
0)
•p
o
0)
s-
o
o
IO
(NJ
(O
+->
oo
0)
-p
(O
Qi
CO
OJ
3
-------
CM
O
8
CM
i
O
0 o
o o
* CM
CM
+ I
O
o
O
-T 00
h CM
CM
O
O
O
C\J
O
-a
o
0)
Q.
CO
«— TJ
CD
H
O
O
jQ
(13
•!->
OO
0)
-l->
o:
oo
co
cu
(ajnjDJ«dtu0i
(u|ui/|ui)
62
-------
00
s>fc
CM
+ I
o
o
CM
I
O
• o
•o
•o
II
•o
II
•o
00
IO
CO
oc.
3
o
UJ
CM
in
CM
CM
8 in
N
SIO
CM
~™
c
c
(ainpjddujai joj P3409JJOO)
(ujui/jUJ) MOTJ
co
o
o
o
C\J
i
o
Ol
T3
O
d)
3
-(->
(O
S-
OJ
Q.
d)
•o
OJ
-!->
o
O)
S-
o
o
O)
4->
(O
D;
I
co
o>
63
-------
o
o
K)
CM
e
CM
-i CO
o
o
C\J
I
•«->
(O
S-
0)
Q.
to
O
Id
IO
CM
O
2
•a
IO
O
o
CM
(djnjDjeduiej. JQJ
(ujui/iui) M01J
8
64
-------
o
o
*
CM
+
•
-,00
U
CD
(O
DC.
un
•
o
o
o
CM
I
OJ
-o
o
cu
^
-p
to
o>
Q.
CU
S-
o
«—
oo
cu
+->
to
o:
o
cu
O)
IO
O
IO
IO
CM
(8jn|DJdduj8i joj
(U|UJ/|uu)
65
-------
was also determined by observing each pump's counter for one minute and
calculating a flow rate based upon the manufacturer's calibration. (Number
counts/minute times ml/count = ml/minute.) The manufacturer's calibration was
observed to be very accurate at room temperature and 1.5 inches of ^0,
and corresponded reasonably well with the corrected value at lower tempera-
tures.
A significant factor in the failure of the C-200's to perform satisfactorily
at low temperatures is the mercury (Hg) reference cell used in the pump's
electronics. Two voltages are available from the Hg cell and these were
observed to decrease significantly with a decrease in temperature. In units 3
and 4 at -10°C, for example, the two Hg cell voltages were observed to
decrease from about 4.0 and 2.7 volts to 1.7 and 0.4 volts, respectively, in 1
to 2 hours and the pumps stopped operating.
The single size C Ni-Cd cell was demonstrated to be suitable to operate the
C-200 at lower temperatures. The Hg reference cell was removed from the No. 1
unit and reconnected to the pump with long wires.that facilitated testing the
C-200 in the temperature control chamber with the Hg cell at room temperature
outside the chamber. The results of these tests are illustrated in Figure 42.
(The +24°C data from unit No. 1 are repeated in Figure 42.) The -20° and
-30°C data in Figure 42 were acquired with the Hg reference cell at room
temperature. Although there is significant degradation in the -20°C data, the
pump did operate for 8 hours. During this run the Ni-Cd cell voltage
decreased from an open circuit room temperature value of 1.39 volts (1.38
volts driving the pump at room temperature) to 1.23 volts after 8 hours of
operation at -20°C. At -30°C the Ni-Cd cell voltage was 1.26volts when the
run was terminated after 2 hours. These observations suggest that the Ni-Cd
cell was not exhausted and not the principal reason for the unsatisfactory
performance.
Flow Rate/Pressure Drop Character!'stics--
The flow rate/pressure drop characteristics of the five C-200 pumps are shown
in Figures 43 through 47 for room temperature and 0°C. In Figures 43 and 44,
the flow rates computed from the counters of units 1 and 2, respectively, are
also included. The constant flow rates indicated by the counters show the
pump speed to be constant and indicate that a very stable constant flow rate
can be anticipated for a given temperature and pressure differential. The
measured flow rates at 25° and 0°C are observed to vary reasonably linearly
with pressure differential.
Continuously Sample Cold Air--
The C-200 units will not satisfactorily sample cold air at -10°C or below.
Their ability to sample 0°C air continuously has not been demonstrated but the
experimental results reported herein suggests that it will. There is no
evidence to suggest otherwise.
66
-------
8
CM
i
O
IO
CM
CM
O
O
CM
IO
N
O
IO
•jiUDJ0duj9jL joj
(u|iu/|ui) 3JVM MOld
00
t
3
4->
ra
O)
Q.
(0 o>
CD
10
CC o
O z
Is
c
UJ
o
S
OJ
10
CM
XI
(d
+J
oo
CM
s
67
-------
0
z
•»
8
CM
"S
Ot
o E
10 0
CMU
4-
•
1
i
O
(0
10
O
T3
O
CO
CO
Ol
Q.
CO
M_
Q
CO
CO
s-
(U
>
OJ
CO
0)
=1
CT1
o
o
3ivd MOld
68
-------
a>
u
O O f
O
8
O
8
0
<0
CM
O
O
O
CM
I
O)
-o
O
01
3
I/)
10
cu
s-
D-
H_ -M
O C
* ^ £
ro
CM
I/)
S-
cu
O)
4->
rt)
OJ
3
C7)
(U|Ui/|)
MOHJ
69
-------
•o
«-
o
•
o.° fl
b h
* * t
if
o o -
M> II
I
i o' o
if II
? 0 0 .•
6 ///o
0 9 • S>
/ / / s
/ / / *
oof —
II J
0 0 •/
111
/ / /
o o • —
///
' / '
o o • -
III
/ / /
o o •
/* *
/
' •
'7
1 1 1 1
§ ! i i
N n
*
o
«i
o
o
C\J
OJ
-!->
to
CC.
Li.
CM ir>
«3-
O)
3
O>
LU
O
MOId
70
-------
o
o
(VI
I
o
o o
f
o
o*
o« o
o» o
•o o
•O Q
46 o
70 O
10
10
CVJ
o
o
CM
I
o
0)
-o
o
(U
3
to
to
— £
o £
01
*(0
"5 £
* S
~ 5-
d)
OJ
OJ
-t->
(O
I
Ol
S-
3
1
o
tf)
cvr
/.
1
o
o
CM
1
0
in
"~
i |
o
o
^~
31Vd
71
-------
•o
•°. o S
^g- Q fi A) ^™
^J * ^ ^_
|«o<3
0 0
Jl
8" / /° °
O / / o
? o o' • 8
0 / / •
o o •
o o • -
/ / I
III
0 O •
III
0 0 • -
0 O •
o o • —
HI
001
1 V 1 1 1
go o o
o "° o
CM M ~ —
LO
o
z:
-
O OJ
* c t
*•" Q
Q.
^ 3
O)
to ^
«3
U-
CVI ^
cu
en
u_
O
(u|iu/|) 31VM MOU
72
-------
Ease of Recalibration--
The flow rate of the C-200 is readily changed with the screwdriver adjustment
accessible on the pump case. It is as easily adjusted at -10° and -0°C as at
25°C. The only indication of flow is the counter readout and this is somewhat
less convenient than a rotometer. However, adjustments in the field
environment are more complex with either a counter or a rotometer because the
air density, which changes with temperature, and must be taken into account.
Battery Performance—
The pump motor in the Model C-200 is powered by a single size C Ni-Cd cell
rated at 2,000 mA-hrs. The pump current, according to the manufacturer, is
less than 60 mA and the pump should operate for a period of at least 33 hours,
(i.e., 2,000 mA-hrs.760 mA = 33 hrs). The C-200's were observed to operate
for more than 24 hours at both room temperature and 0°C. A preceding section
of this report stated that a Ni-Cd cell will deliver approximately 90 percent
of its room temperature capacity at 0°C. Thus, one can estimate that the
C-200 will operate for a period of about 30 hours at 0°C. The principal
reason for the C-200's marginal performance at subzero temperatures is the low
temperature characteristics of the Hg reference cell and not the Ni-Cd cell.
Diaphragm Reliability and Bearing Design--
The C-200 is somewhat restricted to temperatures of 0°C and above and there is
no evidence that the diaphragms or bearings were effected by these
temperatures or by their exposure to lower temperatures during testing.
Lubricants--
Lubricants were not explicitly tested during these evaluations. They are a
likely source of problems at subzero temperatures, but not at 0°C and above.
Case Ruggedness--
There is no evidence of damage to any component or the case of the C-200 as
a result of the low temperature exposures encountered during these
evaluations. The appearance and layout of the C-200 leaves an impression of
ruggedness and reliability.
DuPont Model P-125
The DuPont P-125 sampling pump incorporates an electronic flow controller that
senses the air flow and controls the speed of the pump motor to achieve a
constant flow. Other distinctive features include a battery charge status
indicator and a flow monitoring indicator that trips and latches if the flow
is interrupted or cannot be maintained near the calibrated value for a period
of time. The P-125 is specified as adjustable to flow rates of 25 to 125
ml/min with a pressure drop of up to 25 in. of 1^0. Significantly, the
P-125 is specified for an operating temperature range of -7° to 49°C. Figure
48 is a photograph showing two P-125 pumps, one with a cover removed to show
the inside including the On-Off switch. Five units were available for this
evaluation.
73
-------
Figure 48 Photograph of the Du Pont Model P-125
Flow Rate Stability with Time—
The flow rate stability of the P-125's was measured at 100 ml/min and 10 in.
of H20, i.e., somewhat near the middle of the specified capability. These
data are shown in Figures 49 through 53 for the five pumps. (As stated pre-
viously, flow rate measurements were made at room temperature; however, the
values plotted in Figures 49 through 53 are corrected for the difference
between room and test temperatures.)
The performances at room temperature are notably stable for 8-hour periods of
operation. With the exception of unit No. 5, the maximum deviation was 1 per-
cent. Unit No. 5 was 5 percent low, and the low flow indicator was ON (lit)
after 8 hours. The performances were also generally good at a nominal 0° and
-22°C. At 0°C, maximum deviations from 100 ml/min over the 8-hour period
ranged from -7 to +16 percent with an average of +4 and a standard deviation
74
-------
CNJ
(VI
0.
a.
3
-------
>-.
£
a
UJ
1
P
cc
Id
h-
t
m
2
x
^
is
f!
~
(VI
i
a
i
H
Z
O
Q.
8
O "
CM *• 0 rt
CM CM _ }?$
1 + 1
I
1
0 x -
1
1 4
1 (
•
o
r
r
/
i
• 0 X
O 4
<
<
<
1 4
^
1
i
3 X
» o x
> 1
9
3 ^
X
.
00
CM
d
*
in
*• ^1
o.
"cu
T3
O
s:
n
CU
-[_*
<0
Qi
1
-------
£
Q.
IT
Ul
«
5 5
m o
i +
o«
1.
o«
o
s
00
8
00
oo
d
LO
CM
<0
10
0)
(O
Q.
Ul +->
I ^
•- t
o
to
(O
-p
C/1
O)
^
(C
(M
un
O)
CD
MOHJ
77
-------
(T
i
e
i
i
2
£
CvlCM
I I
* *>
3 I
+ i
-i 00
8
(8Jn|OJ8dui8i
(ujui/|UJ)
8
tn
CM
i
Q.
0)
0)
CL
O)
S-
o
M-
(J
Ol
&.
o
o
(O
OJ
4->
fO
o;
i
CO
LO
o;
78
-------
a.
a.
o:
UJ
£
UJ
i
10
cv.
0
0
CM
I
O
3
i
0
0
X <
.° -
v3 ^X
*• oa
1
0-
o
o
oo
V
>a
CO
-------
of +8.7 percent. At -22°C, the temperature corrected flow rates tended to be
high by about 15 to 25 percent, but they were stable over an 8-hour period.
It is notable that three of the five P-125's maintained a useful flow rate at
-30°C for an 8-hour period, and that useful flow rates could be maintained at
-30° and -40°C by substituting a power supply for the battery pack. Thus, it
is evident that the battery pack is a principal performance limiting factor in
the DuPont P-125 at -30° and -40°C. Therefore, useful sampling at these
temperatures may be possible with a larger battery pack. During the -30°C
experiments, battery voltages were monitored on unit Nos. 3, 4, and 5.
In Nos. 3 and 5, the voltages decayed from initial values of about 1.4 V/cell
to about 1.1 V/cell after 8 hours, and the flow rates remained high. In Unit
No. 4, the battery voltage decayed from 1.35 V/cell to 1.02 V/cell after 4
hours and to 0.84 V/cell when a power supply substitution was made, and
acceptable performance restored at 4.5 hours. Comparable observations were
made for all five units at -40°C.
No data was acquired at -50°C. However, all units were maintained in an
operational mode for 8 hours at -50°C. None of the pumps were running after
8 hours. Unit No. 4 was observed to have a ruptured diaphragm and the flow
controller on Unit No. 2 appears to have failed during this exposure. (Unit
No. 2 runs constantly at full speed at room temperature, and there is no
evidence of failure elsewhere in the unit.)
Flow Rate/Pressure Drop Character! stics--
The flow rate of the five P-125's as a function of pressure drop, with
temperature included as a parameter, are shown in Figures 54 through 58 for
Units 1 through 5, respectively. As in previous data, the 0° and -20°C data
are included both as measured at room temperature and as corrected. In the
aggregate, these data show remarkably linear characteristics between +25° and
-20°C. These results are further evidence of the P-125's generally excellent
performance at temperatures as low as -20°C.
A tendency toward higher flow rates at lower temperatures is evident in both
the flow rate/pressure drop and the flow rate stability data. The exception
is the flow rate/pressure drop data for Unit No. 2. These data are also
incompatible with the flow rate stability data for Unit No. 2. No explanation
for this difference is offered except to note that the flow controller on Unit
No. 2 eventually failed.
Reliability of Calibration over Temperature--
The data plotted in Figure 59 are the 1.5 hour data points from the flow rate
stability data. The rationale for selecting these data points was discussed
in a previous section, i.e., these data include day-to-day variances in
preference .to variances due to battery degradation. A single curve
corresponding to the data plotted for Unit No. 1 is drawn and is considered
somewhat characteristic of the other P-125's. It shows a characteristic
increase in flow rate (measured at room temperature) with decreasing
temperature that peaks in the general vicinity of -20°C, and a subsequent
decrease back toward the room temperature value. Unit No. 1, for example, was
80
-------
i
CM
Q.
g
O <
o
?
<
4
4
<
«
I
1
8
CM
•
/
/
/
I
orrected
o
0,
o
8
1 4
J «
1
1
|
<
/
"
<
i
1
•
01
+->
03
CsL
LD
S_
(U|Ui/|UJ) 31VH MOId
81
-------
o
8
II O '
0
1
1
0
1
1
0
1
I C
1
> (
e •< < -i
§!/ 1
* ' 1
> 9* <
« ' J
• 0 «l <
i 1
N 0 0 0 0 « <
a H
• c
K) ,
9
,1
i
«• i ;
L_ 1
I
S 1
'
i
> c
1
> (
1
J
> « <
I!
1 •
' * 11
1
1°
> * < o
i jr
> <» <
i / 1
o o <• <
[
II III
00 < • <
(
1
> <
1
i in \
> < •<
i /i/
6 o < •<
1
i/ 1/
O O < •< -|
1
i/ 1/
I«X1 «q
If 1
0 tf «
11 J
o I** ••*•
1 * 1
1 1 1 1
§ 8 8 g
(U|UJ/|ui) 31VM M01J
82
J ev.
-4 2
-4 w
H
.E I
1/1
&-
O)
Ol
-!->
(O
O£
LO
ir>
OJ
s-
=3
-------
l\L
i ii
< « o • H
i r
i i
1
— (
P <
1
* c
il
||
ill
1 0<
1 0'
1
1
1
> C
!>/•><
m^
> <
<
i
1
> •
ii
1
.
_ .
1
' 1
/
/
> •
i
/
' il
3 O O • -
II
I •
<
-------
/ r
1
^ ^p o 9
/ 1
C> '
Z < O •
: i II
2 .
§
4
<
<
4
I
£
|
o
0 <3 0<
* 8
<3 I <
1
«
«
«
«
«
4
,
1
3 ••
i
40 o
(30 C
30 c
if
30 (
30 <
II
1
3 <0 C
3
3
3
1
1
30 c
1
O
^
A
0 1
O
i •
> •
> •
> •
CJO O
Jl
10 C
,
• •
100 •
1 ' )
•
0
o
K>
CM
_
»
••^
i 1
**r
O
A
LO
CVJ
CM 1~H
"" D-
^
n
OJ
O ^
LO
O)
s_
CL.
1 4J
w -5 £
•£ 5t
I5
< ^
Ol
OJ
<5
U.
r~-
LO
£
3
^ ^
CM
^^
0 0*0 0
O IO O K>
CM — —
(Ujiu/iui) 3iva MOId
84
-------
IO
i
•h
IO
W
Q.
I I
O •
1 J
fO
i
00
LO
s_
3
cn
CM
MOId
85
-------
IO
CM
— CM ro * «
ggggg
• O O < x
(O
s-
+J
O
c:
*
to
(O
UJ
2£
UJ •!-
9 2§ u,
Ul C\J
|_ <1J l-H
ro
o;
(U
-o
O
O)
01
8
8
O
IO
(u|ui/|uj) 3J.VM M01J
86
-------
pumping 119, 136, 132, and 95 percent of its 25°C value at temperatures of
-2°, -22°, -30°, and -40°C, respectively. If these data are corrected forthe
difference between room temperature (the measurement temperature) and the test
temperature, the corresponding readings would be 108, 114, 107, and 74
percent.
Continuously Sample Cold Air--
After the completion of other testing, including an 8-hour period at -50°C in
an operating mode, three of the P-125 units were placed in an environmental
chamber at -20°C on six separate days and observed to operate routinely for 8-
hour periods. (Of the other two units, one had a ruptured diaphragm and the
other a defective flow controller.) There is no evidence of failures
occurring and performances are generally good at -20°C.
Ease of Recalibration--
The flow rate adjustment on the P-125 is located inside the housing and it is
not designed for adjustments in the field environment. Flow rate adjustments
at -20°C were easily accomplished in the laboratory where the flow rate could
be monitored.
Battery Performance--
The P-125's are powered by 4, size AA Ni-Cd batteries. Size AA cells are
characteristically rated about 450 to 500 mA-hr. A distinctive feature of the
P-125 is an LED indicator that indicates that the batteries have an adequate
charge for a day's operation.
The results of these evaluations indicate that the P-125 battery pack is
adequate to operate the pump satisfactorily for 8 hours at -20°C and, with
some selection, to operate the pump satisfactorily at -30°C for 8 hours. At
-30°C, and even at -40°C, pump performance degradation was clearly due to
battery degradation, and performance could be restored by substituting a power
supply for the battery pack. (Three of the five units evaluated performed
well at -30°C using these internal battery packs.) Thus, in contrast to the
other pumps evaluated, the battery pack was the principal performance limiting
factor at low temperatures. One can conclude, therefore, that good
performances can likely be achieved with the P-125 at -40°C with a larger
battery pack. It is not reasonable to consider operating at -50°C from a
battery pack.
Diaphragm Reliability—
A diaphragm ruptured in one of the P-125's while operating at -50°C. The
others survived this exposure without any evident damage. While this fact is
noted, the reader is advised again that no statement of reliability can be
made with any significant degree of confidence.
Bearing Design and Lubricants--
Bearings and lubricants are likely sources of problems at low temperatures.
Lubricants, particularly, become stiff and tend to "load" the bearings or
pump. The fact that the battery pack was a principal performance limiting
factor in the P-125 and that good performance was achieved at -40°C with a
87
-------
power supply suggests that the bearings and lubricants are suitable for use at
these low temperatures.
Case Ruggedness--
Other than the one ruptured diaphragm, there is no evidence of any damage or
degradation to the pump case or components as a result of testing and exposure
to low temperatures during these evaluations.
Accuhaler Model 808
The configuration of the Model 808 differs significantly from the other pumps
evaluated. It utilizes a limiting orifice that attaches integrally to the
pump to control the flow rate. The pump motor drives a pin-cam mechanism
that, in turn, initiates the following cycle of events. An exhaust valve to a
pump cavity opens and a diaphragm (i.e., a piston with a flexible diaphragm
seal) is driven into the pump cavity causing it to exhaust. At maximum
compression, i.e., minimum cavity volume, the exhaust valve closes and the
piston/diaphragm drive mechanism springs back to a null position, opening the
motor circuit. The piston/diaphragm mechanism, driven by a spring, returns to
an extended, maximum volume position as air bleeds into the cavity through
the limiting orifice. At the extended, maximum volume position, the motor is
re-energized to repeat the cycle. Each cycle pumps a volume of air equal to
the displacement of the piston-diaphragm assembly, and the cycle rate is
largely controlled by the limiting orifice. Several sets of orifices are
available and the orifice is readily changed with a wrench or nutdriver.
Figure 60 is a photograph showing two of the Accuhaler Model 808 personal
sampling pumps. A panel covering the batteries has been removed and wires
added to facilitate testing from a power supply.
figure 60 Photograph of the MDA Model 808
-------
Flow Rate Stability with Time—
The flow rate stability of the Model 808's was measured using the test
apparatus illustrated in Figure 61. (The sample flow through the Model 808
pulsates and cannot be measured with a rotometer.) The restricting val\te was
set independently to drop 1.5 in. of H20 at a flow rate of 100 ml/min.
ROOM
SAMPLING
MANIFOLD
DRY COMPRSSED
AIR
\
TEST CHAMBER
PRESSURE
REGULATING
VALVE
SOAP BUBBLE
FLOW METER
PUMP
Figure 61 Test Apparatus for the Accuhaler, Model 808
Sample air was drawn through the restricting valve from the dry air manifold.
When flow measurements were made, the sample was temporarily drawn from the
soap bubble meter.
The limiting orifice that attaches to the pump is the only flow control option
available to the user. For a given orifice, the stability of the pump is
likely to be enhanced if the pressure drop across the orifice is larger than
the drop across the sample collection medium (i.e., the restriction valve in
Figure 61.) During these evaluations, the testing on the Model 808's was done
with the restriction valve in Figure 61 set to drop 1.5 in. of H20 at 100
ml/min, and a 100 ml/min limiting orifice was used on the pump. The flow rate
89
-------
stability and the variations in flow rate with pressure drop across the
restriction valve would be enhanced by a more restrictive orifice. The flow
rate would also be less.
The flow rate stability data for the Model 808 is shown in Figures 62 through
66. The flow rates were measured at room temperature, but the values plotted
are corrected to the respective test temperature. At 25°C, the flow rates are
reasonably uniform and tend to be within about 5 to 10 percent of 100 ml/min.
Unit No. 3 is an exception. Its flow rate was low at all temperatures. Unit
No. 5 is somewhat different as well in that a significantly large peak occurs
early in the test period.
The pumps did not operate for an 8 hour period at 100 ml/min. It is clear
that a lower flow rate would tend to extend the operating period
significantly. The pump operates on a periodic basis—a quasi "duty cycle"--
in that the battery circuit is closed, the motor cycles the cam for one-third
of a revolution, and the circuit is opened until sufficient sample air bleeds
through the collector and limiting orifice to allow the diaphragm to expand
and close the circuit for another cycle. Therefore, the load on the battery
pack is directly proportional to the flow rate, and the period of operation
should also be directly proportional to the flow rate.
The temperature corrected flow rates plotted in Figures 62 through 66 show a
very complex pattern as a function of temperature. In the other pumps, the
flow rates measured at room temperature tended to increase at low temperatures
because the air moving through the pump was colder and more dense than the
measure air. In the Model 808's, the measured flow rates tended to decrease
at low temperature (but obviously less than the temperature corrected values).
There are several likely influences. It is likely that the diaphragm tends to
stiffen at low temperatures and influence the rate at which the
diaphragm/piston "system" returns to a closed-circuit position pulling in a
sample. The motor speed and battery degradation are improbable influences
because the motor can either cycle the cam or it does not.
The orifice characteristic may be an influence. It was observed that the flow
rate/pressure drop characteristics of the orifice are such that, with the flow
rate measured at room temperature, the characteristics do not change
significantly with temperature. (Thus, if the flow were measured at the test
temperature, it would be less by the ratio of the absolute test (sample)
temperature to the measurement (room) temperature for a given pressure drop.)
Thus, one of the two flow limiting factors is an integral part of the pump,
and its characteristics effectively change with temperature.
It is also observed that the time the pump operates decreases with
temperature. Several factors will influence the time of operation. Since the
flow rates decrease with temperature, the motor will cycle less and demand
less of the battery. However, each cycle is likely to require more energy at
low temperatures because of diaphragm, bearing, and lubricant "stiffness."
Moreover, the battery capacity also degrades with temperature.
90
-------
-i 00
g
oT
o
00
tc
UJ
I
i
OT
Ul
I
oc
Ul
i
OT 5
£ OT
O ^*
oo
o
CO
o>
-o
o
0)
3
O)
Q.
d)
OT s_
tr o
.
ui oj
It
(C
4->
00
(O
o:
CM
ID
(1)
S-
CJ)
1
8
/ 1
" o
o
1
o
00
1
o
(0
1
^
(ujui/iui) 3ivy
91
-------
5
ft eg
55 O
•a
o
s:
a>
3
4->
(O
O)
a.
(U
4->
O
-------
to
g
00
or
ui
X
I
fc
3
(0
(E
UI
i
o
c
b.
J3 £
i H
-4
(O
oo
0)
4->
(O
OJ
8
(U|UJ/|UI)
93
-------
(ujiu/iui)
94
-------
IT)
1.
1°
Ul .
35 oo
£8
p oe
F ui
(•jn;aiadui»i
(u|UJ/|ui) M01J dNOd
95
-------
Flow Rate/Pressure Drop Characteristics--
In lieu of measuring the flow rate/pressure drop characteristics of the
cycling Model 808's, the characteristics of two of the limiting orifices were
measured at +23° and -20°C. The 23°C data are shown in Figure 67, and the
-20°C data doesn't differ significantly. These orifices are not critical; the
pump doesn't have that capacity. Nor does the orifice operate at a given
point on a curve, but cycles about an average value.
Reliability of Calibration over Temperature--
The 1.5 hour data points from the flow rate stability experiments are plotted
for the five Model 808 pumps in Figure 68. There is considerable variance
between these pumps, and it would be somewhat arbitrary to state a calibration
reliability figure. A curve is drawn through the data descriptive of Unit No.
2, and it shows a linear decrease in the flow rate (measured at room
temperature) with decreasing temperature, and decreases by about 15 percent
between 25° and -40°C. If the data from Unit No. 3, which tended to have low
flow rates, is neglected, the calibration stability over the temperature range
is reasonable.
Continuously Sample Cold Air--
After the conclusion of other testing, the Model 808's were operated on six
separate days at -20°C. The pump would operate from 4 to 6 hours with the 100
ml/min orifice, and for more than 8 hours with smaller orifices. These pumps
operate satisfactorily at -20°C; the battery is the principal limiting factor.
At lower flow rates, the stability and the time of operation on a single
battery charge will increase.
There is no evidence of damage or degradation due to operations at -20°C,
other than the diminished capacity of Ni-Cd batteries at low temperatures, and
it is believed that the Model 808 can operate at a low temperature for a long
period of time.
Ease of Recalibration--
The only flow control option available on the Model 808 is the optional
limiting orifice that attaches to the pump. Several "sets" of orifices are
available and these are easily changed.
Battery Performance—
The Model 808's operate from two size AA Ni-Cd cells, each with a rated
capacity of 500 mA-hrs. At high flow rates, i.e., at 100 ml/min, the battery
is a principal performance limiting factor. A distinctive feature of this
system is the periodic operation of a drive motor which operates a cam. The
battery pack operates the motor to cycle the cam which, in turn, opens the
battery circuit until the next cycle. Thus, the battery is open circuited
for a time after each cycle. The battery life is dependent upon the rate at
which this cycle is repeated, i.e., the flow rate. The pumps performance is
relatively independent of the battery as long as the battery can cycle the
cam. When it can no longer cycle the cam, the system simply stops. Between
-20°C and +25°C, the Model 808's can operate for about 5 to 7 hours at 100
ml/min. At lower flow rates, it can operate beyond an 8-hour period.
96
-------
8
"
4
Q.
<3
°OL
50
100 ml/min
200
100 150
FLOW (ml/min)
Figure 67 Characteristics of Two Model 808 Limiting Orifices
97
-------
CO
o
00
00
o
00
O)
•o
o
— cj 10 ^j- m
o o d o o
z z z z z
• o
in
OJ
3
-(->
(T3
OJ
Q.
o
I
x<3
u
c
*
I/)
(C
o
to
JO
to
o
(O
D;
I
CO
O)
3
CD
in
o
in
if)
CVJ
(uiuj/|UJ)
M01J
-------
Diaphragm Reliability, Bearing Design, and Lubricants--
There is no evidence of failure or performance degradation that can be
attributed to either the diaphragm, bearings, or lubricants of the Model 808.
It is likely that diaphragm stiffness and cycle-life will change with
temperature, and the bearings and lubricants are also likely to become stiff
and tend to load the motor at low temperatures. These characteristics were
not explicitly evaluated.
Case Ruggedness--
There is no evidence of damage or degradation to the case or other pump
components as a result of the exposures encountered during these evaluations.
Spectrex Model PAS-1000
The Model PAS-1000 is a motorless, diaphragm-type pump driven by an oscillator
and coil at a relatively high frequency. Figure 69 is a photograph showing
two of these units. The wires evident in the photograph were added to
facilitate testing with a power supply other than the pump's batteries.
A maximum flow rate for the PAS-1000 was observed to be less than 800 ml/min.
The flow rate stability tests were run at 200 ml/min at 1.5 in. of H?0,
and the flow rate/pressure drop characteristics were measured with trie flow
rate set at 200 ml/min at 1.5 in. of H2U and also set to a maxmimum.
Figure 69 Photograph of the Spectrex PAS-1000
99
-------
One of the two units was Inoperative when received because of defective, i.e.,
"cold," solder connections. These were repaired to restore operation. One of
the two battery chargers was also defective and would not fully restore the
battery pack. Subsequent to this discovery, the two battery packs were
charged from the same charger.
Flow Rate Stability with Time—
The flow rate stability of the two PAS-1000's is illustrated in Figures 70 and
71. Test conditions were an initial flow rate of 200 ml/min across 1.5 in. of
HpO at room temperature. As illustrated, the 26°C data increased over
the 8-hour period. (It may be significant that the chamber (room) temperature
also increased over that period from an initial 20° to about 26°C.) Unit No.
2 stopped operating after 6.5 hours at 25°C. Battery pack voltage and current
were not being monitored at the time, but subsequent data suggests that this
failure to operate was not a battery pack problem.
At 0° and -10°C, the flow rate on Unit No. 1 changed significantly while
operating from the battery pack. No explanation is offered for the increase
near the end of the day. The test temperatures remained constant and the
battery voltage continued a normal decay. (The terminal voltage decayed from
1.3 to 1.09 V/cell at 0°C, and from 1.32 to 1.04 V/cell at -10°C. over the
test period.) The Unit No. 2 flow rates also decayed very quickly at 0° and
-10°C. At 0°C, a power supply was substituted for the battery pack after 1.5
hrs. when the battery pack voltage was 1.19 V/cell. At -10°C, a substitution
was made after one hour when the battery voltage was 1.2 V/cell. In neither
instance was a significant increase in flow rate achieved. It can be stated
that the rapid decay in flow rate was arrested when the pooler supply was used.
This could suggest that the flow rate was extremely sensitive to power supply
voltages. However, the experiments with Unit No. 1 indicate that it is not.
Flow rate variations with power supply voltages were not measured. At -30°C,
neither unit would pump air. (The test apparatus could indicate flows of less
than 5 ml/min.)
Flow rate stability tests were also run at -22°C. For Unit No. 1, the coolant
source failed and the test temperature rose to 0°C over a 1.5 hour period.
The pump flow rate also increased over this period. When the coolant source
was restored and the temperature reduced again to -22°C, the flow rate decayed
to about 71 ml/min, i.e., except for the perturbation with temperature, the
flow rate was somewhat like the 0° and -10°C performances. Unit No. 2 dropped
to a flow rate of less than 10 ml/min after 2 hrs.
Flow Rate/Pressure Drop Character!'stics--
The flow rate/pressure drop characteristics of the PAS-1000's are shown in the
curves of Figures 72 and 73 for Units 1 and 2, respectively. These data show
the PAS-1000 flow rates to be very sensitive to both temperature and the
pressure drop across the pump. These data were acquired after the pumps were
adjusted for 210 ml/min at 1.5 in. of H20 at 25°C, except for the 25°C
data with the flow control set to a maximum. These latter data were acquired
because the PAS-1000 was advertised as a U /min pump.
100
-------
15
(K
X
UJ
-------
8 8
JOj p»p»jjoo)
102
-------
./
.•y °
o
JL_
CM
00
S-
CM
CM
t-s.
en
•r-
u.
(
-------
o>
8
e
1
O
£
7
•
/
.•
i/
!•/
y*
r\
o
i-
J «'J
/ I
1
o y o*
M
-------
Reliability of Calibration over Temperature-
After 1.5 hours of operation, the measured flow rates of Unit No. 1 were 217
ml/min at 25°C, 142 ml/min at 0°C, and 115 ml/min at -10°C. Thus, the
calibration had decreased by 35 percent at 0°C and by 47 percent at -10°C.
Ease of Recalioration—
Calibration adjustments on the Model PAS-1000 are easily made with a
multiple-turns potentiometer. There is no flow indicator on the unit, so
these adjustments are not routine field adjustments. The calibration does
vary with time and temperature.
Battery Performance—
The PAS-1000's are powered by 7, size C Ni-Cd batteries. There is no evidence
to suggest that the pumps' performances were negatively influenced by battery
degradation. Terminal voltages were high when power supply substitutions were
made, and improved performances did not result from these substitutions.
Diaphragm Reliability-
There is no evidence that the diaphragm was damaged or degraded as a result of
the testing and exposures encountered during these evaluations.
Bearing Design and Lubricants—
The oscillator/coil design of the PAS-1000 eliminates the usual motor and pump
bearings. The oscillator/coil construction should reduce relative motion
between components to a minimum and reduce low temperature problems associated
with bearings and lubricants.
Case Ruggedness—
There was no case or component damage or degradation evident as a result of
these tests and exposures.
Si pin Model SP-1
A single Model SP-1 available for evaluation had a minor binding problem that
may have influenced some of the experimental results. A second unit was
received as a replacement but not in time to be tested. Figure 74 is a
photograph of the Model SP-1.
Flow Rate Stability with Time—
The Model SP-1 flow rate was measured at several test temperatures with the
flow control set to a maximum, i.e., approximately 240 ml/min, and a pressure
differential of 1.5 in. of H20. Figure 75 shows the measured flow rates,
corrected for the difference between the test temperature and measurement
(room) temperature, as a function of time at 25°, 0°, and -20°C. Flow rates
computed from the pump's counter is essentially the same as the measured and
corrected flow rate plotted, and a separate curve is not drawn.
105
-------
-Figure 74 Photograph of the Sipin Model SP-1
At 25°C, the flow rate remained stable over the 8-hour period. The failure to
operate after 5 hours at 0°C is thought to be due to a binding problem evident
at the pump-counter interface. This problem, clearly evident at times, was
not always evident. It is difficult to estimate what influence it had on
other performances. At -20°C, the measured flow rate had the unusual charac-
teristic illustrated. The flow rate computed from the pumps counter does not
have the same anomaly. Except for the anomaly, the measured and corrected
flow rates and the flow rates computed from the counter compare very closely.
The Model SP-1 did not operate long at -30°C, probably because of the effect
of temperature on a Hg reference cell used in the electronics.
106
-------
00
Q.
oo
QJ
-o
O
CD
S-
(O
01
CL
O)
-------
Flow Rate/Pressure Drop Characteristies--
The flow rate/pressure differential character!si tics of the Model SP-1 are
included in Figure 76. The values plotted are corrected to correspond to the
measurement (room) temperature. The flow rates vary linearly with pressure
differential. (At all test temperatures, and the measured flow rates would
appear much more uniform.)
Reliability of Calibration over Temperature--
The measured 1.5 hour data points from the flow rate stability data show the
flow rate to decrease from the 25°C value by 7 and 17 percent at 0°C and
-20°C, respectively.
Continuously Sample Cold Air—
The Model SP-1 was operated at -10°C on several days after the completion of
other testing. Its performance, influenced by the binding problem described
earlier, was somewhat erratic, i.e., it's period of operation varied
significantly from day to day.
Ease of Recalibration—
The flow rate of the Model SP-1 is easily adjusted with a screwdriver
adjustment control. Reference numbers are on the control, and these could be
used to facilitate field adjustments. However, there is no indication of flow
rate other than the counter.
Diaphragm Reliability, Bearing Design, and Lubricants--
The diaphragms, bearings, and lubricants were not explicitly evaluated. There
is no evidence of diaphragm or bearing degradation or failure. The binding
discussed earlier in this section occurred in a slotted fitting interfacing
the pump and counter and not in a motor or pump bearing.
Case Ruggedness—
There is no evidence of case or component damage as a result of the low
temperature exposures encountered during these evaluations.
SUMMARY
All of the personal sampling pumps evaluated are sensitive to temperature
changes between 25°C and -50°C. There is considerable variance in the
sensitivity of the different models evaluated, and some variance within a
single model. All of the pumps evaluated are useful sampling instruments at
0°C, and none are functional at -50°C. (Pumps which perform poorly at 25°C
also perform poorly at 0°C.) Between these extremes, the different models are
effected differently by temperature. From among these pumps, the industrial
hygienest can select a pump suitable for most any application to a low
temperature of -20°C. At lower temperatures, his options are severely
limited.
None of the pumps tested were useful at -50°C. The Ni-Cd batteries used in
these personal sampling pumps are not useful at this low temperature. (We do
not know of any other battery system, primary or secondary, that is superior
to the Ni-Cd system for this application.) If a suitable power source were
108
-------
o o
to
( 0Jn|DJ9dW9J. JOj. P9J09JJOQ)
109
-------
available, the pumps themselves would not be suitable. None would operate
reliably and most would not operate at all.
At -40°C, the Model P-125 would provide some sampling capability, i.e., a
reasonable but somewhat erratic flow rate for a period of 2 to 4 hours. With
an adequate power supply, e.g., a larger Ni-Cd battery capacity or a remote
battery pack maintained at a higher temperature near the body, the -40°C
sampling may be reasonably good.
At -30°C, the Model P-125 can provide a useful sampling capability for about 8
hours on its internal battery pack. (Some selection of battery packs and
pumps may be necessary.) The MSA Models G and S are marginally useful for the
higher, nominally 2a /min flow rates, and the MDA Model 808 is marginally
useful for lower flow rates, i.e., less than 100 ml/min. The Models G and S
flow rates will be significantly lower than 2a /min, and they will operate for
about 7 to 8 hours. The MDA will have a maximum flow rate of about 50 ml/min
and the period of operation will be less than 4 hours.
At -20°C, the MSA Models G and S, the DuPont Model P-125, the Model 808, and
the Sipin Model SP-1 will all provide a useful, 8-hour sampling capability.
The Models G and S and the P-125 will perform reasonably well as compared to
their 25°C capabilities. The Model 808 will be limited by battery capacity if
an 8 hour period of operation is desired. The Model SP-1, which may have been
limited by an uncharacteristic anomaly, will provide a reasonable pumping
capability at -20°C. The Sipin has a Hg reference cell which does limit its
performance at lower temperatures.
It is emphasized again that a comparison of the pumps evaluated was not an
objective of this effort. The pumps are rated or specified differently by
their manufacturers and, consequently, comparison is somewhat unfair. This is
especially true of the pumps which were not tested as CMDPSU units. However,
in our opinion, there is an obvious quality of construction and room
temperature performance about some of these pumps which should be
acknowledged. The MSA Models G and S, the MSA Model C-200, and the DuPont
P-125 perform well at room temperature and give an impression of quality
construction. (The C-200, however, should not be considered for application
below 0°C because of the Hg reference cell.) This is not to say that the
other pumps are not adequately constructed and some advantages cited for each.
The SP-1, for example, is well constructed, and the simplicity of the system
can be advantageous. The Model 808 has a different operating mechanism which
could have some advantages. It should, for example, be relatively insensitive
to battery status until the pump fails to operate at all. The Model 2392-PS
is similar in many respects to the Models G and S. It is lighter, has a
smaller battery pack, and performs well for 8 hours at room temperature. The
Model PAS-1000 is significantly different from any of the other pumps. It
does not contain a conventional motor driven pump, but operates a diaphragm
pump in a small displacement/high frequency mode. There are some inherent
advantages to such a system at low temperatures. However, in the present
model, the pump is very sensitive to pressure differential and the high
frequency sound can be somewhat objectionable.
no
-------
RECOMMENDATIONS
We concluded that the personal sampling pumps evaluated were not designed for
use in sub-zero environments. Instead, they were designed and optimized for
use in temperatures near 25°C. We conclude further that, in response to the
solicitation for pumps for this evaluation, the standard models were submitted
without significant changes to optimize these units for low temperatures
because of insufficient time and incentive. Consequently, standards based
upon the results of these evaluations will tend to be premature and somewhat
pessimistic. Standards recommended in a subsequent section should be con-
sidered as preliminary, subsequent to further investigations and develop-
ments .
INVESTIGATIONS
The Working Environment In Cold Climates
The Alaskan environment has been adequately defined in this report, and the
interested reader can find a more detailed description in the references
cited. What is needed is a definitive description of the Alaskan working
environment. It is doubtful that personnel work in -50°C temperatures, for
example. Moreover, at low temperatures, their options for adjusting pump
flows or observing operations will be significantly limited. In a cold
environment, a useful mode of operation for a personal sampling pump would be
to protect the unit by wearing it underneath the outer clothing. As a
minimum, the pump could be configured so that the batteries could be worn near
the body and thus maintained at a warmer temperature.
It is recommended that the working environment encountered in the Arctic Area
be investigated, along with various options regarding the use of personal
sampling pumps in such an environment. These investigations would involve
placing a knowledgable observer in that environment for several weeks at
select periods to evaluate the environment and further test the use of
personal sampling pumps. Selected pumps, configured for optional use
practices, would be worn and monitored in the working environment. A small,
portable, controlled environment instrument package could be configured to
monitor the pump's performances. These investigations would enhance the
development of reasonable standards pertinent to the use practices and
performances of personal sampling pumps in cold environments, and it is
recommended that they be conducted. As a part of these investigations, the
influence of the cold, dense air on the distribution of contaminants and its
significance in terms of toxicity and detector tube responses should also be
considered.
Battery Investigations
There is a dearth of information pertaining to the characteristics of Ni-Cd
batteries at low temperatures. Since these batteries are the best choice for
operating personal sampling pumps and other portable apparatus, it is reason-
able to investigate their characteristics between room temperature and -50°C.
Ill ::
-------
It Is recommended that a compliment of commercially available Nl-Cd batteries
in all of the standard sizes be tested to rate their performances as a func-
tion of temperature and discharge rate as compared to their performances at
25°C and ten-hour discharge rate. Ideal maintenance practices should be
followed. It is further recommended that the cycle life of a complement of
Ni-Cd cells be measured at the ten-hour discharge rate as a function of
temperature. The capacity of the cells under test should be derated to
correspond to the capacity at the test temperature as determined in the
preceding investigations.
Sampling Devices
These investigations have not addressed the issue of the effect of temperature
on the various sampling devices which are used with personal sampling pumps.
The sampling devices are an integral part of the system, and it is recom-
mended that these be characterized at low temperatures in a subsequent study.
DEVELOPMENTS
There is much that can be done to enhance the performance of personal sampling
pumps at low temperatures. A thoughtful selection of materials, components,
and lubricants with low temperature performance as an objective is an obvious
initial step: It may also be advantageous to configure the pump differently,
e.g., such that the battery pack is separate and can be selected to provide
different capacities. Subsequent to the investigations prepared in the
preceding section, it is recommended that NIOSH sponsor the development of a
pump optimized for use in the working environment encountered in cold
climates. This environment would be defined as a result of the recommended
investigations.
The recommended development effort is not envisioned as an extensive under-
taking, and the manufacturers of promising pumps may undertake the project on
their own resources if a use environment and a potential market is defined.
In either event, the resultant pumps would have to be reevaluated with the
defined environment as a test condition. If a suitable pump did not evolve,
further development efforts would need to be considered. The results of the
evaluations described herein suggest that personal sampling pumps suitable for
use in -30°C environments can be obtained with a little effort. (The Model
P-125 may be suitable now with selected batteries.)
RECOMMENDED USE TECHNIQUES
For the personal sampling pumps evaluated, it is recommended that they not be
used at temperatures lower than tabulated below:
112
-------
Pump Low Temperature Limit (°C)
MSA Model G -20
MSA Model S -20
RAC Model 2392-PS 0
MSA Model C-200 0
Du Pont Model P-125 -30
MDA Model 808 -20
Spectrex Model PAS-1000 0
Si pin Model SP-1 -10
It is further recommended that each pump be calibrated for the desired flow at
the use temperature. To the extent that the pump's environment can be
moderated (by wearing under clothing, for example) it should be done. However,
the moderated environment needs to be defined. It is conceivable that the
working environment need not be moderated for some of the pumps evaluated.
The manufacturers use and application instructions should be considered as a
guide and followed to the extent possible. The user should be mindful that
these instructions assume near room temperature conditions. In the inter-
pretation of sampling results, the effects of a more dense air sample should
be considered. The effect of air density on contaminant levels and detector
responses has not been considered herein.
The battery pack should be used with care regardless of temperature. Recharg-
ing should be carried out with the charger supplied by the manufacturer
according to instructions and at room temperature. In lieu of this option,
the batteries should be charged at a constant C/10 rate for about 15 hours, or
at a C/20 rate for much longer periods. (C is the ampere-hour rating of the
battery.) The battery should not be recharged unless a significant portion of
its capacity has been dissipated, e.g., C/3 ampere-hours.
113
-------
STANDARDS RECOMMENDED BY RTI
The standards recommended below are an adaptation of the CMDPSU standards in
30 CFR 74 to personal sampling pumps for use in cold environments [Ref. 1].
These standards reflect the needs of OSHA and industry for pumps suitable for
sub-zero industrial hygiene surveys as reflected in 30 CFR 74 and the
performances observed in the pumps evaluated. These recommended standards are
arranged in a numbered paragraph format to conform to the format usually found
in the Federal Register. A statement of rationale is included in those
instances where it is considered helpful. With reference to 30 CFR 74,
references to the sampling devices have been deleted. There is an obvious need
to evaluate the performances and responses of sampling devices and detectors in
low temperatures, but these were not objectives of this evaluation program.
1. Purpose
The regulations in this part set forth the requirements for approval of
personal sampling pumps designed for use in sub-zero environments.
2. Sampler unit
The personal sampling pumps shall consist of a pump unit adaptable to a
sampling head assembly and, if rechargeable batteries are used in the pump
unit, a battery charger.
3. Specifications of the personal sampling pump
(a) Pump unit—(1) Dimensions. The overall dimensions of the pump, hose
connections and valve or switch covers shall not exceed 8 inches in height, 6
inches in width and 4 inches in thickness.
(2) Weight. The pump unit shall not weigh more than 4 pounds.
(3) Construction. The case and all components of the pump unit shall be
of sufficiently durable construction to endure the" wear of use in sub-zero
environments to -30°C. (Rationale: The pumps evaluated did not perform
acceptably at temperatures below -30°C. Thus, durability at -30°C is a
suitable objective.)
(4) Exhaust. The pump shall exhaust into the pump case, maintaining a
slight positive pressure which will reduce the entry of foreign materials into
the pump case. (Rationale: This requirement, repeated here from 30 CFR 74,
does eliminate positive pressure applications such as filling gas bags.
However, it seems as important a consideration in an environment of blowing
snow as in a coal-dust environment. Moreover, the emphasis in these tests was
personal sampling.)
(5) Flow rate adjustment. The pump unit shall be equipped with a
convenient means of flow rate adjustment. To prevent accidental adjustment,
the flow rate adjuster shall be inside the pump case or recessed in the pump
case and shall require the use of an adjusting tool. (Rationale: This
requirement is intended to be less restrictive than the similar requirement in
30 CFR 74. All pumps will not have a flow rate indicator. However, all pumps
shall provide a means of adjusting flow rates either in the field or in a
laboratory.)
114
-------
(6) Battery. The power supply for the pump shall be a suitable battery
pack located in the pump case or in a separate case which attaches to the pump
case by a permissible electrical connection.
(7) Pulsation. When a pump's flow rate is 1.0 £/min or greater, the
irregularity in flow rate due to pulsation shall have a fundamental frequency
of not less than 20 Hz. This requirement shall not apply when the flow rate is
less than 1.0 &/min.
(8) Belt clips. The pump unit shall be provided with convenient means
of securing the pump to a wearer's clothing.
(9) Recharging connection. A suitable connection shall be provided so
that the battery may be recharged without removing the battery from the pump
case or from the battery case if a separate battery case is used.
(10) Flow rate indicator. Pumps specified by their manufacturers for
flow rates greater than 1 /min at room temperature shall have a flow rate
indicator as an integral part of the pump unit and the flow rate indicator
shall be calibrated within +/- 5 percent for flow rates between 1 and 2 1/min
and temperatures between 40° and -20°C. The manufacturer shall provide
temperature and pressure operation ranges and correction factors. Pumps
specified by the manufacturers for flow rates less than 1 1/min at room
temperature shall provide for an indication of the total volume of air pumped,
or an indication that the pump has failed to maintain a selected flow rate
over the sampling period. In these instances, the manufacturer shall also
provide temperature and pressure operation ranges and correction factors. The
specified flow rates shall correspond to a specified maximum pressure
differential.
(11) Flow rate range. There is a need for a wide range of sampling flow
rates. Manufacturers of pumps approved under this part shall specify a range
of flow rates that can be maintained for 4- and 8-hour periods at specified
pressures by the pump, and the pump flow rate shall be adjustable over this
range.
(12) Flow rate consistency. The 8-hour flow rate shall remain within
+/- 15 percent of a specified value over an 8-hour period and over a
temperature range of 40° to -20°C. (Rationale: The better pumps evaluated
conformed to this requirement.)
(13) Duration of operation. Pumps approved under this part shall be
specified for either 4 or 8 hours of operation at the specified flow rate and
pressure differential across the pump in the temperature range of 40° to
-20°C.
(14) Battery charger. The battery charger shall be operated from a 117
volt, 60 Hz power line. It shall be provided with a cord and polarized
connector so that it may be connected to the charge socket on the pump or
battery case. The battery charger shall be fused, shall have a grounded power •
plug, and shall not be susceptible to damage by being operated without a
battery on charge. The battery charger shall be capable of operating at either
a 16- or 64-hour charge rate. It shall be capable of fully charging the
battery in the pump unit in the stated times and shall not overcharge a
discharged battery in 16 hours when operating at 16-hour charge rate or in 88
hours when operating at the 64-hour charge rate.
115
-------
REFERENCES
1. Code of Federal Regulations. Title 30, Part 74.
2. Nelson, 6. 0. 1972.Controlled Test Atmospheres. Principles and
Techniques. Ann Arbor Science Publishers, Inc. Ann Arbor, Michigan.
3. Berry, Mary Clay. 1975. The Alaska Pipeline: The Politics of Oil and
Native Land Claims. Indiana University Press, Bloomington.
4. Caldwell, Lynton K. 1971. Environment: A Challenge for Modern Society
Doubleday & Co., Anchor Books, Garden City, New Jersey.
5. Climatography of the United States. September, 1959. U. S. Dept. of
Commerce N060-29.
6. "Alaska." 1969. The World Book Encyclopedia. Field Enterprises
Educational Corporation.Vol. 1: A. Chicago, Illinois.
7. Bryson, Reid A., and Hare, F. Kenneth. 1974. Climates of North America.
(World Survey of Climatology, Vol. 11). Elsevier Scientific Publishing
Company. New York.
8. Climates of the States. 1974. U. S. Dept. of Commerce, NOAA, Washington,
D.C.
9. Cicchetti, Charles J. 1972. Alaskan Oil; Alternative Routes and
Markets, Resources for the Future, Inc., Washington, D. C.
10. Gardey, John. 1976. Alaska: The Sophisticated Wilderness. Stein and
Day. New York.
11. Local Climatological Data. 1965. U. S. Dept. of Commerce, Weather
Bureau.
12. "Alaska Briefs." October, 1975. Weatherwise 28:93, 135+, 183, 227+,
April - August.
13. "North Alaska Briefs. December, 1975 - February, 1976. " Weatherwise.
Vol. 28: 278-9, Vol. 29: 49+.
14. Jowett, C. E. 1973. Electronics and Environments, John Wiley & Sons,
New York.
15. Bell, J. H., Jr. 1963. Cryogenics Engineering, Prentice-Hall, Englewood
Cliffs, New Jersey.
16. Wigley, D. A. 1971. Mechanical Properties of Materials at Low Tempera-
tures. Plenum Press. New York.
17. Engineering Design Handbook, Environmental Series, Part TWO; National
Environmental Factors. April 1975. AMC Pamphlet No. 706-116. Head-
quarters, U. S. Army Material Command, Alexandria, Virginia.
18. Martin, H. L., et.al. 1968. Effects of Low Temperatures on the Mechan-
ical Properties of Structural Metals. NASA SP-5Q12(01). for NASA.
Office of Technology Utilization.
19. Symposium on Effects of Low Temperatures on the Properties of Materials.
March 19, 1946. Spec. Tech. Publn. No. 78. Philadelphia: ASTM.
20. Dummer, G.W.A., and Griffin, N. 1960. Electronic Equipment Reliability
John Wiley & Sons, New York.
116
-------
21. Levy, Herbert. April, 1965. "Wirewound Resistors." Electronics World
V. 73, No. 4.
22. Ward, Donald 0. April, 1965. "Common Problems in Specifying Resistors."
Electronics World. Vol. 73, No. 4.
23. Electronic Components Handbook. 1957. Electronic Components Lab,
Wright Air Development Center. McGraw-Hill.
24. Nieders, H. July, 1965. "Electrolytic Capacitors." Electronics World
Vol. 74, No. 1.
25. "Ceramic Capacitors." July, 1965. Electronics World. Vol. 74, No. 1.
26. Reference Data for Radio Engineers. 1972. ITT. Howard W. Sams & Co.,
Inc., 5th edition, New York.
27. Lamphier, Walter C. July, 1965. "Plastic-film Capacitors."
Electronics World. Vol. 74, No. 1.
28. Robinson, William M. July, 1965. "Paper Capacitors." Electronics
World. Vol. 74, No. 1.
29. Rothenstein, E. M. July, 1965. "Mica Capacitors." Electronics World.
Vol. 74, No. 1.
30. Linear Intergrated Circuits. February, 1975. National Products
Catalog.Santa Clara, California.
31. Jolly, W. P. 1972. Cryolectronics. John Wiley & Sons. New York.
32. Rittenhouse, John P., and Singletary, John B. 1969. Space Materials
Handbook. NASA SP-3051. For NASA, Office of Technology Utilization.
33. Electronics World. October, 1965. (Special Issue on Batteries).
Vol. 74, No. 4.
34. Bauer, Paul. 1968. Batteries for Space Power Systems. (NASA SP-172).
TRW for NASA. Redondo Beach, California.
117
-------
APPENDIX
TABLE 1A
COMPANIES RECEIVING REQUESTS FOR LITERATURE DESCRIPTIVE
OF PERSONAL SAMPLING PUMPS
Company
Air Sampling Division, OR
The American Society of Safety Engineers, Inc., IL
Anatole J. Si pin Company, NY
Anderson 2000, Inc., GA
Bacharach Instrument Company, PA
Bendix Corporation, MD
The Canadian Society of Safety Engineers, Inc., Ontario
Century Systems Corporation, KS
CSE Corporation, PA
Denver Equipment Division, CO
Devco Engineering, Inc., NJ
E. I. du Pont de Nemours & Company, DE
Edmont Wilson Company, OH
Industrial Products Company, PA
Industrial Safety Equipment Association, Inc., VA
Joy Manufacturing Company, CA
Lefco Engineering, CA
Matheson Gas Products, NJ
Microchemical Specialties Company, CA
Mine Safety Appliances Company, PA
Quality Control Equipment Company, IA
Research Appliance Company, PA
Safety Equipment Distributors, Inc., VA
SKC, Incorporated, PA
Spectrex Company, CA
Staplex Company, NY
Thermo Systems, Inc., MN
Weather Measure Corporation, CA
Will son Products Division, PA
119
-------
RESEARCHTRIANGLEINSTITUTE
POST OFFICE BOX 12194
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27709
SYSTEMS AND MEASUREMENTS DIVISION
Gentlemen:
The Research Triangle Institute (RTI), under contract to the National
Institute for Occupational Safety and Health (NIOSH), is planning an
evaluation of personal sampling pumps suitable for use in cold, sub-zero
environs such as the Alaskan oil fields. Pumps such as the Coal Mine Dust
Personal Sampler Units (30 CFR 74) and Detector Tube Units (42 CFR 84) are
especially of interest for the planned evaluation. Other pumps and proto-
types which are suitable for use in low temperature environs will also be
solicited.
The evaluation program will be conducted under NIOSH Contract No.
210-76-0124. The NIOSH Project Officer is Mr. Charles S. McCammon, DPSE-
MRD, Robert A. Taft Laboratories, 4676 Columbia Parkway, Cincinnati, Ohio
45226, Telephone (513) 684-2591. The Project Director at RTI 1s Dr. Richard
Whisnant, and the Project Leader is Mr. Carl D. Parker.
Manufacturers of personal sampling pumps are urged to participate in
this evaluation program. Vie are initially seeking to determine the avail-
ability of personal sampling pumps by requesting brochures, specifications,
and other descriptive Information. Pumps will be selected for evaluation
from the information received. Subsequent correspondence will include
additional information relative to the evaluation program and request spe-
cific pumps for evaluation. We will appreciate receiving this Information
as soon as possible and no later than July 12, 1976. It should be sent to
the attention of the undersigned.
Your response to this request for descriptive literature and your
cooperation with the planned evaluation program will be appreciated. If
you desire additional information, please call the undersigned at
(919) 549-8311.
Very truly yours,
Carl D. Parker
Systems & Measurements Division
CP/smk
120
(919) 549-8311 FROM RALEIGH, DURHAM AND CHAPEL. HILL.
-------
RESEARCH TRIANGLE INSTITUTE
POST OFFICE BOX 12194
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27709
SYSTEMS AND MEASUREMENTS DIVISION
Gentlemen:
We appreciate your prompt response to our recent request for information
descriptive of available personal sampling pumps. We have reviewed the infor-
mation received and concluded that your Models are suitable for the purposes
of the evaluation program. The purpose of this letter is to provide you with
additional information about the planned program and to request your further
participation.
The Research Triangle Institute is conducting this evaluation under con-
tract to the National Institute for Occupational Safety and Health (NIOSH).
The NIOSH Contract No. is 210-76-0124 and the Project Officer is Mr. Charles
McCammon, Robert A. Taft Laboratories, 4676 Columbia Parkway, Cincinnati,
Ohio 45226, (513) 684-2591. The Project Director at RTI is Dr. Richard A.
Whisnant and the Project Leader is Mr. Carl D. Parker.
With the development of United States energy resources in sub-zero envi-
rons, such as in Alaska, there is an increasing need for information relative
to the current air sampling instrumentation and methodology avaflable for in-
dustrial hygiene investigations in sub-zero weather. In field use, cold tem-
peratures may alter the response of such air sampling instruments as personal
sampling pumps. Some basic problems associated with the operation of air sam-
pling instruments in sub-zero environs are, for example, greases and oils freez-
ing, bearings freezing up, piston units freezing and sticking, pump diaphragms
becoming brittle, valves sticking, and decreased battery efficiency. Since it
is necessary to use personal sampling pumps for field industrial hygiene sur-
veys including compliance work by OSHA, data must be obtained to establish the
reliability and operating characteristics of personal sampling pumps in these
sub-zero environs.
The temperature range of interest will be from room 21°C to a lower limit
of -50°C. It is not necessary for a pump to be functional to the lower limit
to be of interest. Further, since battery deficiency will limit the low tem-
perature operation of most pumps to a much higher temperature than -50°C, some
experimental evaluations will be completed using a laboratory power supply
rather than batteries.
121
(919) 549-8311 FROM RALEIGH, DURHAM AND CHAPEL
-------
Page 2
If you have explicit recommendations to enhance the operation of your pump
at(low temperatures, please call them to our attention. If the pumps evalua-
te'!! differ in any way from those obtained with a routine purchase order, it is
essential that these differences be called to our attention.
We are requesting that five (5) units of your Models be sent to us for
evaluation at your earliest convenience, and no later than September 20, 1976.
These should be shipped to RTI to the attention of Mr. Carl D. Parker. Shipment
may be made C.O.D. (include insurance). RTI will return the pumps F.O.B. your
plant (with full insurance coverage).
The following points are also relevant to the evaluation prog~am and are
called to your attention:
1. Pumps submitted for evaluation must be accompanied by a statement
excluding RTI and NIOSH from responsibility for damages which might
occur during the evaluation program. Necessary accessories should
also be included. A complete operational manual featuring operating
instructions, flow diagrams, and circuit diagrams, for example, are
required for the evaluation program and should be included. Material
routinely supplied to purchasers of the instrument should be clearly
identified.
2. Based upon the results of the evaluation program, RTI will recommend
to NIOSH a set of performance standards and quality control standards
applicable to personal sampling pumps for use in sub-zero environs.
Participants will have an opportunity to comment on these recommenda-
tions.
3. Each participant is entitled to see the results of the evaluation of
his pump only. Results of tests on other pumps will be kept confi-
dential until governmental publication of test results.
4. The results of the preliminary evaluation released to the participant
by RTI may not be used for advertising purposes or as a claim of en-
dorsement by NIOSH. Any publication resulting from these evaluations
can, of course, be referenced.
Your prompt reply to this invitation and your cooperation in this evalua-
tion program will be greatly appreciated. If you desire additional infor-
mation, please call the undersigned at (919) 549-8311 or Mr. C. McCammon at
(513) 684-2591.
Very truly yours,
Carl D. Parker
Project Leader
CDP/js
122 -fe U.S. Government Printing Office: 1978-757-141/6773, Region No. 5-11
------- |