SEPA
United States Office of Air Quality EPA-450/4-90-008a
Environmental Protection Planning and Standards May 1990
Agency Research Triangle Park NC 27711
Air
IMPROVE
PROGRESS REPORT
APPENDIX A
-------
EPA-450/4-90-008a
IMPROVE
PROGRESS REPORT
APPENDIX A
By
Marc Pitchford
Environmental Monitoring Systems Laboratory
U. S. Environmental Protection Agency
Las Vegas, NV 93478
And
David Joseph
Air Quality Office
National Park Service
Denver, CO 80228
U S EPTlronosntal Protection Agency
p—iTi 5 I io".ry (£"FL-lo)
Rue,lOn -J I J-j-.J-'-^J
230 S. Dearborn Street. Room 1670
Cliicago, IL 60604
Office Of Air Quality Planning And Standards
Office Of Air And Radiation
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711 '
Mav 1990
-------
This report has been reviewed by the Office Of Air Quality Planning And Standards, U. S. Environmental
Protection Agency, and has been approved for publication. Any mention of trade names or commercial
products is not intended to constitute endorsement or recommendation for use.
EPA-450/4-90-008a
-------
Standard Operating Procedures for
IMPROVE
Particulate Monitoring Network
Analytical Services Division
Crocker Nuclear Laboratory
University of California, Davis
.Davis, CA 95616
July 1989
-------
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
Table of Contents
I. INTRODUCTION 4
A. Role of Procedures in the IMPROVE system 4
B. The IMPROVE Sampler 6
C. Responsibilities for Maintaining the Procedures 7
D. Overview of the Standard Operating Procedure Document 8
II. SAMPLE HANDLING BEFORE SHIPMENT TO THE SITE 10
A. Preparing the Clean Filters 11
1. Module A fine Teflon filters 11
2. Module B fine nylon filters 14
3. Module C fine quartz filters 14
4. Module D PM10 Teflon filters 14
B. Loading the Cassettes 14
C. Verifying the Loading Procedures and Leak-testing the Cassettes . . 18
D. Prepare the Blue Box for Shipment 19
III. SAMPLE CHANGING AT THE SITE 20
A. Preparing for the Weekly Sample Change 21
B. Removing the Exposed Cassettes 21
C. Inserting the Clean Cassettes 22
IV. SAMPLE HANDLING AFTER SHIPMENT FRCM THE SITE
23
A. Receiving the Blue Box 24
B. Reviewing the Field Logsheet 24
C. Unloading and Cleaning the Cassettes 26
1. Module A fine Teflon filters 26
2. Module B fine nylon filters 26
3. Module C fine quartz filters 27
4. Module D PM10 Teflon filters 27
5. IMPROVE Dot Chart 27
D. Checking and Entering the Field Data 28
1. Review the Field Logsheet for Quality Assurance 28
2. Entering the field data into the sample-handling database ... 28
E. Processing and Shipping the Exposed Filters 29
1. Module A fine Teflon filters 29
2. Module B fine nylon filters 30
3. Module C fine quartz filters 30
4. Module D PM10 Teflon filters 31
-------
V. SAMPLE ANALYSIS 32
A. Gravimetric Mass Analysis 32
B. Laser Integrating Plate Analysis (LIPM) 33
C. PIXE/PESA Analysis 36
D. Ion Oircmatograph. Analysis (1C) 37
1. Receipt of the filter 37
2. Filter extraction , 38
3. Ion analysis 38
4. Data transfer to UCD 38
E. Thermal Optical Reflectance Analysis (TOR) ., 39
VI. PROCEDURES FOR DATA PROCESSING 41
A. Introduction: The Equations of Concentration and Uncertainty ... 41
1. Volume 41
2. Gravimetric Mass 41
3. optical Absorption 42
4. PIXE Analysis 42
5. PESA Analysis 42
6. Ion Analysis 42
7. Carbon Analysis 42
B. Entering the Data into the Concentration Database 43
C. Validating the Data 44
D. Preparing Magnetic Tapes and Floppy Disks 44
E. Preparing the Seasonal Summaries 44
VII. PROCEDURES FOR SAMPLER MAINTENANCE 45
A. Evaluating Sampler Modifications 45
B. Calibrating the Flow Audit Device 45
C. Preparation for Annual Site Visit 46
D. Annual Site Visit 46
-------
Standard Operating Procedures for
IMPROVE
Particulate Monitoring Network
July 1989
List of Tables and Figures
Table 1. Filter used in standard IMPROVE sampler 7
Table 2. Filter identification code for Teflon filters 13
Table 3. Sample identification code and color code for all filters .... 15
Figure l. Schematic of IMPROVE Sampler
Figure 2. Flow diagram for the procedures for sample handling before
the shipment of the clean filters to the sampling site.
The starred procedures (*) are done by the carbon contractor . . 10
Figure 3. Flow diagram for sample changing at the site 20
Figure 4. Flow diagram for the procedures for sample handling after
the receipt of the exposed filters from the sampling site ... 23
Figure 5. Schematic of the LIPM system 33
-------
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
I. INTRODUCTION
A. Role of Procedures in the IMPROVE system
The nature and extent of the particulate monitoring program based on the
IMPROVE sampler require carefully designed operating procedures. The
procedures have developed out of our experience in operating monitoring
networks since 1973. Some of the major factors influencing the
procedures are as follows.
1. There are presently 48 sites using the IMPROVE sampler (20 IMPROVE,
19 NFS, 7 NESCAUM, 2 Lake Tahoe), requiring the processing of over
500 filters each week. The large flow of samples requires
sample-handling procedures in the laboratory to be as efficient as
possible. To handle this workload, the procedures in the laboratory
have been separated into well-defined tasks.
2. The system uses multiple filter media, with a given filter following
one of 6 different pathways. It most of these pathways, it is
necessary to identify the clean filter as well as the exposed
filter. This complexity requires a sample-identification system
that is simple and reliable. The procedures must include numerous
cross-checks, but cannot be allowed to become cumbersome. The
procedures must incorporate a good inventory system so that all
filters can be accounted for at any time.
3. Most of the sites are located in pristine regions where the ambient
concentrations of fine particles are extremely low, often 1 to 2
orders of magnitude below urban levels. The low loadings require
procedures that minimize and monitor any contamination of the
samples. For example, every box of clean filters is given a date on
which the filters are to be installed in the sampler, in order to
avoid filters remaining at the sites too long and picking up extra
contamination. In addition, the filters remain in sampling
cassettes from Che time they leave the laboratory until the time
they return.
4. Since the samples are collected twice a week, it is essential that
the downtime of the samplers be kept to a minimum. This requires
procedures that ensure a steady flow of clean filters to the site.
The procedures must also include internal checks of the field
information to identify problems with the sampler and with sample
collection. It is essential that the problems be identified and
rectified rapidly.
5. The particulate data are to be used in various source-receptor
models to determine causes of visibility impairment. The data must
be able to stand legal scrutiny. The procedures must include
sufficient quality assurance checks to provide the most accurate and
dependable results possible.
Page 4
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
6. The changing of filters at the sampling site is handled by persons
who are involved in the program less than an hour a week.
Therefore, the field procedures must be simple and self-evident. By
making the normal routine as simple as possible, the field operators
can concentrate on abnormalities.
Page 5
-------
IMPROVE STANDARD OPERATING PROCEDURES
JULY 1989
B. The IMPROVE Sampler
The IMPROVE sampler consists of four independent filter modules and a
common controller, as shown in Figure 1. Each module has its own inlet,
PM2.5 or PM10 sizing device, flow rate measurement system, flow
controller, and pump. The 4 pumps are housed in a separate unit to
isolate their vibration from the rest of the sampler.
PM10 teflon
sunshleld
wires
and
vacuum hoses
in conduit
Figure 1. Schematic of IMPROVE Sampler
power
Page 6
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
Each module in the standard sampler has two filter cassettes, with the
first cassette collecting samples on Wednesday and the second on
Saturday. Each filter cassette is color-coded, with one set of colors
referring to the module (red, yellow, green, blue) and the second set to
the first or second cassette within the module (black, white). Table 1
lists the major data for each module, including the filter media, the
color codes, and the analytical techniques.
Table 1. Filters used in standard IMPROVE sampler.
module particle filter color . analytical methods
size material code
A PM2.5 Teflon red mass, LIPM, PIXE, PESA
B PM2.5 nylon yellow 1C
C PM2.5 quartz green TOR
D PM10 Teflon blue mass
The filters are transported to and from the site in sealed cassettes
using specially designed shipping containers. The field operator
removes the cassettes with exposed filters and inserts the new cassettes
without directly handling the filters. With this system it is possible
to have multiple filters in a single cassette. The C module has two
quartz filters in tandem, with the second filter to monitor the artifact
caused by adsorption of organic gases; this second filter is not
routinely analyzed. For NFS congressionally mandated sites, the D
module has an impregnated filter following Che Teflon filter in order to
measure sulfur dioxide gas.
C. Responsibilities for Maintaining the Procedures
The overall responsibility for the procedures belongs to the Project
Manager. He is responsible for appropriate coordination between
personnel to maintain the procedures. As a senior scientist, the
Project Manager is responsible for characterizing and calculating the
various artifacts and uncertainties in the collection and analysis.
This includes proposing QA tests using the Davis Field Station.
The Quality Assurance Manager is responsible for the details of the
procedures. He works closely with the Field Support Manager to verify
that the proper procedures are being followed in the. sample-handling
laboratory and in the field. He is responsible for validating the
results of the various UCD analytical systems arid the final data,
including the data from the external contractors.
Page 7
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
The Field Support Manager is responsible for the procedures in the
sample-handling laboratory and for procedures in the field. Included in
the laboratory procedures are the procedures for measuring mass and
absorption. He coordinates with the external contractors to maintain a
proper flow of samples and data. He coordinates with the Field Engineer
to correct sampler problems immediately and oversees the annual
maintenance schedule.
The Field Support Manager is also the sample-handling laboratory
supervisor. Assisting him are three full-time laboratory technicians
and several student technicians.
The maintenance of the samplers is the responsibility of the Field
Engineer and the Field Technician. This responsibility includes the
•annual audit and routine maintenance. These persons are also
responsible for conducting QA tests at the Davis Field Station,
including tests to consider the feasibility of sampler modifications.
The PIXE/PESA Manager is responsible for the procedures to be followed
in the analysis of the samples by PIXE and PESA.
The Support Group is responsible for developing and maintaining the
computer hardware and software used in the procedures. The software
includes several programs for sample handling, programs for the PIXE and
PESA analytical methods, programs for data reduction and validation, and
programs for data presentation and transmission. The Head of the CNL
Computer Support Division is the Computer Support Manager. The other
members of the group are the Quality Assurance Manager and the PIXE/PESA
Manager. All three members of the Computer Support Group have
considerable experience in aerosol analysis. The group coordinates with
three additional staff members who develop and maintain software
relating to operating procedures.
D. Overview of the Standard Operating Procedure Document
The procedures are separated into 6 segments. The first 3 segments
concern sample handling: before shipment to the site, at the site, and
after return from the site. These procedures are designed to maintain a
smooth flow of samples, assure good inventory control, and minimize
contamination. The fourth segment treats the several analytical
techniques used in the program. The purposes of the analytical
procedures are primarily to maintain the highest quality analyses and
secondarily to maintain efficiency. The fifth segment examines the
procedures to produce the concentration database from the analytical
results, to validate the data, and to generate the seasonal data
summaries. This section includes procedures to compare all the data in
order to detect possible errors.The final segment deals with procedures
for routine and unscheduled maintenance of the samplers in the field.
At the beginning of each of these segments there is a flow diagram or a
summary of the procedures of the segment.
Page 8
-------
IMPROVE STANDARD OPERATING PROCEDURES
JULY 1989
Supplemental information is included in appendices to these segments.
Appendix 1 lists the various forms used in the three sample-handling
segments. Appendix 2 describes the IMPROVE sampler and the field
procedures in detail. Appendices 3 to 7 give more, detailed operating
procedures of the various analytical procedures. Appendix 8 provides
additional lists for the annual inspection and maintenance of the
samplers, and forms for flow audits.
Page 9
-------
IMPROVE STANDARD OPERATING PROCEDURES
JULY 1989
II.
SAMPLE HANDLING BEFORE SHIPMENT TO THE SITE
The handling procedures in this section cover all actions from the purchase
of the flltJrS up to the shipment of the clean filters to the sampling site
Figure 2 gives the flow diagram for this portion of the procedures. UCD is
responsible for all actions in this section except for the purchase and
pre-firing of the quartz filters, the responsibility of the carbon
contractor.
A
fine Teflon
B
fine nylon
C
fine quartz
D
total Teflon
purchase
filters
i
'
measure
pre-LIPM
i
F
measure
pre-MASS
i
<
purchase
filters
purchase
filters
1
pre-fire
filters*
i
f
send to
UCD*
,
>
purchase
filters
measure
pre-MASS
i
•
load clean filters into cassettes
attach identification fags to cassettes
leak-test all cassettes
send cassettes with clean filters to site
Figure 2. Flow diagram for the procedures for sample handling before the
shipment of the clean filters to the sampling site. The starred procedures
(*) are done by the carbon contractor.
Page 10
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
A. Preparing the Clean Filters
This section includes the procedures for preparing the four types of
clean filters used at the IMPROVE sites: fine Teflon, fine nylon, fine
quartz, and PM10 Teflon. Because the procedures differ significantly
between the types, each type will be treated individually. This section
includes procedures from two analytical methods: gravimetric mass (MAS)
and the Laser Integrating Plate Method (LIPM). The procedures are
discussed in detail in sections V-A and V-B.
1. Module A fine Teflon filters
a. Purchase Teflon filters (Gelman Teflo) from commercial vendor.
b. Prepare collection masks.
The collection mask fits underneath the fine Teflon filter in
the cassette and reduces the area of collection. The primary
purpose is to improve the sensitivity of the PIXE and PESA
analyses by concentrating the particles. The mask also reduces
the mass artifact by isolating the filter from the 0-ring. The
decrease in the area of collection is limited by filter
clogging; in regions of high concentrations the flow rate can
be reduced below the acceptable range.
The collection masks are prepared from the inert paper spacers
that Nuclepore Corporation provides between Nuclepore 47 mm 8 urn
polycarbonate filters. The spacers have a very light coating of
Apiezon-L grease. Many years of experience have shown that this
paper will not transfer mass to the filters. The spacers are
retained in their original factory containers when 47 mm
Nuclepore filters are processed for other studies. These
containers are labelled and sealed until required as stock for
preparing the collection masks.
The masks are prepared using specially machined double-action
cutter punches. The cutter punches simultaneously cut the 25 mm
outer diameter and the desired collection diameter. Both punch
and cutter are concentric, centering the collection area in the
25 mm mask. It is imperative that the cutter punch be handled
gently and the heat-treated and ground edges be protected. The
edges are very sharp and can cause injury. If the cutter punch
is dropped or not correctly used, damage to the cutter punch
will result and it will have to be remachined.
The procedures for preparing the collection masks are as
follows.
i. Obtain a sealed and marked container of 100 Nuclepore
spacers from the laboratory supply.
Page 11
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
ii. Obtain the correct punch for the desired collection area,
generally 2.2 cm .
iii. Obtain a thick pad of clean, pure Teflon material from the
laboratory supply retained for this purpose. This pad,
approximately 1/2-inch thick, provides a cutting surface
that will not dull or change the geometry of the cutter
punch faces.
iv. Place the Teflon material on the base of the mandrel press.
Open the mandrel press enough to accommodate the cutter
punch assembly when resting on the Teflon. Initially, place
no more than 5 spacer papers on the Teflon pad and center
them under the cutter punch. Gently, but firmly, apply
pressure to the assemby with the press arm until a distinct
cutting sound is heard and felt. Open the press and observe
that 25 mm masks with the desired hole have been produced.
Remove the completed masks and put them in a clean
container. Discard the excess spacer material.
v. Verify that the mask geometry is correct. Using a vernier
inside/outside caliper from the laboratory supervisor,
measure the inside diameter of the fabricated masks. Make
several measurements and determine the mean. For a 2.2
cm^collection area, the diameter should be 16.73 + 0.01 mm.
(The collection area of the actual particles on the filters
has been found to equal this measured area.)
vi. With practice and experience, up to 10 masks can be produced
at a time. If incomplete cutting occurs, reduce the number
of pieces of stock in the press back to 5.
vii. When all 100 spacers in the original supply have been
converted to masks, seal the masks in their container. Wrap
the cutter punch with tape and return it co che protective
foam box. Return the Teflon pad to its storage.
c. Process a group of 50 filters through the LIPM station.
At this station the filters are given filter identification
numbers and the precollection LIPM values (pre-LIPM) are
measured. The filter identification number is used for the
Teflon filters (modules A and D) to associate the precollection
values of LIPM and mass with a given filter. The filter
identification codes are given in Table 2. The filter
identification system is used in addition to the sample
identification system that specifies the sampling site and the
date of collection. The correspondence between the two
numbering systems is made after the exposed filter returns from
the site.
Page 12
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
Table 2. Filter identification code for Teflon filters,
module collection area code
fine Teflon A 1.1 cm;? TFnnnn.M2
fine Teflon A 2.2 cm2, TFnnnn.Ml
fine Teflon A 3.8 cm2 TFnnnn.MO
PM10 Teflon D 3.8 cm2 TFnnnn.U
i. Prepare 50 Filter Identification Tags (red Avery dot labels)
with the next sequential filter identification number.
(Check last assigned number on the A-Preweight Logsheet and
increment by one.) Obtain a petri box of 5o petri dishes and
place the Filter Identification Tags on the petri dishes.
Record the numbers on the A-Preweight Logsheet.
ii. Insert clean Teflon filters from the supply into the 50
Quickie Mounts of the LIPM system.
iii. Measure the 50 pre-LIPM values following the procedures
described in section V-B. (The pre-LIPM value is the
intensity of light transmitted through the clean filter.)
Record each pre-LIPM value alongside the filter
identification number on the A-Preweight Logsheet.
iv. Place each filter in its proper petri dish. Place the petri
dish back in the petri box and label the box.
d. Measure the tare masses of 50 filters (pre-MASS).
The filters are weighed in batches of 50 filters. Within this
batch the work is separated into groups of approximately six
filters. Between each group of 6 filters, the calibration of
the balance is checked. If necessary, the balance is
recalibrated.
Remove 6 petri dishes from the appropriate petri box and measure
the pre-MASS values of the filters following the procedures
described in section V-A. Record the pre-MASS value of each
filter alongside the proper filter identification number on the
A-Prew'eight Logsheet. Return each filter to its petri dish
after weighing and return the petri dish to the petri box.
&. For each of the 50 entries on the A-Preweight Logsheet, enter
the filter identification number, the pre-MASS value, and the
pre-LIPM value in the sample-handling database.
Page 13
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
2. Module B fine nylon filters
Purchase nylon filters (Gelman Nylasorb) from commercial vendor. No
further processing is necessary.
3. Module C fine quartz filters
a. (carbon contractor) Purchase Pallflex 2500QAT-UP quartz filters.
b. (carbon contractor) Pre-fire the quartz filters at 900° for at
least 4 hours, following operating procedures developed by the
contractor.
c. (carbon contractor) Ship the carbon filters to UCD. These are
identified by lot number.
d. (UCD) Receive the prefired quartz filters and store in the
freezer until required for the loading sequence.
4. Module D PM10 Teflon filters
a. Purchase Teflon filters (Gelman Teflo) from commercial vendor.
b. Prepare 50 Filter Identification Tags (blue Avery dot labels)
with the next sequential filter identification number. (Check
last assigned number on D-Preweight Logsheet and increment by
one.) Get a petri box of 50 clean petri dishes and place the
Filter Identification, Tags on the petri dishes. Record the
numbers on the D-Preweight Logsheet.
c. Measure the 50 pre-MASS values following the procedures
described in section V-A. (The pre-MASS value is the tare mass
of the filter.) Record each pre-MASS value alongside the filter
identification number on the D-Preweight Logsheet. Place each
filter in its proper petri dish after the weighing and return
the petri dish to the petri box. Label the petri box when
finished.
d. For each of the 50 filters on the D-Preweight Logsheet, enter
the filter identification number and the pre-MASS value into the
sample-handling database.
B. Loading the Cassettes
The procedures in this section are triggered by having a Blue Box with
empty cassettes. The Blue Box is a shipping container that was
specifically designed to transport the weekly supply of filter cassettes
between UCD and the sampling sites. The blue case is molded ABS
material with metal reinforcement. The cassettes are vibrationally and
thermally insulated by a soft polyurethane foam liner. The case is
Page 14
-------
IMPROVE STANDARD OPERATING PROCEDURES
JULY 1989
water-tight and held closed with cam locks. Molded on the front of the
case is the message "KEEP OUT OF DIRECT SUNLIGHT." On the front there is
a flush-mounted mailer with a metal frame and an indented section to
display the site-date tag. The mailer has a reversible prepaid mailing
label. The UCD address is on one side of the mailer, while the site
address is on the other. Each Blue Box is assigned to a given site.
For each site there are 5 to 7 Blue Boxes in the system.
1. Prepare the Field Logsheet with the site and date sequence.
a. The Blue Box will have the site and date for the desired sample
change listed on the outside label. Verify that this date
agrees with those of the next tags on the sheet of preprinted
sample identification numbers for this site. Update the IMPROVE
Dot Chart to indicate that sample preparation is in progress.
b. Record the appropriate sample identification numbers
corresponding to each sample on the Field Logsheet. Table 3
gives the code for the sample identification numbers and the
associated color codes for all of the modules. (The color code
is used to mark the cassettes.)
Table 3. Sample identification code and color code for all filters.
module filter color code sample identification code
A
A
B
B
C
D
D
1 red/black
2 red/white
1 yellow/black
2 yellow/white
1 green/black
2 green/white
1 blue/black
2 blue/white
site-date-Al
site-date--A2
site-date-Bl
site-date-82
site-date-GlP
site-date-CIS
site-date-C2P
site-date-C2S
site-date-Dl
site-date-D2
(primary)
(secondary)
(primary)
(secondary)
The site code is the standard four-letter NFS code. (e.g. ROMO)
The date is that for the Tuesday of the sample change, (e.g. 1/7/89)
2. Check the cleanliness of the module A cassettes and load them with
clean Teflon filters.
a. Remove the protective red cap from the Al cassette (red/black)
and verify its cleanliness.
Page 15
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
b. Remove the cassette top and the lock ring. Check that the top
has a single flat gasket. Recheck the condition and the
cleanliness of the collection mask, the drain disk, and the
ethylene propylene (EP) 0-ring. Replace the drain disk, if it
has discarded during cleaning or if it is dirty. Replace the EP
0-ring, if it was removed during cleaning or if it deformed. Be
careful to return the filter grate with the finer grids oriented
up and the large bar grids down.. If necessary, reclean the
filter grate and 0-ring by gently brushing with the brush
reserved for nonquartz cassettes.
c. Verify that the correct collection mask is installed; if the
mask is dirty or worn, replace it with a new one.
d. Remove the next petri dish from the fine Teflon petri box.
Transfer the filter identification tag from the petri dish and
the sample identification tag from the pre-printed sheet to the
top cap, locating them between the exposed ridges. Center and
mount the clean drain disk, the collection mask, and the Teflon
filter on the grate. The rough and extended edges of the filter
should mate with the mask. This orientation will present an
open, smooth, and flat surface. Gently place the lock ring on
the cassette, taking care to keep the filter centered and not
damage the filter support ring. Place the cap with a single
flat gasket over the lock ring and tighten it. Keep all
elements centered and make it finger tight. Reinstall the
protective red cap designated "A" over the cassette.
e. Record the filter identification number on the Field Logsheet.
f. Put the cassette in the Blue Box.
g. Repeat procedure for the A2 casette (red/white).
3. Check the cleanliness of the 47 mm module B cassettes and load them
with clean nylon filters.
a. Check the cleanliness of the Bl cassette (yellow/black)
following the procedures for module A.
b. Transfer the Sample Identification Tag from the pre-printed
sheet to the cassette. There is no filter identification number
for the nylon filter.
c. Obtain a clean nylon filter from the supply and load it into the
cassette directly on to the diffusion and support grid.
d. Put the cassette in the Blue Box,
e. Repeat for the second nylon filter (B2, yellow/white).
Page 16
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
4. Check the cleanliness of the module C cassettes and load them with
clean quartz filters.
The module C cassettes contain two sections, each with a quartz
filter. The first or primary filter collects all the particles and
adsorbs organic gases, while the secondary filter only adsorbs
gases.
a. Disassemble each section of the Cl cassette (green/black) and
check the cleanliness. Assure that two flat silicone rubber
gaskets are used in the cassette cap of the first section. Two
gaskets are needed to provide spacing so that the hold down
retainer of the cyclone manifold will fit. If there is still
any glass filter debris, remove it with the brush reserved for
quartz cassettes. Replace the drain disk if it was removed
during cleaning or if it is dirty. Replace the 0-ring with a
Viton 0-ring, if it was removed during cleaning or if it is
deformed.
b. Load a clean quartz filter from the supply into each section of
the cassette.
c. Record the lot number for each filter on the Field Logsheet.
d. Transfer the Sample Identification Tags from the pre-printed
sheet to each section of the cassette. The primary filter has
the suffix "P", while the secondary filter has the suffix "S."
e. Put the cassette in the Blue Box.
f. Repeat for the second quartz cassette (C2, green/white).
Check the cleanliness of the module D cassettes and load them with
clean Teflon filters.
a. Check the cleanliness of the Dl cassette (blue/black) following
the procedures for module A.
b. Insert a clean Teflon filter from the module D container box.
Do not use a collection mask with this filter. Mount the filter
with the ribbed side down and the smooth side up. ("Up" means
facing the incoming air.)
c. Attach the Filter Identification Tag and the Sample
Identification Tag to the outside of the cassette.
d. Record the filter identification number on the Field Logsheet.
e. Put the cassette in the Blue Box.
Page 17
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
f. Repeat for the module D2 cassette (blue/white).
6. If a dynamic field blank (DFB) cassette is in the DFB queue, include
it in the current Blue Box. (The DFB is identified by having a
yellow cap on the hose end.) Do not include more than one DFB per
Blue Box. The type of module D cassette must match the normal type
for this site, with respect to the impregnated filter. Repeat the
procedure for a cassette of the given module. Add the suffix DFB to
the sample-identification number. Include the standard letter
describing how to install the DFB.
7. This completes the loading of the cassettes. Transfer the Blue Box
to the leak-test area. Leave the Blue Box open.
C. Verifying the Loading Procedures and Leak-testing the Cassettes.
At this point, all the cassettes have been loaded with clean filters,
have their identification tags, and are in an open Blue Box, along with
the accompanying Field Logsheet.
1. Verify that all cassettes are present and have been loaded with
clean filters.
2. Verify that the filter- and sample-identification numbers and che
quartz-lot numbers correspond to those recorded on the Field
Logsheet.
3. Record these numbers on the Mailer Record Logcard.
4. Mount a cassette on the leak-test element designated for the
specific module, following the standard color code. Hand tighten
the nut. Remove the protective red cap and place che appropriate
leak-cap gently over the open cassette face. There are two
leak-caps, one for 25 nun and one for 47 mm. Open the vacuum valve
and observe the flowmeter. If the flow rate is less than 1 L/min,
the system is considered to be airtight. If the flow rate exceeds
this value, find and remove the leaks (recenter components, replace
0-rings, etc.).
Special care is required for the module C quartz cassettes. Because
of the fragility of the filters, it is not possible to get as good a
seal; the criterion for this module is 3.5 L/min. Begin by
tightening the secondary filter, and then tighten the primary
filter. Be careful to tighten each filter segment only to a "firm"
hand tightness. An absolute leak seal cannot be guaranteed by
further tightening. Excessive tightening will cause the lock ring
to sever the edges from the filter. This will cause the cassette to
leak significantly, and the flow meter will rise sharply. The
filter will appear to warp. Replace all filters which have been
overtightened.
Page 18
-------
JUL.Y 1989
5. Replace the protective red cap designated for the module.
6. Repeat the leak- test for the remaining cassettes in the Blue Box.
7. Record the current date on the Mailer Record Logcard in the column
marked "date shipped. "
8. File the Mailer Record Logcard in the Mailer File chronologically by
site. Verify that the this card follows the previous card by 1
week.
D. Prepare the Blue- Box for Shipment.
1. Make certain the reusable, reversible mailer label on the Blue Box
displays the site destination.
2. Update the IMPROVE Dot Chart by filling in the site/week box with
yellow ink.
3. Place the Blue Box in the mailing tub for dispatch.
Page 19
-------
IMPROVE STANDARD OPERATING PROCEDURES
JULY 1989
III. SAMPLE CHANGING AT THE SITE
Sample changing is performed every Tuesday by the field operator. Each site
has a manual (Appendix 2) with the procedures and a description of the
program and the sampler. The procedures are also written on the inside of
the doors for the controller, module A, and module D. Figure 3 gives the
flow diagram for sample changing.
receive container of clean filters from UCD
record date and temperatures
start 30-minute timer
measure and record flow rates
record elapsed time
remove cassettes with exposed filters
JL
insert cassettes with clean filters
measure and record flow rates
send container of exposed filters to UCD
Figure 3. Flow diagram for sample changing at the site.
Page 20
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
A. Preparing for the Weekly Sample Change
1. When the Blue Box is received at the site, record on the Receipt
Log, the date received and the date written on the label of the Blue
Box. The Blue Box should be received 1 to 2 weeks before the date
for sample-changing.
*
2. Keep the Blue Box in a clean, dry, and cool location.
3. For the Tuesday sample change, take two Blue Boxes to the site: the
empty Blue Box with last Tuesday's date (Old Blue Box) and the full
Blue Box with the current Tuesday's date (New Blue Box).
B. Removing the Exposed Cassettes
1. Remove the Field Logsheet from the Old Blue Box. The front side
will have been filled in last Tuesday. On the reverse side, record
the present date and time, your initials, and the current, the
minimum, and the maximum temperatures.
2. Open the control module and verify the date and time of the
controller clock. Turn the bypass timer to 30 minutes. This will
start all four pumps.
3. For each module (A, B, C, D) in turn, do the following:
a. Record the small gauge before pressing either filter switch.
This should be between 16 and 25 "Hg, depending on the
elevation. If it drops below this, the problem might be
internal pump failure or leaks in the hose. Call UCD.
b. Press the filter 1 toggle switch and record the small gauge and
the magnehelic gauge. Repeat for filter 2. The readings should
be near uhe green line marked on the gauges. See manual or
instructions on door for troubleshooting help.
c. Record both elapsed times in hundredths of hours and zero both
timers. Times of 24.00 hours are expected.
d. Remove each cassette from the holder by first unscrewing the
hose from the solenoid. Then unscrew the knurled knob holding
the cassettes enough to lift the support bracket and remove the
cassette. Do not remove the knob completely. Get the
protective red cap marked for this module from the old box. Put
the cap on the cassette and place the cassette in the Old Blue
Box. When finished with module D, place the completed Field
Logsheet in the Old Blue Box.
Page 21
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
C. Inserting the Clean Cassettes
1. Remove the Field Logsheet from the New Blue Box. Check the date on
the box and on the Field Logsheet; if this is not the present date
for the week, make a note on the Field Logsheet. Fill in the
current date and time and your initials.
2. Insert the cassettes for each module in turn, selecting the two
cassettes for each module:' red for A, yellow for B, green for C,
and blue for D. Remove the red cap from the black (filter 1)
cassette and put the cap in the New Blue Box. Insert the cassette
into the left side of the manifold (under the clamp) and attach the
hose onto the left (filter 1, black) solenoid. Hand tighten the
nuts on the cassette. Do not use a wrench or pliers to tighten
them. Make certain both cassettes are firmly mounted on the
manifold and tighten the knurled knob.
3. Record the gauges for each module in turn.
a. Check that the elapsed timers were reset.
b. Record the small gauge before pressing the toggle switches.
c. Press the filter 1 toggle switch and record both gauges. If
either gauge is not near the green line, check the
troubleshooting guide inside the module A or D door.
d. Press the filter 2 toggle switch and record both gauges.
4. Store the Field Logsheet in the New Blue Box.
5. Reverse the mailing label on the Old Blue Box with the exposed
filters, so that Che label shows the UCD address. Do not add
postage; the shipment is prepaid, First Class. Send the Blue Box
to UCD.
6. If a dynamic field blank is sent with the Blue Box, it is
accompanied by a form letter describing the special steps required.
The dynamic field blank is needed to monitor any contamination in
the system. It follows the same path as an actual sample, except
that no air is drawn through. The cassette with the dynamic field
blank has a yellow cap on the end of the hose and has the suffix DFB
on the end of its identification number (e.g. GUMO 09/17/89
AlDFB.) Do not remove the yellow cap. The hose will be tagged with
the color appropriate for that module. The blank is inserted on the
maniford in place of one of the two manifold caps, so that there are
three cassettes on the manifold. When the cassettes are removed on
the following Tuesday, replace the cap on the cyclone manifold.
Page 22
-------
IMPROVE STANDARD OPERATING PROCEDURES
JULY 1989
IV. SAMPLE HANDLING AFTER SHIPMENT FROM THE SITE
The handling procedures in this section cover all actions from the receipt
of the Blue Box with exposed filters up to the transfer of the filters to
the external contractors or to the queue for cyclotron analysis. Figure 4
gives the flow diagram for this portion of the procedures.
receive cassettes with exposed fillers from site
review field log slice!
discard Invalid samples
enter info database
transfer filters and Identification lags from cassettes to petrl dlslies
A
fine Teflon
send to
Ion contractor
send to
carbon contractor
n
fine nylon
C
fine quartz
I)
total Teflon
Figure 4. Flow diagram for the procedures for sample handling after the
receipt of the exposed filters from the sampling site.
Page 23
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
A. Receiving the Blue Box
1. Immediately upon receipt of the Blue Box of exposed cassettes from
the site, update the IMPROVE Dot Chart by placing a blue dot in the
site/week box, on top of the yellow coloring.
2. Clean the outside of the Blue Box.
B. Reviewing the Field Logsheet
The Field Logsheet is reviewed by the person who removes the exposed
filters. The review is conducted prior to the removal. The Field
Logsheets will be reviewed again by the laboratory supervisor prior to
their entry into the sample-handling database.
If correction or explanation is required, draw a single line through the
existing information using a fine red pen, and enter the correct
information. Comments should also be made using the fine red pen.
1. Verify that the actual date of the sample change corresponds to the
sample date of the sample-identification numbers. (Actual dates
that are 1 or 2 days before the sample-identification date are
acceptable.)
If the dates do not agree, and the samples were not collected over
multiple periods, replace each Sample Identification Tag with one
indicating the actual Tuesday sampling date. The discrepancy in
dates could occur if the field operator used the incorrect Blue Box.
2. Check for omissions. If possible estimate the correct reading from
other readings on the Field Logsheet, or from readings on the
previous Field Logsheet. If the solution is not evident, check with
the laboratory supervisor and call the field operator. Clf the
correct data cannot be determined the sample will have to be
invalidated.)
3. Review the comments by the field operator and take action
accordingly. If an equipment or hardware malfunction is reported,
take action to correct the problem.
4. Verify that all closed-solenoid gauge readings (system vacuum
without sampling air flow) are greater than 16 inches mercury.
5. Verify that magnehelic values range from 0.2.0 to 0.60.
6. Verify that filter pressure drops measured on the small gauge are in
the following approximate ranges.
A < 1.5 inches mercury
B < 2.0
C < 5.0 " "
D < 1.5
Page 24
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
Verify that the sampling durations are approximately 24.00 hours.
For sites running with small generators known to have variable
frequency (CANY, ARCH, and ISRO), if the times range from from 23.6
to 24.4 hours, correct the times to 24.00 hours. (The controllers
use quartz clocks and are not affected by the line frequency.) This
variation reflects a frequency variation of ±1 hertz.
If the samples are not valid they are generally removed from the
normal system at this point and put in the Unusable Archive. The
action must be recorded on both the Unusable Archive Inventory and
the Field Logsheet. Add the suffix, listed below, appropriate to
the action to the sample-identification number on both the Sample
Identification Tag and the Field Logsheet (e.g. ROMO 3/15/88A1XX).
Include a short explanation on the Field Logsheet of why the sample
is invalid, such as "Cassette loose and moved from sample port" or
"Tubing attaching support broken." Transfer the filter to a petri
dish and put the petri dish in the Unusable Archive. If there are
questions concerning validity, check with the laboratory supervisor.
If necessary, call the field operator. The three acceptable
suffixes and the accompanying reasons for invalidating a sample and
archiving it in the Unusable Archive are:
a. DB: double sample. The samples were exposed for longer than 24
hours, that is, over more than one period.
b. PF: pump failure. The vacuum gauge readings with the solenoids
closed are less than 15 inches of Hg.
c. XX: invalid for other reasons. The sample ran for less than 12
hours, the filter has a hole, the filter has obvious problems
(not centered properly, grill upside down), the filter was
dropped or contaminated by water, the cassette is broken.
9. In two cases, when the filter clogs or when the filter is not
installed in the sampler, the invalid samples are given a suffix,
but are not removed from the normal system at this point. The
samples will be analyzed but the data will be discarded during che
data validation procedures. Add the suffix appropriate to the
action to the sample-identification number on both the Sample
Identification Tag and the Field Logsheet. If there are questions
concerning validity, check with the laboratory supervisor. If
necessary, call the field operator. The suffixes and the
accompanying reasons for including a sample in this category are:
a. CG: clogged filter. The final magnehelic reading is less than
1/2 of the initial reading, resulting in unreliable flow rate
measurements. (This sample is kept to document the factors
involved in clogging.)
Page 25
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
b. TB: transport blank. The filter was returned without having
been installed in the sampler. (The transport blank is used to
determine artifacts from all causes except adsorption of gases.)
10. Enter the date and your initials as reviewer at the end of the Field
Logsheet.
C. Unloading and Cleaning the Cassettes
1. Module A fine Teflon filters
a. Remove the protective red cap from the Al cassette (red/black)
and verify its cleanliness. If required, wipe the inside of the
cap with a Kimwipe dampened with ethanol.
b. Remove the cassette top and the lock ring. Transfer the fine
Teflon filter to a petri dish and transfer the accompanying
Filter Identification Tag and Sample Identification Tag from the
cassette to the petri dish.
c. Place the petri dish in the fine Teflon storage tube.
d. If an incorrect collection mask was used, write the actual mask
used (e.g. "1.1 cm mask") on the Field Logsheet with the red
pen. Change the filter-identification number on the Field Log
Sheet and on the Filter Identification Tag to the correct mask
code.
e. Check the condition and the cleanliness of the collection mask,
the drain disk, and the EP 0-ring. Discard the drain disk if
dirty and the 0-ring if deformed. Clean the filter grate and
0-ring by gently brushing with the camel-hair brush reserved for
nonquartz cassettes. Discard the collection mask if the visible
aerosol is not uniformly circular.
f. Reassemble the cassette.
2. Module B fine nylon filters
a. Transfer each nylon filter to a petri dish and transfer the
accompanying Sample Identification Tag from the cassette to the
petri dish. There will be no Filter Identification Tag.
b. Transfer the petri dishes to the nylon petri box.
c. Record the sample-identification number on the RTI-Ion
Contractor Inventory.
Page 26
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
d. Clean the cassettes following the procedures for module A.
3. Module C fine quartz filters
The module C cassettes contain two sections, each with a quartz
filter. The first or primary filter collects all the particles and
adsorbs organic gases, while the secondary filter only adsorbs
gases.
a. Transfer each fine quartz filter to a petri dish and transfer
the accompanying Sample Identification Tags from the cassette to
the petri dish. Keep the primary and secondary filters
separate.
b. Place the petri dishes in the quartz petri box. When the box is
full, return the petri box to the freezer.
c. Record the sample-identification number on the DRI-Carbon
Contractor Inventory.
d. If two filters were inadvertently used for either the primary or
the secondary filter, process them as a single filter, and write
a comment on the Field Logsheet. (This occasionally occurs
because the quartz filters tend to stick together.)
e. Make certain that two flat silicone rubber gaskets are used in
the cassette cap of the first section. Disassemble each section
of the cassette and clean. Gently brush away glass filter
debris with the brush reserved for quartz cassettes. Make sure
no debris is retained under the lip of the anti twist lock ring.
Brush the debris into a waste receptacle. Be careful not to
raise or breathe any filter debris dust during this operation.
The quartz debris is highly electrostatic and can contaminate
other samples. Discard Che drain disk if dirty and the 0-ring
if deformed.
4. Module D PM10 Teflon filters
a. Transfer each PM10 Teflon filter to a petri dish and attach the
accompanying Filter Identification Tag and Sample Identification
Tag to the petri dish.
b. Place the petri dishes in the PM10 Teflon storage tube.
c. Clean the cassettes following the procedures for module A.
5. Find the date for the next unprocessed week for this site on the
IMPROVE Dot Chart. Make a new tag for this site-date and place the
tag in its place on the Blue Box. Transfer the Blue Box to the
queue for clean filter processing.
Page 27
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
D. Checking and Entering the Field Data
This step consists in reviewing the Field Logsheet a second time and
entering the data on the Field Logsheet into the sample-handling
database. It can be performed at any time after the Field Logsheets are
initially reviewed and the filters are removed from the cassettes. The
second review is conducted by the laboratory supervisor or his
representative prior to the entry of the field data into the
sample-handling database.
1. Review the Field Logsheet for Quality Assurance
a. Verify that there is an entry for every space on the sheet.
b. Verify that the correct identification codes were made for all
samples. Verify that the dates are all correct or have been
corrected on the Sample Identification Tags and the Field
Logsheet.
c. If the data indicate any mechanical problems, verify that all
necessary corrective action has been initiated. If necessary,
call the field operator.
d. Check for consistency between the module B pressure drop and the
magnehelic readings. Low magnehelic readings may indicate that
the denuder has caused condensation to accumulate in the high
pressure side of the magnehelic. Call the field operator and
have .him remove the hose, drain the water, and re-attach it.
This operation requires less than a minute.
e. Verify that the temperatures have been reported in Celsius; if
not, convert them.
f. Check for discrepancies in the elapsed time. If one of the
elapsed times is less than the others for trie same sampling
period, find the reason for the difference. Either correct the
value or discard the sample.
g. Review the comments of the field operator.
h. When you are satisfied that the Field Logsheet is complete and
correct, place your initials as reviewer above the date changed
on the front side of the Field Logsheet.
2. Entering the field data into the sample-handling database
The entry is performed through the computer program DOLOGS. The
steps of this program are as'follows.
Page 28
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
a. To the question of which record is to be entered, enter the. site
code (4 characters) plus the subcode "1", and enter the date
encoded in the sample-identification number (e.g. "GLAC
05/02/89"). It is essential that this entry be correct.
b. The program will now call up each entry individually, following
the layout of the Field Logsheet.
c. If a character is typed incorrectly, use the backspace to delete
the character, and then type the correct character,
d. If incorrect data has already been entered, use the displayed
menu to modify the data.
e. When the last entry has been completed, the program will
automatically return to the site-date question for a new Field
Logsheet.
f. When the data for a Field Logsheet have been completely entered
into the data base, use a date stamp and a red inkpad to stamp
the current input date in the upper left corner of the reverse
(final) side of the Field Logsheet.
g. Place the date-stamped Field Logsheet in the file folder for
that site and season maintained in the IMPROVE file cabinet.
The Field Logsheets should be arranged in sequential order.
E. Processing and Shipping the Exposed Filters
At this point, the exposed filters are in petri dishes with their
identification tag(s), and the dishes are in the appropriate storage
container. Sach filter type now follows a separate pathway.
1. Module A fine Teflon filters
a. Determine the post-MASS
In this segment, the post-MASS is determined and the filter
transferred to a slide mount.
i. Get a petri dish from the fine Teflon storage tube.
Determine the post-MASS following the procedures of section
V-A.
ii. Record the post-MASS on the Mounted Samples Logsheet.
iii. Load the filter into a slide mount.
Mark the white side of the slide mount with a blue indelible
wide-tip felt pen in four places. Write the
sample-identification number on the left, write the
Page 29
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
filter-identification number on the top, and write the site
code on the right. Read the position number corresponding
to the sample date on the position chart on the wall and
mark this number on the bottom edge of the mount. Place
this bottom edge into the temporary slide tray first; this
number will be hidden by the tray structure.
iv. When the temporary slide tray is full, transfer the tray to
the LIPM analysis queue.
b. Measure the post-LIPM value for each sample in the temporary
tray, following the procedures described in section V-B. Record
the post-LIPM value for each slide on the Mounted Samples
Logsheet.
c. Enter the sample-identification number, the
filter-identification number, the post-MASS, and the post-LIPM
into the sample-handling database, for each of the 50 lines on
the Mounted Samples Logsheet.
This action connects the filter-identification number and the
sample-identification number, permitting the computer to get the
proper pre-MASS and pre-LIPM values. At this point the computer
will determine the net mass and optical absorption of the
sample, check that the values are in the acceptable range, and
print a message to the technician. If the values are not
acceptable, the technician will check the entries.
d. Transfer the slides from the temporary tray into permanent
site-specific trays. (Each site-specific tray will hold the
samples for one season. The position in the tray corresponds to
the sampling period within the season.) The number at the bottom
of the slide mount will give the position in the tray. Place
the tray in the queue for PIXE analysis.
2. Module B fine nylon filters
When four petri boxes (200 nylon filters) have accumulated, ship the
boxes to the ion contractor via First Class Mail, along with the
Nylon Filter Inventory.
3. Module C fine quartz filters
Retain the petri boxes of exposed quartz filters in the freezer
until the large shipping container of 10 boxes (500 filters) is
filled. Add blue ice to the shipping container and ship to the
carbon contractor by Federal Express or UPS Second Day so that the
filters will remain cold during transport. Include the Quartz
Filter Inventory.
Page 30
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
4. Module D PM10 Teflon filters
a. Get a petri dish from the PM10 storage tube. Determine the
post-MASS following the procedures of section V-A. Record the
post-MASS on the PM10 Archive Logsheet. Enter the
sample-identification number, the filter-identification number,
and the post-MASS into the sample-handling database.
b. Return the filter to the petri dish. Place the petri dish in
the next position in the 50-position petri box for permanent
archiving. Record the sample and filter-identification numbers
on the PM10 Archive Logsheet. When the Archive petri box is
full, label the box and archive it.
Page 31
-------
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
V. SAMPLE ANALYSIS
A. Gravimetric Mass Analysis
For sample-handling procedures before and after the gravimetric mass
analysis, see sections I-A-1, I-A-4, III-E-1, and III-E-4. For the
startup precedures every morning and afternoon, see Appendix 3 for
Cleaning and Calibration of the Electrobalance, and IMPROVE Gravimetric
Controls.
Never turn the electrobalance off.
1. Remove the filter from the petri dish by gently slipping the flat
forcep under the outer polyolefin support ring. IMPORTANT: The
forceps should contact only the outer support ring of the filter.
Do not touch the deposit area. Place the filter momentarily on the
antistatic strip, with the aerosol side up. Discard the petri dish
if this is an A filter. Repeat for approximately 5 more filters.
2. Enter the appropriate identification numbers on the appropriate
logsheet:
module A precollection A-Preweight Logsheet
module D precollection D-Preweight Logsheet
module A postcollection .... Mounted Samples Logsheet
module D postcollection .... PM10 Archive Logsheet
3. Remove the filter from the antistatic strip and place it aerosol
side up in the center of the weighing pan but offset toward the
right side of the pan by approximately 2 mm. The offset is required
so that the filter can be removed from the pan after weighing
without putting stress on the balance. If the filter is
inadvertantly centered on the pan, use a second pair of forceps to
nudge the filter very gently to a position where the flat forceps
can again grasp the filter support ring only.
4. Close the glass door and allow the electrobalance to stabilize.
This will require approximately 30-45 seconds. Note that the
digital weight readout will slowly decrease until stabilized.
5. When the balance is stabilized, record the mass value on the
logsheet.
6. Remove Che filter by opening the glass door and gently grasp the
filter support ring with forceps. IMPORTANT: Be very careful not
to grasp the edge of the weighing pan while removing the filter;
pulling the balance pan could damage the balance. Slowly remove the
filter from the balance cavity and place it either in the same petri
dish (D filter) or on a prepared PIXE slide mount with the aerosol
side up (A filter).
Page 32
-------
IMPROVE STANDARD OPERATING PROCEDURES
JULY 1989
7. After every 6 measurements, allow the balance to stabilize without a
filter to observe the "zero." If this value exceeds ±2 micrograms,
recalibrate the balance following the procedures of Appendix 3.
8. Whenever the balance is not in use, close the balance glass window.
B. Laser Integrating Plate Analysis (LIPM)
The LIPM system is used to measure the optical absorption of the
particles on the fine Teflon filters. The absorption by the particles
on the filter is smaller than the absorption by particles in the
atmosphere because of the layering of particles on the filter. A
correction to the measured value, based on the areal density of
particles on the filter, is made at the time of data processing. A
schematic of the system is given in Figure 5. Light of 633 nm
wavelength from a He(Ne) laser is diffused and collimated to provide a
uniform beam of light of approximately 0.7 car at the sample. The light
transmitted through the sample is collected with an Oriel 7022
photodiode detection system. The decrease in light intensity is
produced by both absorption and large-angle scattering. (Light
undergoing small angle scattering will be collected by the
detector.) The blank Teflon filter does not absorb light, but it does
scatter light; therefore, it is necessary to measure the transmission
of the blank filter. For the particles on the filter, the absorption is
the primary cause of decrease in light intensity, with only a small
amount of scattering.
LIGHT SHROUD
-FILTER „ OPAL
GLASS
MASK
(VARIABLE)
RADIOMETER
Figure 5. Schematic of the LIPM system.
Page 33
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
The UCD LIPM system is designed Co handle samples in 18x24 mm slide
mounts arranged in linear trays. Prior to the analysis of the
pre-collection clean filters it is necessary to put the filters in slide
mounts; specially modified slide mounts (Quickie Mounts) have been
machined to simplify the process. These Quickie Mounts keep the filters
centered but permit the filters to be inserted and removed easily. The
exposed filters are arranged in permanent slide mounts and linear slide
trays prior to the post-LIPM analysis.
The system is calibrated at the beginning of every analytical session,
following the procedures of Appendix 4. The system is adjusted to give
a reading of 0.000 when then Beam Attenuation lever is closed and 0.750
when no sample is present. A set of standard samples are analyzed and
the measurements are compared with the standard values. If the
measurements differ by more than a preset value, approximately equal to
the uncertainty of the measurement, the calibration is repeated. If
differences still exist, the system is examined for problems.
The steps for the pre-LIPM analysis is as follows. See section I-A-1
for procedures before and after the pre-LIPM measurement.
1. Clean the flat metal forceps with ethanol and transfer 50 Teflon
filters into Quickie Mounts. (The "Quickie Mount" is a modified
slide mount used to hold the filters during the precollection LIPM
analysis.) Note that each filter support ring has two distinctive
sides, the smooth side and the lipped side.
Place the filter in the Quickie Mount with the smooth side facing
the black side of the mount. This orientation is followed in other
Air Quality Group protocols and assures that smooth side is always
the aerosol side, and that the smooth side always faces LIPM and
PIXE detectors. This standardized procedure also assures that when
aerosols cannot be visibly seen, the orientation on the filter is
known.
2. Place each Quickie Mount in the 40 position slide tray with the
black side to the front. This will assure the black face with
filter smooth aerosol side faces the detector. Note that the
position 1 corresponds to the sample identification 1 or 51 in the
identification sequence.
3. Place the filled slide tray to be prelasered in the LIPM system.
Adjust the calibration level to 0.750 using the vernier multipler
extension arm.
4. Push the slide in position L fully into the system. Observe and
record the value on the A-Preweight Logsheet under "Pre Laser".
Representative values should be between 0.350 and 0.500.
5. Periodically verify that the calibration level of 0.750 is
maintained when the slide changer is fully out. Continue until all
filter prel'aser values are obtained and recorded.
Page 34
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
6. Remove the 40 position slide tray from the LIPM system and orient it
such that the openings of the Quickie Mounts are vertical.
Carefully remove the filters sequentially and place them in their
corresponding petri dishes. After closing each petri dish, return
it to its petri box.
7. Repeat steps 4-6 for any remaining filters in the group of 50.
When the petri box is full (50 petri dishes), close it and attach a
l-l/2"x3" white Avery label to the container showing the sequential
filters enclosed, the date of the pre-LIPM analysis, and your
initials. Also include a space to enter the data for gravimetric
analysis.
Example: TF 4551.Ml thru 4600.Ml.
Prelasered 5/15/89 KW
Preweighed
Place the completed container and the A-Preweight Logsheet next to
the electrobalance to facilitate the pre-Mass analysis.
The post-LIPM analysis follows the post-MASS analysis (cf. section
III-E-1). At this point in the process, the exposed filters have been
mounted on permanent slide mounts arranged in temporary 40-position
linear slide trays. Each filter has been centered on the mount with the
aerosol face towards the black side of slide mount, and the sample date,
the filter-identification number, the site code, and an archive position
number have been written on the white side of the mount. Accompanying
the tray i's a Mounted Samples Logsheet. This logsheet identifies the
position, the sample-identification number, the filter-identification
number, the post-MASS and provides the location to record the post-LIPM
values. The procedures for the post-LIPM analasis are as follows.
1. Calibrate the LIPM system as in the pre-LIPM analysis.
2. Place the 40-position slide tray in the holder on the LIPM
apparatus. Adjust the vernier multiplier extension lever to read
0.750 with the slide changer fully out.
3. Sequentially cycle slide mounts and filters into the system. Record
the post-LIPM value for each sample in the Postlaser position on the
Mounted Samples Logsheet.
4. Periodically verify that the calibration level of 0.750 is
maintained when the slide changer is fully out.
5. When all samples in the tray have been analyzed, transfer the slide
mounts to the permanent linear slide trays organized by site.
6. Leave the LIPM system on, with the slide changer out and the cover
closed.
Page 35
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
7. Enter all the data from the Mounted Samples Logsheet into the
sample-handling database and put the logsheet in the permanent file.
C. PIXE/PESA Analysis
The UCD Particle Induced X-ray Emission Analysis (PIXE) system has been
optimized to analyze lightly loaded aerosol samples. The system uses a
4.5 MeV proton beam from the Crocker Nuclear Laboratory cyclotron to
excite x-rays in the sample. The analysis program controls the
cyclotron beam and the changing of the samples. The x-rays are measured
by two detectors. The first detector measures the entire spectrum from
Na to Pb, but is optimized for elements with x-rays below the main line
of Fe. The second detector is optimized for Fe and above. With the
two-detector system, it has been possible to reduce the minimum
detectable limit of elements heavier than Fe by a factor of more than 2.
The PESA system runs concurrently with the PIXE system, and determines
the concentration of hydrogen in the sample. The method is most
effective when the filter material does not contain any hydrogen, as is
the case with teflon.
1. Run the dBIII program to produce the instruction files for the
PIXE/PESA analysis. The program will determine the flow rates, the
elasped time, the volume/unit area (used in PIXE to calculate
concentration in mass/volume), the mass, the optical absorption and
other collection parameters and validity flags. The program will
control which samples to analyze and will permit the calculation of
concentrations at the time of analysis.
2. Run the PIXE system, using a source to verify the correct operation
of the detector system.
3. Run Che PIXE standards dray of approximately 40 standards. These
are single element, double element, and multi-element standards.
Verify the calibration of the system and update renormalization
values to compensate for small shifts in the system.
4. Run the PESA standards tray immediately after the PIXE standards
tray. This tray includes several mylar standards plus a series of
blank Teflon filters. Enter the calibration value from the mylar
standards into the analysis parameter file. Record the hydrogen
values for the blanks. Select the cleanest blank to be used as a
system blank for PIXE to estimate the background of x-rays in the
spectra.
5. Reanalyze a tray of filters from an IMPROVE site analyzed during the
previous IMPROVE analysis session. Compare the concentrations of
the two analyses for all major elements, including hydrogen. If the
Quality Assurance Manager and the PIXE/PESA Manager find the
calibration and reanalysis acceptable, begin actual analysis.
Page 36
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
6. Run the actual IMPROVE samples. Each tray contains all samples for
a given season from a single site. Initialize the analysis of the
samples in the tray by entering the identification code of the tray
in the acquisition program. Initialize the spectral analysis
program by entering the run number of the analysis. Check the
following features in the PIXE spectral analysis of the sample. If
any problems are detected, consult with the QA Manager, the
PIXE/PESA Manager, or a senior scientist.
a. Is the background completely removed by the two background
steps? That is, do the valleys between the peaks in the final
spectrum go to zero. If not, determine if the problem is an
inappropriate blank. Verify on the output log that the run
number of the blank is correct. Do not revise the fitting
parameters without consulting with the QA Manager or PIXE
Manager.
b. Periodically check the agreement of elements in the overlap
region of the two detectors. Make a note on the PIXE Runsheet
if the ratio of the two Fe's differs significantly from 1.0.
c. Monitor the livetime of the detectors and the cyclotron beam
current carefully. If the livetime drops below 50% or if the
beam integrated charge begins to decrease, ask the cyclotron
operator to either drop or raise the beam current.
d. After every three trays, while the next set is being loaded into
the system, run the QA program to generate the standard
correlation and time plots for the three sites just analyzed.
Note any anamolies for the QA Manager.
7. Rerun the standards at the end of the session and verify the
results.
8. Before ending the run, wait until the QA Manager is able to validate
the results for all of the trays. If problems in the analysis are
determined, select the trays and samples involved and reanalyze
them.
D. Ion Chromatograph Analysis (1C)
The procedures of the ion analysis by the ion contractor, Research
Triangle Institute, are given in complete detail in Appendix 6. This
section will summarize the procedures.
1. Receipt of the filter
The exposed filters are shipped from UCD to RTI in batches of 200
filters. Record the following data on the Sample Log Form when the
shipment is received: the sample indentification numbers, the date
of receipt, and any comments on the condition of the samples.
. Page 37
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
2. Filter extraction
a. Desorb the sample in 15 mL of desorbing solution.
b. Expose the sample and solution to ultrasonic energy for 30
minutes.
c. Allow the sample and solution to sit overnight.
3. Ion analysis
The analyses are performed on two ion chromatographs, Dionex Manual
Model 21201 and Model -14. The following calibration and standards
are run separately on the two units. The model used to analyze a
given sample is included in the reported data, although there does
not appear to be any bias between the models. The steps below are
performed daily.
a. Run a quality control standard (made at least weekly from
independently prepared stock solutions); proceed only if the
value agrees with the predicted value to within 10%.
b. Inject 100 mL of sample and solution into the 1C unit and
analyze for chloride, nitrite, nitrate, and sulfate.
c. At the midpoint of the day, determine the calibration of the
model by running a series of calibration solutions of varying
concentrations.
d. During each day, run a quality assurance sample (EPA Acid
Precipitation Audit sample) and reanalyze one sample (Duplicate
Sample).
e. Periodically run a blank sample of only the desorbing solution.
4. Data transfer to UCD
The data are transferred to UCD via ASCII files on floppy disk in
groups of at least one batch of samples. The following information
for chloride, nitride, nitrate, and sulfate is provided for each
batch. All values are in units of micrograms/filter.
a. The values, mean, and standard deviation of the quality control
standard for each 1C unit.
b. The values, mean, and standard deviation of the quality
assurance standard for each 1C unit.
c. The values for each sample. This includes dynamic field blanks,
travel blanks, laboratory blank, and other quality assurance
samples provided by UCD.
Page 38
-------
E. Thermal Optical Reflectance Analysis (TOR)
The procedures of the carbon analysis by the carbon contractor, Desert
Research Institute, are given in complete detail in Appendix 7. This
section will summarize the procedures.
The measurements are made using an OGC/DRI thermal/optical carbon
analyzer. The method is based on the preferential oxidation of organic
and elemental carbon compounds at different temperatures. It relies on
the fact that organic compounds can be volatilized from the sample
deposit in a helium atmosphere at a lower temperature than elemental
carbon. A small punch is removed from the sample and placed in the
analyzer. The temperature and oxidizing environment are changed in
time. This volatilizes carbon compounds, which are first converted to
carbon dioxide by passing the compounds over heated manganese dioxide;
the carbon dioxide is next reduced to methane. The amount of methane is
then measured by a flame ionization detector.
The principal function of the optical (laser reflectance) component of
the analyzer is to correct for the pyrolysis of organic compounds to
elemental carbon. Without this correction, the organic carbon fraction
of the sample would be underreported and the elemental carbon fraction
would include some pyrolyzed organic carbon. The correction is made by
continuously monitoring the optical reflectance of the filter and sample
using a laser and photodetector. This reflectance, largely dominated by
the presence of black elemental carbon, decreases as pyrolysis takes
place and increases as elemental carbon is liberated during the higher
temperature stages.
1. Receipt of the filter
The exposed filters are shipped from UCD to RTI in batches of 500
filters. Record the sample-identification number for each sample in
the Air Analysis Logbook upon receipt. Store tne samples in a
freezer until analysis.
2. Carbon analysis
a. Remove a punch from the filter and place the punch in the sample
boat.
b. Place the sample boat in the analyzer and begin the automatic
sequence.
c. At the end of the analysis, examine the thermogram for proper
laser response, temperature profiles, realistic carbon peaks,
and the presence of the calibration peak at the end of the
analysis. If a problem is found, analyze another punch from the
sample.
Page 39
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
d. Analyze replicate samples at a rate of one replicate per ten
samples.
3. Validate the analyses following the procedures described in Appendix
7.
4. Provide the results in micrograms/filter to UCD in ASCII files on
floppy disks. Include both the original and the replicate analyses.
Page 40
-------
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
VI. PROCEDURES FOR DATA PROCESSING
A. Introduction: The Equations of Concentration and Uncertainty
1. Volume
The volume is the product of the average flow rate and the sample
duration. The average flow rate is calculated the the gauge reading
recorded on the Field Logsheet and stored in the sample -handling
database. The average flow rate depends on the average temperature over
the sampling period, but not on the temperatures at the times of
measurement. The equations for the average flow rates from the
magnehelic and small gauge readings are:
magnehelic: AFM - a ( Mlb + M2b ) / 2 * (Tavg/Tcal) 1/2
small gauge: AFG - [c - d(Gl+G2)/2] * (Tavg/Tcal)1/2 .
The variables in these equations are:
Ml, M2 initial and final magnehelic readings
Gl , G2 initial and final small gauge readings
a,b,c,d calibration constants during audit
Teal absolute temperature at the time of audit
Tavg average absolute temperature during sampling
In Che following equations, V is the volume and fV is the fractional
uncertainty in volume. The fraction uncertainty in volume equals the
fractional uncertainty in flow rate, since the duration is well defined.
The uncertainty can be estimated from internal and 3rd-party audits.
The value determined in this manner includes both precision and
accuracy. The difficulty in making the estimate for the IMPROVE sampler'
is that the precision of the built-in flow measurement system is as good
as the precision of most audit devices. At present, the best estimate
of internal precision of average flow rate is that it is better than 1%.
and the best estimate of total uncertainty is that it is better than 3%.
All results from internal and 3rd-party audits are recorded in the audit:
database. All calculations are based on a volume uncertainty of 3%.
2. Gravimetric Mass
The equation for the mass concentration is
C - ( M - B ) / V,
where M is the mass difference between the post-MASS and pre-MASS, and 3
is the mass artifact determined from the mean of the controls and the
dynamic field blanks. The uncertainty in concentration is
aC - [ (fV C)2 + (aFB / V) j1/2,
where aFZ is the standard deviation in the controls and field blanks.
The minimum detectable limit (mdl) is 300 ng/nr .
Page 41
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
3. Optical Absorption
The equation for the coefficient for the particles on the filter depends
on the initial and final LIPM measurements, the volume, and the area.
The values are multiplied by 0.97 to compensate for scattering by the
particles. To calculate to coefficient in the atmosphere, it is
necessary to divide the measured coefficient by a factor R that depends
on the areal density of the particles on the filter. The coefficient in
the atmosphere is calculated by the equation
b - (area/V) * 104 * ln(PRE/POST) * 0.97 / R .
4. PIXE Analysis
•The artifact concentrations for the PIXE elements are zero. PIXE
determines the areal density of the given element; to determine
concentration, this areal density is multiplied by (area/V). The
uncertainty of the concentration is the quadratic sum of the volume
uncertainty, the analytical uncertainty of calibration (proportional to
measured value), and the statistical uncertainty associated with the
number of counts in each spectral peak. The 3% volume uncertainty and
the 4% PIXE calibration uncertainty add together to give a total
uncertainty of 5%. The total uncertainty is thus 5% plus statistics,
added quadratically.
5. PESA Analysis
The equations for PESA for hydrogen are the same as for PIXE. Again,
there is no significant hydrogen artifact on dynamic field blanks.
6. Ion Analysis
Analysis of field blanks indicate that there is artifact formation
during the period in the cassettes. The standard deviation of the field
blanks provides an estimate of the uncertainty of the artifact. The
data from replicate samples indicate that the analytical uncertainty is
proportional to measured value, rather than a constant. The equations
of concentration and uncertainty in concentration are
C - ( M - B ) / V,
oC - [(fV C)2+(<7FB/V)2+(fA*M/V)2 ]l/2,
where M is the mass measured by the ion analysis and fA is the
fractional uncertainty in the analysis. The mdl is equal Co twice^ the
uncertainty with no loading
mdl - 2 (aFB/V) .
7. Carbon Analysis
Page 42
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
Organic carbon artifact is caused by contamination in the filter
material, by contact with the cassette, and by the adsorption of organic
gases during collection. Elemental carbon artifact is caused by
contamination in the filter material and by contact with the cassette.
At present, we have insufficient data to to define the various equations
for organic and elemental carbon.
B. Entering the Data into the Concentration Database
1. Run the dBIII program FLOWS to calculate and print all the flow rates.
This program will select the appropriate calibration constants, obtain
the data from the sample-handling database, and calculate the flow
rates. (The calibration database contains a record of every flow rate
calibration with the constants and the date of measurement.) The flow
rates will be calculated and printed. Check all cases when the two flow
rates differ significantly, find the cause, and correct the information
in the sample-handling database.
2. Run the dBIII program TRAY to create the instruction files for PIXE/PESA
and begin the process of transferring data from the sample-handling
database to the concentrations database. This program will include the
following information in the instruction file, using the data from the
sample-handling database:
a. The module A average flow rate from the magnehelic readings.
b. The module A sample duration (elapsed time).
c. The volume / area.
d. The fine mass concentration and uncertainty.
e. The PM10 mass concentration and uncertainty,
f. The coefficient of absorption and uncertainty.
g. The filter-identification number for module A.
h. Information on the start time and date, and the site name.
i. Instructions on which samples to analyze.
3. These data are transmitted from this instruction file to the PIXE/PESA
spectra files, and from there to the PIXE/Pesa output files.
4. When the data from the external contractors are available for a complete
season, run the program IONS to generate similar output files for the
ion and carbon concentrations. This program combines the collection
data from the sample-handling database with the masses from the external
contractors to determine the concentrations and uncertainties.
Page 43
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
5. Run the program PROM to combine the output files from the last two steps
to produce concentrations database files. Each file contains all the
data for a single site and season; this information is encoded in the
file name (e.g. ACAD1A1D.C88).
C. Validating the Data
At this point, the data from the external contractors have undergone an
internal analytical validation process. The gravimetric mass and absorption
values have been check for consistency. • The PIXE/PESA analyses have been
check for internal consistency. In this segment of the procedures the
different variables are intercompared, and the results are examined for
anomalous variations with time.
The procedures here are followed by the Quality Assurance Manager.
1. Run the programs CORFIL and CORCOF, using the data in the concentrations
database, to generate desired correlation plots for all sites. This
should be done for the following pairs: Si and Fe, S*3 (Teflon,PIXE)
and S04-(nylon,1C), mass and H, mass and reconstructed mass, OC and
organic mass by hydrogen and sulfur.
2. Run the program IMPSUM to generate time plots of major variables.
3. Run the program DBP to generate a wide variety of statistical
comparisons.
4. Check the results for systematic variations.
5. Check individual anomalies; look for errors in transcribing data.
D. Preparing Magnetic Tapes and Floppy Disks
1. Run program DBP to create ASCII versions of the concentrations database
files.
2. PIP the files to the magnetic tape or to a floppy disk, as desired.
E. Preparing the Seasonal Summaries
1. Run the program IMPPG2 to prepare a table of concentrations and a table
of averages for key physical variables at each site for the season.
2. Run the program IMPSUM to prepare a page of time plots of key physical
variables for each site for three-month period.
3. Use the averages files from IMPPG2 to prepare contour maps of key
physical variables for the network.
Page 44
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
VII. PROCEDURES FOR SAMPLER MAINTENANCE
The following procedures are to be followed by the Field Engineer and the Field
Technician. Sections A, B, and C are to be followed at Davis, while section D is
done in the field.
A. Evaluating Sampler Modifications
If a modification to the sampler is being considered, first set up an
experiment at the Davis field site to test the modification and its impact
on the entire sampling process thoroughly. (The modifications will normally
be made during the annual site visit.)
B. Calibrating the Flow Audit Device
The flow audit device is used to audit the flow rate of the samplers. The
device is first calibrated at UCD using a spirometer and a dry gas meter.
1. Set up the calibration system to be the same as with a normal sampling
system: the air enters through the measuring orifice, then passes
through the cyclone, a cassette/valve assembly to imitate a loaded
filter, a critical orifice, and the pump. Connect the survey spirometer
to the exhaust side of the pump. The spirometer measures the displaced
volume over a given time.
2. Start the calibration by turning on the pump and adjusting the valve so
that the reading on the audit magnehelic is 1.0 "H20.
3. Start the spirometer chart recorder rotating and mark the displacement
in liters. Repeat for two to four revolutions of the chart recorder, or
until the measurements are consistent.
4. Compute the flow and record the flow on the Audit Calibration Logsheet.
5. Verify the flow rate using the dry gas meter. Install the meter Co the
intake of the cyclone and take a 5-minute measurement. Repeat the
calibration procedure if the spirometer and dry gas meter measurements
do not agree.
6. Repeat the calibration procedure for magnehelic readings of 1.0, 0.8,
0.7, 0.6, and 0.4 "H20.
7. Calculate a best-fitting logarithmic equation, following the procedures
of Appendix 2.
8. Record the date, temperature, and atmospheric pressure at the time of
calibration.
9. Label the orifice meter with the calibration date, temperature,
pressure, and 5-point logarithmic equation.
Page 45
-------
JUJ-Y
10. Enter the orifice meter calibration data into the audit database.
C. Preparation for Annual Site Visit
Each site will be visited at least once per year, normally in spring or
summer. The steps prior to the visit are
1. Notify the appropriate land management departments about one month prior
to the scheduled site visit with information on the site, approximate
date of the visit, and the name of the visiting technician.
2. Notify the site operator two weeks prior to scheduled visit. Coordinate
the exact date of arrival with the site operator. Determine if the site
operator is having any problems with the sampler and/or sample handling.
3. Conduct a Site Review
Prepare a folder for each site to contain a site checklist, a
Maintenance Checklist (Appendix 8), a copy of the previous flow rate
calibration, and any other pertinent information.
a. Review the site summary for the previous year and note any problems
on the site checklist.
b. Review the description of site location and note any missing
information, such as site pictures.
c. Review the Field Logsheets for the previous year and note any
problems with flow rates and defective parts on the site checklist.
d. Consult with the Sample Handling Laboratory Supervisor about any
special problems for the site that should be considered during the
visit.
e. Review the concentration data with the Quality Assurance Manager,
with special emphasis on (1) any inconsistencies that may be
attributed" to faulty sampler operation, and (2) any change in
elemental composition indicating changing emissions. Record any
significant results on.the site checklist.
D. Annual Site Visit
The site visit will include any necessary modifications to the sampler,
replacement of the nitrate denuder, a complete inspection of the entire
sampler and replacement of any defective or deteriorating parts (such as
hoses and 0-rings), an audit of the flow rate calibration, and optional
operator training. The steps are
Page 46
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
1. Review the performance of the site with the field operator. Review the
site information summary sheet. Obtain any missing information for the
description of site locations, including taking site pictures.
2. Perform an initial inspection following the Maintenance Checklist
(Appendix 8).
a. Take the final readings of any exposed filter and record the values
on the Field Logsheet.
b. Inspect the controller module.
Read the program of the control clock. Check the control clock
override switches by turning the sampler off. Turn the bypass timer
to thirty minutes and verify the delayed pump startup. Record any
damage on the Maintenance Checklist.
c. Inspect the filter modules.
Record on the Maintenance Checklist any damage to solenoid values,
elapsed timers, relays, toggle switches pressure guage and
magnehelic. Check the maximum vacuum of all of the pumps. Verify
the proper installation of cassettes. If they are improperly
installed, discuss the proper procedures with the site operator.
d. Inspect the pumphouse and the pumps and record any damaged parts.
3. Conduct the initial flow rate audit.
a. Fill out sampler calibraton log information. Include the data,
altitude, temperature and orifice meter calibration constants.
fa. Install the calibration cassettes in filter 1 for each module.
c. Leak-test the system,
d. Return the bypass timer to thirty minutes.
e. Audit the four flow rates indicated on the sampler calibration
logsheet for each module.
f. Calculate the orifice meter value corresponding to the indicated
flow rates. Adjust for the ambient pressure and temperature using
the calculated correction factor.
g. Press the filter 1 switch and record the vacuum ("Hg) and magnehelic
values.
h. Remove the aluminum cap at the bottom of the stack and insert the
orifice meter into the stack opening.
Page 47
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
i. Press the filter 1 switch and record the value from the orifice
meter magnehelic. (This value should correspond to 23.0 1pm.)
j. Install the valve inline between the filter and the critical
orifice.
k. With the orifice meter in place, press the filter 1 switch and
adjust the valve until the orifice meter magnehelic indicates the
desired flow rate.
1. Remove the orifice meter from the bottom of the stack and insert the
aluminum cap.
m. Press the filter 1 switch and record the vacuum ("Hg) and magnehelic
values.
n. Repeat step c.
o. Repeat steps f, g, and h.
p. For auditing flows higher than the nominal flow rate, use an empty
cassette.
q. Repeat steps i and j until you have finished audit.
r. For the PM-10 module audit, insert the orifice meter where the stack
attaches to the cassette holder and leave it in throughout the audit
procedure.
s. Zero all the magnehelics.
Perform the scheduled maintenance following the Maintenance Checklist
(Appendix 8) ,
Final Calibration
a. Calibrate the fine filter modules to 2 . 5yum cut point by inserting a
23.0 1pm critical orifice.
b. Calibrate the PM-10 module to lO-O^m cut point by inserting an 18.9
1pm critical orifice.
c. Check all cyclone 0-rings to insure that they are seated.
d. Perform leak-test.
e. Perform the multipoint flow audit again.
f. Use a permanent pen to mark the maximum, nominal, and minimal flow
values on the faces of the magnehelic and vacuum gauges .
Page 48
-------
IMPROVE STANDARD OPERATING PROCEDURES JULY 1989
•g. Replace the calibration cassettes with the proper monitoring
cassettes.
h. Record the initial readings on the Field Logsheet.
i. If the audit took place on Wednesday or Saturday, start the sampler
by pressing the appropriate override switch. Record the length of
time the sampler was not operating on the Field Logsheet.
j. Calculate the logarithmic equations for all modules using the data
collected in the multipoint audit.
6. Enter the Calibration Equations
After returning from the sites, verify the logarithmic equations for all
modules at every site. Input the site visit information into the
calibration database system. Record the logarithmic equations, the site
code, the temperature at the time of the final multipoint audit in
degrees Celsius, the elevation factor, and the date.
Page 49
-------
-------
APPENDICES
APPENDIX 1: Logsheets Used in Sample Handling
APPENDIX 2: IMPROVE Sampler Manual
APPENDIX 3: Gravimetric Mass Startup Procedures
A. Cleaning and Calibration of the Electrobalance
B. IMPROVE Gravimetric Controls
APPENDIX 4: LIPM Startup Procedures
APPENDIX 5: PIXE/PESA Procedures
APPENDIX 6: Ion Contractor Procedures (RTI)
APPENDIX 7: Carbon Contractor Procedures (DRI)
APPENDIX 8: Maintenance Checklist (Annual Site Visit)
-------
Appendix A—1
Logsheets Used in Sample Handling
-------
-------
IMPROVE SOP APPEND JULY 1989
APPENDIX 1: Logsheets Used in Sample Handling
A. Balance Log--Record of calibration date and significant events for
each balance at UCD.
B. LIPM Calibration Logsheet--Record of calibration data for UCD LIPM
system.
C. IMPROVE Dot Chart--Portion of large chart posted on the laboratory
wall, indicating the status of all filters by the use of several
colors.
D. A-Preweight Logsheet--Record of data for module A filters, before
collection.
E. D-Preweight Logsheet--Record of data for module D filters, before
collection.
F. Mailer Record Logcard--4"x6" card, record of date that each filter
was shipped to sampling site.
G. Field Logsheet--Record of data for sample change. The front side is
for clean filters and the reverse side for exposed filters.
H. Mounted Samples Logsheet--Record of data for module A filters, after
collection.
I. PM10 Archive Logsheet--Record of data for module D filters, after
collection. Accompanies each box of 50 filters in PM10 Archives.
J. External Contractor Inventory--Sample identification numbers of all
exposed filters sent to the Desert Research Institute or to the
Research Triangle Institute for analysis. Accompanies each box of 50
filters in shipment.
K. Unusable Archive Inventory--Record of sample identification numbers
with problem code for samples that are damaged or otherwise invalid.
Accompanies each box of 50 filters in archive.
-------
IMPROVE SOP APPEND
JULY 1989
A. Balance Log--Record of calibration date and significant events for
each balance at UCD.
!>.•»( o
Temp
Tlmf
7.pro
.003/
.000
Cnl lbr.il Ion
199. 976/
200. OOO
r>0 . OOO
mp, . si d.
50.011
20.000
mg. std.
20.OOA
1 n ( 1 1 .1 1
-------
IMPROVE SOP APPEND
JULY 1989
B. LIPM Calibration Logsheet--Record of calibration data for UCD LIPM
system.
UATE/Initials
H 82 83
c
,O5O
o'f'fe
7Y\
30?
.-*/!
LASER CALIBRATION
«5 116 «7 «8 «9 I S10
.SZ8
• 25'J
.!
-------
IMPROVE SOP APPEND JULY 1989
C TMPROVE Dot Chart--Portion of large chart posted on the laboratory
' wall, indicating the status of all filters by the use of several
colors.
HBOCStBBS -OS OUT IW1£, JSBWZBIIwa.
ia^K&ai i0101 '\ s\ i f\ •*! i i i i i i ii i i i i i i i i i i i i r
"\ "i i i i i i i i i i i i i i i i i i i i i i i
i i i i i i i i i i i i i i i
i i i i i i i i i i i i i i i
_. . , . ..
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
t f\ \ i i i i
I I ) I I I I I I I I ! I I I I I I I I I I I I I
'\ '\ '\ '\
'\ f\ '\ r
'i 'l 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
'l 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
i i i i i i i i i i i i i i i i i i i i i i i i
DOM 01 '\ ^ r] *T1
1 1
101. 1" I I I I I I I I I I I I I I I I I I I I I I I I I
iu/i "r •*! *r '\ '\ i i i i i ii i i ii T i i i ii i i i i
'l *"l 'l I I I I I I I I I I I I I I I I I I I I I I I
-------
IMPROVE SOP APPEND
JULY 1989
D. A-Preweleht Logsheet--Record of data for module A filters, before
collection.
PREWEIGHTS PREWEIGHTS
IMPROVE "A" FILTERS, MASKED
Identfftcslton | Pre Wvlghl
iTF 0S&I. Ml
2TF ttTPZ.. Ml
3 TF &70 3 .H!
4 TF #S7>/. M/
s TF £^. Ml
7. TF &S£iJ:lJ
s. TF- ££rc?£Lk(j
9 TF f-?^-}, MI
10 TF f^/D, Ml
11 TF P)S/I . HI
12 TT P>5"/4?, HI
is TF ft£/3. M/
14 TF ^/^ Hi
V^, /$*?
<-//./ ^
y^< /^^
L/ 1 . {;££)
15 TF- h5/5" H)
16 TT «£r///, M/
17 TF fi5/?._MJ i
is TF b5/8. Ml
19 TF £6"/9 /U/
20 TF f->5=2D /-( |
21 TF fc «"2 I Ml
22 TF 0,^52.. HI
PreLMer
0. '^W2
0- 3fr 1
D. 4M-0
O , H 31
(p , Lf (f^
O ' fLfL-
0 . T^T?
O_. ^5/
Idwtlinaflon
26 TF AC4J/a.Ml
27 TF ^521- .Ml
28 TF £63B. M I
26 TT- P;^ ^cf )-/ /
30 TF ft5 30. M 1
31. TF flS4?!. Ml
32. TF ^32 M)
33 TF ft5^?^? Ml
^j, L/97 j 34 TF P>5 B'i , M 1
C <-P~\i.
D ^L
O •^"*~\
D . ^^&
f^ LL£~ *T
t_' ' J^ ^
/"7 , C/£j /—
D. ^^"9
0 . <^f. «?
D.^0
£> , 45;^
o, V-T"?-
35 TF ££r*r. Ml
36 TF ^5 3^5 , M /
37 TF £537. M/
ss TF p-.53%. M/
Pre Weight
Pre Laser
O -H T I
O.^WD
o . f i T
/^ u ^ i
/"^i W-O^»
£ , 40?-
0 -395
D ^tob
O . 2,>0
O . 3T-/?
O , VC6"
n . ^.-6"
D. W4 !
39 TT- f?S39. H/ /O. ^D? i
40 Tr H5W, M/
41 TF £^74 / , U )
42 TF ^542. M/
43. TF J55"_
-------
IMPROVE SOP APPEND
JULY 1989
E. D-Preweight Logsheet--Record of data for module D filters, before
collection.
PREWEIGHTS
PREWEIGHTS
TP fr I
U
23.
IMPROVE TV FILTERS, UNMASKED
McntmatkNi
1.TF (0l5\. U
2.T? to/52 U
3. TF" fc(=;?. l(
4 TF faisH. U
5 TT" l/> IS5", U
e. 7T fc i 5 i, . U
7. TP It \S1.U
B. TF It IS& U
s TF (s\S9. 6/
20 7T kil-D, U
21 TF fo / ~i \ . I ,/
22 TF /r n 7 IJ
24 TP u 1 74- ///
25 TP Ic 1 15 . tJ
MwtDflcaflon
26. TF
27 TF
28. TF
ki^b.U
h m u
/el TB. iJ
29 TF /e/ ? 9. L/
so TF
31. TF
/ol&D.ti
fa { ft 1 , L/
32. TF /oi&Z-.U
33. TF
34. TF
Lol %3. U
klft'i-. U
35 TF (el %. Li
36 -TF fc>l&tn.lJ
37. TF fr>iP>T-.IJ
38. TF Lf\r)n,\A
39 TF L"/?^. /x/
40 TF bl^D.U
41. TF /firtl.U
42 TF
/0/92.L/
43. TF Ce/93 M
44. TF 0/9f, L/
45. TF (tftS.U
46 TF
47 Tp
48 TF
-------
IMPROVE SOP APPEND
JULY 1989
F. Mailer Record Logcard--4"x6" card, record of date that each filter
was shipped to sampling site.
BOX f
DATE OUT
Channel
Module
Code
Al
A2
Bl
B2
Cl
C2
Dl
02
SI
S2
Teflon Filter
ID
Ounrtz Filter
Lot 1
S02 Filter
Lot I?
-------
IMPROVE SOP APPEND JULY 1989
G. Field Logsheet--Record of data for sample change. The front side is
for clean filters and the reverse side for exposed filters.
Samlcr Pkrtieulae* sa^lin? mewocfc ri«ld
1; (bl*di)
rtlttr 2: (**UM)
a on aaacltt
Svaplvt Particulat* SM^linq Mtwork ritld
Operatori
<*«*• /_ ti«»: dayi Sun nan Tut Initial!
nodule A («DI
FUtet 1: (black)
nodul* A («DI ^ 2_ (¥||1
filter i. (blacJt) ' .
fllttr 2: iMftiut . __^_ nedult • (Ylxioil
Filter li (black)
Filter 2: 1«*iltel
niter :• iwttiwi noduu c (caosNt
riit*r 1 (black)
Filter 2: iwttitt)
nodule D (BLUE)
Filter I: (black)
, , iiux. rllMr Jt (-|tit|
'— 1: (blao*t)
" wsamlvr i
-------
IMPROVE SOP APPEND JULY 1989
H. Mounted Samples Loesheet--Record of data for module A filters, after
collection.
IMPROVE/CRITERIA MOUNTED SAMPLES
. SAMPLE ID FILTER ID POST WEIGHT POST LASER
(71 (MO Qs-.oq.RTi fit mi^MI
-------
IMPROVE SOP APPEND JULY 1989
I. PM10 Archive Logsheet--Record of data for module D filters, after
collection. Accompanies each box of 50 filters in PM10 Archives.
IMPROVE/CRITERIA ARCHIVED m SAMPLES
SAMPLE ID FILTER ID POST WEIGHT SAMPLE ID FILTER ID POST WEIGHT
I 42.2-01 |
OS
-------
IMPROVE SOP APPEND JULY 1989
J. External Contractor Inventory--Sample identification numbers of all
exposed filters sent to the Desert Research Institute or to the
Research Triangle Institute for analysis. Accompanies each box of 50
filters in shipment.
F1I.TER INVENTORY FOR EXTERNAL ANALYSIS BY 1^(. X. ~ C1
Filter Iocnttfir.lt Ion Fitter IHptit I f (c.illoii
Site Date Channel Site Date Channel
i- f*iGu og-pq. en of z»._
i. '&)km as -en. en czs 27.
3. iDSt C-Mfo'gf) C.-IS 28.
10.
6. 31.
7. 17.
fl. 13._
9. I 3'.. _
10. 35.
11. 36. _
12. 37._
13. 38
14. 39._
15. 40._
16. 41._
17. 42.
IB. 43._
19. 44.
2O- <••).
21. 66.
22. 47.
23. '.a._
24. 49.
7'j. 50.
InventorIctl liy
-------
IMPROVE SOP APPEND
JULY 1989
K. Unusable Archive Inventory--Record of sample identification numbers
with problem code for samples that are damaged or otherwise invalid.
Accompanies each box of 50 filters in archive.
ARCHIVE OF UN-USABLE IMPROVE/CRITERIA SAMPLES
PROBLEM
IDENTIFICATION FILTER ID CODE
1 SKMD /pV
z7T)}JT'c»J:lt'?>cif\\
3. CCL.A ctf- /*.#n? i
4 ^F.LL t4'2b'ffir>\
5.
6.
7,
8
9.
10.
11.
12.
13.
14
15.
16
17.
18.
19.
20
21
22.
23
M.
25
TF- 7V9.MI
TF-?«f5l>U
__
TF «rt M
yy
?H
XX
T>R
PROBLEM
IDENTIFICATION FILTER ID CODE
26
27
28
29.
30.
31.
32
33
34
35
36.
37
38
39.
40
41.
42.
43.
44.
45
46.
47
46
49.
50
P-i
i
ARCHIVE
-------
-------
IMPROVE Sampler Manual
Version 2
January 1988
Robert A. Eldred
Air Quality Group, Crocker Nuclear Laboratory
University of California, Davis CA 95616
Table of Contents
1. Overview 3
2. General Description of the Sampler 6
3. Site Location 9
3.1 Site Selection Criteria 9
3.2 Power Requirements 9
4. Sample Changing Procedures 10
4.1 General Description 10
4.2 Details of Sample Change 12
4.3 Trouble Shooting 16
5. Control Module 18
5.1 General Description 18
5.2 Operating the Time Clock 18
5.3 Programing the Time Clock 19
5.4 Electrical Circuitry 21
6. Filter Module 23
6.1 General Description 23
6.2 Flow Rate Measurement 24
6.3 Flow Rate Audit 26
6.4 Flow Rate Calibration 28
6.5 Electrical Circuitry 32
6.6 Nitrate Denuder 33
7. Pump House 34
7.1 General Description 34
7.2 Pump Specifications 34
7.3 Electrical Circuitry 35
8. Sampler Stand 37
9. Acknowledgments 39
-------
IMPROVE Sampler Manual
Version 2, 1988
Improve Sampler Manual
Version 2
January 1988
List of Table and Figures
Table 1. List of IMPROVE and NPS Criteria sampling sites .... 4
Table 2. Density factor at IMPROVE sites 30
Table 3. Density factor vs. altitude 30
Figure 1. Map of IMPROVE and NPS Criteria sampling sites 4
Figure 2. Layout of IMPROVE sampler, without pump house 7
Figure 3. IMPROVE sampler field log 11
Figure 4. Reading the min/max thermometer 13
Figure 5. Layout of fine filter module 14
Figure 6. Layout of PMlO filter module 15
Figure 7. Electronic time clock with cover plate removed 18
Figure 8. Time clock programs for standard IMPROVE cycle 20
Figure 9. Electrical schematic for control module 22
Figure 10. 50% cutpoint for cyclone vs. flow rate 23
Figure 11. Flow rate audit log 27
Figure 12. Flow rate calibration log 29
Figure 13. Electrical schematic for filter module 32
Figure 14. Nitrate denuder 33
Figure 15. Layout of wall of pump house 34
Figure 16. Electrical schematic for pump house 35
Figure 17. Electrical schematic for relay box 36
Figure 18. Outdoor stand for IMPROVE sampler 38
-------
-------
IMPROVE Sampler Manual
Version 2, 1988
1. OVERVIEW
Fine aerosol particles affect remote areas primarily by impairing
visibility and secondarily by producing acid precipitation. These fine
particles are generally manmade, although some are produced by smoke and
windblown dust. (Most naturally produced particles are coarse and have a
smaller effect on visibility and acid rain..) In the case of sulfur, fine
sulfate particles are produced by the transformation of sulfur dioxide gas
in the atmosphere. Measurements of the concentration and composition of
these fine particles are necessary to determine the extent of the problem
and possible sources of the particles.
The National Park Service and the Environmental Protection Agency have been
monitoring particulate concentrations at national parks, monuments,
forests, wildlife refuges and other remote sites since 1979 using stacked
filter samplers. Coarse particles are collected on nuclepore filters and
fine particles on teflon filters. The coarse filters are analyzed for
mass, while the fine particles are analyzed for mass, optical absorption,
hydrogen and elements Na to Pb, including sulfur and the soil elements.
From the measured data we can calculate the concentration of organic
particles.
The particulate monitoring program has been expanded to include other
federal agencies with the establishment of the IMPROVE (Interagency
Monitoring of PROtected Visual Environments) program, designed to determine
the extent and causes of visibility impairment at selected class I areas
throughout the United States. The National Park Service maintains
additional sampling sites through the NFS Criteria Pollutant Monitoring
program. The two programs use the same sampler and nearly identical
sampling protocols and are operated by the Air Quality Group of Crocker
Nuclear Laboratory at the University of California at Davis. The sites
selected as of October 1987 are shown in Figure 1 and listed in Table 1.
A new sampler was designed for these networks called the IMPROVE Modular
Aerosol Monitoring Sampler that collects three samples of fine particles
(smaller than 2.5 tm) and one of respirable particles (smaller than 10 pm).
At the NFS Criteria Pollutant Monitoring sites, a fifth filter measures
gaseous S02. The entire unit is modular, with four filter modules, a
controller module and a pump house containing four pumps. The modules are
mounted either on an outdoor wood stand or on a wall of an air quality
building. The IMPROVE sampler retains the simplicity of the stacked filter
sampler but adds several features, including additional filters for
measuring nitrates and carbon, twice the flow rate to improve sensitivity,
an improved flow rate measurement system, and fewer sample changes.
The samplers require a weekly change of filter cassettes by field personnel
provided by the cooperating federal agencies. Each site receives from
Davis a weekly box of 8 filter cassettes and a log sheet. After the filter
change, the 8 cassettes of exposed filters are returned to Davis for
analysis.
-3-
-------
IMPROVE Sampler Manual
Version 2, 1988
Figure 1. Map of IMPROVE and NPS Criteria Sites
Table Is List of IMPROVE and NPS Criteria Sites
IMPROVE Network
ACADIA National Park
BIG BEND National Park
BRIDGER Wilderness
BRYCE CANYON National Park
CANYONLANDS National Park
CHIRICAHUA National Monument
CRATER LAKE National Park
DENALI National Park and Preserve*
GLACIER National Park*
GRAND CANYON National Park
GREAT SMOKY MOUNTAINS National Park
JARBIDGE Wilderness
MESA VERDE National Park
MOUNT RAINIER National Park
ROCKY MOUNTAIN National Park
SAN GORGONIO Wilderness
SHENANDOAH National Park
SUPERSTITION Wilderness
WEMINUCHE Wilderness
YOSEMITE National Park
NPS Criteria Network
ARCHES National Park
BADLANDS National Park
BANDELIER National Monument
GREAT SAND DUNES National Monument
GUADALUPE MOUNTAINS National Park
HALEAKALA National Park
HAWAII VOLCANOES National Park
ISLE ROYALE National Park
LASSEN VOLCANIC National Park
PETRIFIED FOREST National Park
PINNACLES National Monument
POINT REYES National Seashore
REDWOOD National Park
VIRGIN ISLANDS National Park
VOYAGEURS National Park
YELLOWSTONE National Park
* Also an NPS Criteria Site
-4-
-------
IMPROVE Sampler Manual
Version 2, 1988
In order to obtain a complete signature of the composition of the
particles, a variety of analytical methods are used.
A fine teflon filter — gravimetric analysis (mass)
laser integrating plate (optical absorption)
Particle Induced X-ray Emission (Na to Pb)
Proton Elastic Scattering (hydrogen)
Forward Alpha Scattering (H, C, N and 0)
B fine nylon filter ion chromatography (nitrate)
C fine quartz filter — combustion analyzer (organic carbon
and elemental carbon)
D respirable particle teflon filter — gravimetric analysis (mass)
S impregnated quartz filter — ion chromatography (S02)
Filters A and D are analyzed by the Air Quality Group at Davis
Filter B is analyzed by Research Triangle Institute
Filters C and S are analyzed by Desert Research Institute
The analytical results are included in 3-month seasonal summaries, which
are distributed to the cooperating agencies and to the local sites. In
addition, interpretative studies are performed by the participating
contractors, the cooperating agencies and by other research groups. The
data are available to local resource managers on a variety of media.
-------
IMPROVE Sampler Manual
Version 2, 1988
2. CTNERAL DESCRIPTION OF HIE SAMPLER
The IMPROVE modular aerosol monitoring sampler was designed and built by
the Air Quality Group at Davis specifically for the IMPROVE program. The
following design criteria were followed:
—Accuracy and precision in particulate collection including air volumes,
particle size cuts and time intervals.
—Reliability and ruggedness under extreme environmental conditions,
without requiring air conditioned shelters.
—Ease of siting and operation, with protocols to eliminate potential field
errors and to minimize the time demands on the field personnel.
—Flexibility in terms of present and future sampling requirements, with
ability to increase or decrease the number of simultaneous samples and to
match the system sensitivity to the specific site.
—Attractive and professional appearance, to permit inclusion in public
displays, if desired.
—Reasonable costs.
The sampler is composed of six modular units: a control module, four
filter modules and a pump house, as shown in Figure 2. The control and
filter modules are all contained in identical gray fiberglass enclosures,
20 inches high by 17 inches wide by 11 inches deep. These are mounted
either outdoors on a wood stand with sunshield and work table, or on the
walls of an air quality building. The aluminum pump house measuring 3 feet
wide by 2 feet deep by 4 feet high will contain four pumps, a cooling fan
and two heaters. The hoses and wires are enclosed in teflon coated metal
conduit.
The control module contains a 7-day, 4-channel electronic time clock, a
30-minute bypass timer for sample changing plus appropriate relays. The
module has a cooling fan that turns on at 85°F and a neater that turns on
at 20°F.
-fi-
-------
IMPROVE Sampler Manual
Version 2, 1988
sec.
Figure 2. Layout of IMPROVE modular Aerosol Monitoring Sampler,
without pumphouse
The three fine particle filter modules are almost identical. The inlet
protects against rain and eliminates bugs and particles larger than around
15 urn. One of the units has a nitrate denuder to remove gaseous oxides of
nitrogen from the air. The air passes through identical cyclones, designed
to remove particles larger than 2.5 im at a flow rate of 21.7 I/fain. The
airstream passes through a filter, which collects all the fine particles.
In the standard configuration, each module contains two solenoids and two
elapsed time indicators to handle two filter cassettes. The unit is
designed to be able to accomodate up to four cassettes by adding additional
solenoids and elapsed time indicators. Each unit also contains two gauges
to measure the flow rate and two toggle switches for use during sample
changing. Each module is connected to its own pump in the pump house.
-------
IMPROVE Sampler Manual
Version 2, 1988
The PM10 module differs from the fine modules in that the fine inlet and
cyclone are replaced by a commercial PMlO inlet similar to those used for
standard dichotomous samplers (virtual impactors), but designed to operate
at 18.9 1/min. All particles smaller than 10 j/m are collected on the
filter. The filter cassettes are mounted vertically up rather than down as
in the fine modules.
The flow rate of 21.7 or 18.9 1/min is regulated by a critical orifice
calibrated for the intended filter medium. The flow rate is measured
before and after the collection by two independent methods. The first
measures the pressure drop across the filter and uses the equation for flow
rate through a critical orifice. The second method calculates the flow
rate by measuring the pressure drop across a fixed orifice.
The filters are transported and installed in cassettes; the filters are
handled only under controlled laboratory conditions. This is essential in
maintaining quality control and in simplifying field protocols. The
nitrate module uses 47mm cassettes while the others use 25mm cassettes.
The active area of the sample collection on the fine teflon is reduced from
3.8 mm to 2.2 cm by using masks in the cassette behind the filter, in order
to improve the analytical minimum detectable limit for elemental analysis .
The use of cassettes permits adding extra filters to the system without
revising the field protocols. For example, at some sites, the PMlO filter
will be followed by an impregnated quartz filter to capture gaseous S02.
The following filter media are to be used in the standard system, along
with the corresponding analytical measurements:
module A 25mm teflon fine mass, absorption, elemental (H, Na-Pb)
module B 47mm nylasorb fine nitrate
module C 25mm quartz fine organic and elemental carbon
module D 25mm teflon PMlO mass
-8-
-------
IMPROVE Sampler Manual
Version 2, 1988
3. SITE LOCATION
3.1 Site Selection Criteria
The specific sampling site should be selected with the following criteria:
—The site must be away front local combustion sources, such as diesel
generators, automobiles, chimneys, and dumps.
—The site must not have obstructions that would hinder sampling
representative aerosols, such as trees or buildings.
—The site should not be located in small valleys subject to nonregional
conditions.
—The site must have 25A of current at 120V.
—The site must be accessible for filter changes in all weather conditions.
—The site should be located near existing particulate monitoring stations
in order to provide continuity.
3.2 Power Requirements
The maximum current for the standard system is 23 amps and lasts for a few
seconds during the startup of the pumps. The normal current with the pumps
in operation is 14 amps. The heater in the controller module adds 0.7 amps
and a 60W lamp in the pumphouse adds 0.5 amps. The annual energy
consumption is 4,000 kWh.
-9-
-------
IMPROVE Sampler Manual
Version 2, 1988
4. SAMPLE CHANGING PROCEDURES
4.1 General Description
This section summarizes the protocols to be followed for the weekly sample
change. Section 4.2 will provide more details on the procedures and
section 4.3 will discuss trouble shooting. The instructions are posted
inside the doors of the control module, module A and module D. You will
need last week's shipping box to put in the cassettes currently in the
sampler and this week's shipping box with the new cassettes.
The filters.are to be changed once a week, either on Sunday, Monday or
Tuesday. Because of other air quality instruments, most operators will
make the change on Tuesday. The shipping boxes containing the clean
filters will be identified by the date for Tuesday. The box should be
received 1 to 2 weeks before the specified date. A receipt log is provided
to document the receipt of these boxes. If the box is not present at the
time it is to be used, please call Davis to discuss procedures. We will
have you use another box if available. In any case, you must remove the
exposed cassettes before Wednesday: if any filter is run for two periods
it will be invalid. You will need to have at the site the empty shipping
box for the exposed filters in the sampler and the full shipping box with
clean filters.
Inside the shipping box will be eight cassettes and a log sheet. Each
cassette will be identified by both colored tape and a coded label. The
two systems are redundant; if you follow the colored tape you can ignore
the code. If "date" is the Tuesday date on the shipping box, the cassettes
will be as follows:
MODULE FILTER COLORED TAPES LABEL DESCRIPTION
A 1 red, black date-Al single 25mm cassette
A 2 red, white date-A2 (teflon)
B 1 yellow, black date-Bl single 47mm cassette
B 2 yellow, white date-B2 (nylasorb)
C 1 green, black date-Cl double 25mm cassette
C 2 green, white date-C2 (quartz)
D 1 blue, black date-Dl single 25mm cassette
D 2 blue, white date-D2 (teflon)*
*If S02 is to be measured at NPS Criteria Pollutant Monitoring
site, the D filter will be a double 25mm cassette.
The log sheet is two-sided, with the initial data on one side and the final
data on the other. These are shown in Figure 3. The initial data is to be
entered after inserting the clean filters. The final data is to be entered
the following week before removing the exposed filters. The log sheets
correspond to the 8 cassettes in the shipping box. The log sheets must
remain with the shipping box.
-10-
-------
IMPROVE Sampler Manual
Version 2, 1988
nvKNI _»f>l*t rntle-UU (Mpllnf Wtwk M«M la*
DATUM
IMUf tn_»rtlnt elna -_u*ttMl
Initial!
•Mil }«lfl
-
Hfnchtlle
f«•(•!>
NBrfuK » OBI
riu«t It (MxM
MlUt It t
»i«w
HffDttta (•>(«» tnmrinf i)q fton Tu» InltUll
Operator!
r»"tel
Figure 3. IMPROVE sampler field log.
-11-
-------
IMPROVE Sampler Manual
Version 2, 1988
The protocols to be followed at the sampler for the weekly sample change
involve 5 steps.
step 1: Record the following on the reverse side of last week's log
sheet: date; current, min and max temperatures; your
initials; and any comments on sampler problems, construction
near sampler, etc.
step 2: Check the time and date on the controller clock and turn timer
to 30 minutes.
step 3: For each module (A,B,C,D) record 5 gauge readings and 2 elapsed
times.
step 4: For each module (A,B,C,D) remove the exposed cassettes, placing
in last week's shipping box along with the log sheet.
step 5: Get the cassettes and the log sheet in this week's shipping box.
For each module (A,B,C,D) insert the new cassettes, matching
colored tapes (A-red, B-yellow, Ogreen, D-blue; black on left,
white on right).
step 6: For each module (A,B,C,D) record 5 gauge readings.
4.2 Details of Sample Change
This section provides more details on the five steps in the sample change.
step 1: Remove the log sheet from last week's shipping box. The front
side should have been filled in the previous week. On the
reverse side, record the present date and time and your
initials. Record the current temperature, the minimum
temperature (since the last change) and the maximum temperature
on the min/max thermometer. Figure 4 shows how to read the
thermometer. Use the Celsius scale. Press the reset button
until the min/max indicators drop to the mercury. (The three
temperatures are needed for precise flow rate measurements.)
step 2: Open the control module. Verify the time of the time clock.
Turn the bypass timer to 30 minutes. This switch turns on the
four pumps, and disconnects the auto control of the solenoids.
Thus the solenoids will all be closed.
If the change takes longer than 30 minutes, you must reset the
switch. It is best to do this before the 30 minutes expires.
If the pumps turn off while there is a good vacuum behind the
solenoids and if the solenoids are closed, the pumps will
generally not restart. Therefore if the bypass timer turns off,
briefly open one toggle switch in each module to destroy the
vacuum, before restarting the timer.
-12-
-------
IMPROVE Sampler Manual
Version 2, 1988
WIN
MAX
read minimum
temperature at
bottom of blue
indicator
read maximum temperature
at bottom of
blue indicator
read current temperature
at top of mercury
(either side)
Figure 4. Reading the min/max thermometer. Note that the left side increases
as one reads down. Use bottom of blue indicators to read minimum (left side)
and maximum (right side) and top of mercury to read current temperature. When
finished recording values, hold down reset button until both blue indicators
drop to mercury.
step 3: For each module (Ar B, C, D) in turn, do the following. Layouts
of the fine and PM10 modules are shown in Figures 5 and 6.
a. Before pressing either filter switch, record the small
gauge. This should be 15 to 25 "Hg (depending on the
elevation) and indicates the pump is working properly.
If it drops below the indicated yellow line, look for
leaks in hose and hardware between the solenoid and
pump and call Davis.
b. Press the filter 1 toggle switch and record the small
gauge and the magnehelic gauge. The insert in Figure 5
shows how to read the small gauge below 5 "Hg. Section
4.3 discusses the significance of the gauge readings
and how to use them for troubleshooting.
c. Repeat substep b for filter 2.
d. Record both elapsed times (in hundredths of hours).
Times of 24.00 hours are expected. Zero both elapsed
timers. In extreme cold the reset buttons may not
operate easily. In this case it will be necessary to
keep a running total and determine the durations by
subtraction.
-13-
-------
IMPROVE Sampler Manual
Version 2, 1988
INLET
TACK
•bottom
cop for
stock
I r^
I 1
, j PUMP VAC.
I 1
O
Toggle
•0FILTER 2 \\
l'
Dwyer 2001 LT
Magnehellc n
0-I.O"H20 '
AIR FLOW
^
(-• -• i
ocuum line
to pump
•Cup for
coarse particles
ff
Figure 5. Layout of fine filter module
-14-
-------
IMPROVE Sampler Manual
Version 2, 1988
4 x- INLET STACK
"^V (HAS PM-IOHEAD)
CRfTfCAL
ORIFICE
Vocuum line
to pump
Figure 6. Layout of PM10 filter module
-15-
-------
IMPROVE Sampler Manual
Version 2, 1988
step 4: Remove each cassette by first unscrewing the hose from the
solenoid. Then unscrew the knurled knob holding the cassette
enough to lift the support bracket and remove the cassette. Do
not remove the knob completely. Put the red cap on the cassette
and place it in last week's shipping container. When finished
with module D, place the completed logsheet in also.
step 5: Remove the blank log sheet from this week's shipping box. Check
the date on the box and on the log sheet. If this is not the
proper date for the current week, make a large note on the log
sheet. (Also call Davis.) Fill in current date and time and
your initials. If this is change does not follow the removal of
the exposed cassettes, record the current temperature. Insert
the cassettes for each module in turn, selecting the two
cassettes for the module: red for A, yellow for B, green for C,
and blue for D.. Remove the red cap from the black (filter 1)
cassette, insert the cassette into the left side of the manifold
(under the clamp) and attach the hose and nut onto the left
(filter 1, black) solenoid. Place the red cap in the empty
shipping box. Repeat for the cassette marked with white tape
(filter 2) inserting on right side of manifold and right
solenoid. Hand tighten the nuts on the hoses. Do not use a
wrench or pliers to tighten. Make certain both cassettes are
firmly mounted on the manifold and tighten the knurled knob.
step 6: Record gauges for each module in turn, in the same way as with
the exposed filters:
a. Before pressing the toggle switches record the small
gauge.
b. Press the filter 1 toggle switch and record both
gauges.
c. Press the filter 2 toggle switch and record both
gauges.
d. Check that elapsed timers were reset. If the elapsed
timer will not reset because of cold, record the
initial time on the right side oth the log sheet.
You could turn off the bypass timer, although it will
automatically turn off after the 30 minutes have elapsed.
4.3 Trouble Shooting
This section discusses how to interpret the readings of the two gauges when
a solenoid is open. The gauges provide two independent measurements of the
flow rate. The equations are given in section 9. The magnehelic measures
the pressure drop across the cyclone or a large orifice; as the flow rate
decreases, the magnehelic reading will also decrease. The small gauge
gives the pressure drop across the filter; an increase in this pressure
drop (due to loading of the filter or a nonstandard filter) will cause the
flow rate to decrease.
-16-
-------
IMPROVE Sampler Manual
Version 2, 1988
If the needles of both gauges are on the green line, then the sampler is
operating at the standard flow rate. " If there are no leaks in the system
and the only problem is a nonstandard drop across the filter, the two
gauges will shift in the opposite direction. For example, a heavily loaded
filter would cause the small gauge to increase and the magnehelic to
decrease. As long as the needles are within the red lines, the flow rate
is acceptable (10%).
Several problems in the vacuum system can be quickly seen by comparing the
two gauge readings. Note that the fine modules (A, B, C) differ from
module D.
1. The cassette or cassette plug is not placed firmly on the cassette
manifold. (The plug is a problem for module B.) Modules ABC:
the small gauge will read in the acceptable range, but the
magnehelic will read low. In module D, the gauges will not
indicate when the cassette is not seated firmly, so special care
must be taken.
2. The. hose is not be connected properly to the solenoid. Modules
ABC: both the magnehelic and the small gauge for one filter will
read low. Module D: the magnehelic will be acceptable, but the
small gauge will read low.
3. Foreign matter is blocking the critical orifice. Modules ABC and
module D: both the magnehelic and the small gauge for both
filters will read low. Call Davis for instructions.
4. Module B only. The denuder has slipped and is partially blocking
the inlet. The magnehelic will read high. Call Davis for
instructions.
The following table summarizes the possible errors:
module vac gauge magnehelic possible problem
ABC normal low cassette or plug not seated
ABC low low hose not connected properly at solenoid
critical orifice clogged *
leak between solenoid and pump **
B normal high denuder in inlet has dropped
D low normal hose not connected properly at solenoid
D low low critical orifice clogged *
leak between solenoid and pump **
* should occur for both filters in module
** the vac gauge will also be lower than normal with both soleniods
closed—should occur for both filters in module
-------
IMPROVE Sampler Manual
Version 2, 1988
CONTROL MODULE
5.1 General Description
The control module consists of a 4-channel 7-day time clock, a 30-minute
bypass timer, five control relays, a fan with thermal switch set for 85°F,
a heater with thermal switch set for 20eF, three time delay relays, a 200
VA transformer and a terminal strip. The electrical circuit is discussed
in section 8. The time clock is discussed in section 5.
The time delay relays and thermal switches are located under the blue
panel. The time delay can be adjusted using the DIP switches. The thermal
switch can be adjusted by removing the small cap on the front of the switch
and rotating the set screw.
5.2 Operating the Time Clock
The time clock in the control module operates the pumps and the filter
solenoids. Each of the four channels is independently programmable on a
weekly cycle. The clock has 4 to 7 day battery backup? if this is
exceeded so that the clock is either blank or always reads 12:00 AM, the
time and memory will have to be reset. The clock has a specified operating
range of -4° to +1228F. Figure 7 shows the clock face and the program
buttons. These buttons are normally covered by a plate.
Auto Auto Aufo Auto
On | Off On I Oft On ) Off On I Off
w xir
Chan. 1 Chan.2 Chan.3 Chan. 4
10 10 10 10
NN \ 1 i
IMP
AM ^
PM Zl
*lh -
, no-no
' U U'U U
Mo Tu We Th Fr Sa Su
GRflSSLIN digi 128-45/4
Reset Override
Override
±lll I/O 1
o
I/O 2 |/o 3 i/o 4
Cancel Wrff« Read h+ fTlH*
Set time Ix . IMP h— m —
Figure 7. Time clock with cover plate removed, showing clock face and
program buttons. The status shown is Tuesday, 8:00 AM with all four
channels off.
-------
IMPROVE Sampler Manual
Version 2, 1988
The clock uses 12-hour notation, with the hour and minute indicated
directly and the AM/PM shown by a black rectangle at the left.
The day of the week is indicated by a black rectangle on the bottom.
The status of each of the four output channels is indicated by black
rectangles at the top, with I-in (on) and O-out (off). When a channel n is
on, the solenoid for filter n is open and a sample is being collected for
all four modules.
IPMr +lh and Ix are not used and should have no black rectangles. The four
screwqs at the top should all be set to auto.
To set or adjust the time of the clock, the "set time" button must be held
down while pressing one or more other buttons. To set the time, it is
necessary to difine: the day of the week, the hour (using the "h+" and
"h-" buttons) and the minute (using the "mf" and "m-"). Be sure the AM/PM
setting is correct. To adjust the minutes, hold down the "set time" button
and press the "rot" or "m-n button. To adjust for daylight savings time,
hold down the "set time" button and press the "h+" button in spring
(beginning daylight savings) and "h-" button in fall (end of daylight
savings).
5.3 Programming the Clock
Each program consists of the following three elements:
1. 1 to 7 days
2. time of day
3. on/off command for 1 to 4 channels (open/close solenoids 1 to 4)
To enter a program, enter the above 3 elements, and then press the "write"
button. Continue until all programs have been entered. If you hesitate
longer than 15 seconds between button-pushing, the clock reverts back to
the time output mode and the program is not entered. To turn a channel on,
press the I/O button once. To turn a channel off, press the I/O button
twice. Every command is registered immediately with a black rectangle or
as hour:minute.
For the standard IMPROVE network, four programs must be entered, as shown
in Figure 8. For a typical entry press the day, "h+" once, "nvf" once and
the appropriate channel I/O either once or twice. Note that channels 3 and
4 are unused. The order of programs does not matter. An additional
program can be entered at any time.
You may cancel a program completely or modify one or more elements. Press
the "read" button until the desired program is reached. To cancel
completely, press "cancel". To change one or more elements, press the
appropriate buttons and then press "write".
-19-
-------
IMPROVE Sampler Manual
Version 2, 1988
The override buttons at the top can be used to reverse the status of any
channel (on to off, off to on). This continues in effect until cancelled
by another override or by being reset at the next program time.
The override buttons may be used during installation when calibrating the
magnehelic. For example, pressing the "override" button for channel 1 will
turn on the pumps and open all solenoids number 1.
The override button may also be used if the programs are entered after a
desired change. Suppose a program was entered at 8:10 that was to have
activated channel 1 at 8:00. Since the clock checks only for programs for
the present minute, channel 1 will not turn on. By pressing the override
button it will then turn on. It will turn off at the programmed time.
— lx
Chan. 1 Chan.3 Chan. 3 Chan. 4
10 10 10 10
_ I
IMP
AM
PM
Mo Tu We Th Fr Scr 5u
Chan. I Chan.3 Chan.3 Chan. 4
10 10 10 10
\ \ II
IMP
AM
Mo Tu We Th Fr So Su
program 1: Wed, 12:00 AM, 1 on program 2: Thu, 12:00 AM, 1 off
Chan. I Chan.3 CSan.3 Chan. 4
' 0 10 |0 10
Mo Tu We Th Fr So Su
Chan. I Chan.3 Chan.3 Chan.4
10 10 10 10
NN n / /
IMP
AM ]
PM __
• HI —
^r-D'H ni\
'^•CAHJi'J
— lx
Mo Tu We Th Fr Sa Su
program 3: Sat, 12:00 AM, 2 on program 4: Sun, 12:00 AM, 2 off
Figure 8. Clock face for four standard programs
-20-
-------
IMPROVE Sampler Manual
Version 2, 1988
5.4 Electrical Circuitry
The electrical schematic for the control module is shown in Figure 9. The
four outputs from the time clock (TC-1 to TC-4) each go to a relay. One
output from each relay goes to a terminal which is then connected to a
solenoid and an elapsed time indicator in each filter module. A second
terminal from each relay goes to all of the pumps. Thus when any clock
output goes on then all pumps are turned on.
The 30-minute bypass timer goes to a fifth relay. When this goes on, all
clock outputs are disconnected so that all solenoids will be closed. A
second output from the bypass relay turns on all pumps.
Three of the four pumps have time delay relays (TDR) to produce a small
delay between startups to prevent overload at sites where power is
marginal.
The numbers in square indicate the terminal number. The colors refer to
the wires on the six wire cables between control and filter modules and
between control module and pump house.
-------
IMPROVE Sampler Manual
Version 2, 1988
to
filter
modules
to
relays
•
in pump
house
110 v power
r!T~rl
f '" " Grasslin dioi 128
„ thermal /f~\
11 switch (85°F) ^)
,, thermal >• A A A
11 switch (20°F) hYavter
llOv
i i
24 v
II 30min (r-^] manual
bypass timer V8y override
ip , ii TC~I M-"i
it ii I q /
ro.o xrX
v*r\ o -r f •» x~x
II (rR)
,,TC-3 /W\
UiV
l,TC-4 /fp\
i,(5rVi ^^^
• •CR-IO t prn nrnnnn
jj^n-ic [4.^ |j|ue
' 4-81
j | ^
't"U_fe.__, black
1 2:E1
co a , — , Pump
l|OR-a . |3l red *l
f n o TDR ; _ , .
i|CR-9 i-.-ori /i nranop *? fl white
n^^*'^ |r cnrl ... _, i =; 1 nrnnn *T
f*Q II i
ii^n-ii ,jo _nr| ..,,.._ 1 rl hinn *A
,,CR-I2
1 — '• TTI hlnrk
'
*o
*2 tlm«»H
f /circuits
filters
C 7 nf
T c or
modules
A,B,C*D
~_* ^HIK^BAV
Schematic for IMPROVE Controller
Figure 9. Electrical schematic for control module. TC-n is time clock
output for channel n, CR's are control relays, TDK's are time delay relays.
The other numbers refer to terminal strip. The colors refer to wire on
6-wire cable.
-22-
-------
IMPROVE Sampler Manual
Version 2, 1988
FILTER MOtWLE
6.1 General Description
Both fine and PMlO modules contain identical control panels with two
elapsed time indicators, two toggle switches, a 0-30" Hg vacuum gauge and a
0-1" water magnehelic gauge. They both also contain 2 solenoid valves and
a critical orifice. The layout of the fine filter module is shown in
Figure 5, and that of the PMlO module in Figure 6.
The airstream enters the fine module through an inlet designed to keep out
rain, insects and particles larger than around 15 (M. The airstream than
passes through a 3.66 cm metal cyclone that is identical internally to the
well-tested cyclone design of Walter John and Georg Reischl of the Air
Industrial Hygiene Laboratory used by the California Air Resources Board.
It has a 50% efficiency effective cutoff at 2.5 /um aerodynamic diameter at
a flow rate of 21.7 1/min. The variation of 50% cutpoint versus flow rate
is shown in Figure 10. A 10% decrease in flow rate from 21.7 1/min will
produce a 10% increase in cutpoint diameter. For example, a 30% decrease
in flow rate to 15.3 1/min will increase the cutpoint to 3.5 tm.
S 3-5
o
u
•H
e 3.0
I '•'
o
2.0
\
\
10 15 20
Flow Rate (Ilters/min)
25
30
Figure 10. Diameter of 50% cutpoint vs. flow rate for cyclone, from
W. John and G. Reischl.
-23-
-------
IMPROVE Sampler Manual
Version 2, 1988
The PM10 module contains a standard PMlO inlet designed and built by
Wedding and Associates, that has a 50% efficiciency of 10 tm at a flow rate
of 18.9 1/min. The variation in outpoint diameter with a change in flow
rate from 18.9 1/min is not known. The PMlO inlet is attached to a filter
cassette manifold identical to that used in the fine module, except it is
mounted facing down.
The flow rate is regulated by a critical orifice located downstream of the
solenoids. The flow rate depends on the area of the critical orifice and
on the pressure and absolute temperature of the air as it enters the
orifice. .At flow rates near 20 1/min, the pressure drop across the filter
can be a significant fraction of the ambient pressure, which can procude
small changes in the flow rate. This affects the system in two ways.
First, since the different filter media have different pressure drops, it
is necessary to use critical orifices of different diameter in each of the
four modules. It is not possible to interchange filter media from one
module to another and maintain the desired flow rate. Second, the flow
rate can decrease when the filters become heavily loaded. Based on data
from the SFU network this should rarely be a problem.
The flow rate for a given sample is measured before and after each sampling
period using two methods, as described in the next section. The effect of
deviations in flow rate from the nominal value of 21.7 or 18 1/min is to
change the cutpoint from 2.3 or 10 im rather than produce errors in volume.
6.2 Flow Bate Measurement
The flow rate is measured in two ways as a weekly quality assurance check.
The first measurement is based on the equation for flow through a critical
orifice, and the second on the pressure drop across a fixed orifice. Both
equations depend on pressure and absolute temperature. The pressure is
assumed to depend only on the altitude. (Meteorological variations in
pressure are much smaller than those due to altitude; the average
variation is less than 1%.) The ambient temperature is measured at the time
of the weekly flow rate measurements. In addition, the minimum and maximum
temperatures for each week are measured, allowing a slightly more accurate
determination of average flow rate over the 24 hour sampling .period.
The flow rate through the critical orifice depends on the temperature of
the air and the pressure drop in the filter and in other parts of the
system up to the critical orifice. When the solenoid is open, the value on
the vacuum gauge, AP, (in "Hg) measures this pressure drop. (For the PMlO
module there is a relative large calibration orifice between the
measurement point and the critical orifice, but this decreases the pressure
by around 0.2%.) If T is the absolute temperature (°C +273) and P is the
ambient pressure in "Hg, then the flow rate by this first method is
Q, - Q0 (1- ^f)
where Q9 is the flow rate for no filter, corrected to 280°K, and depends
only on the area of the critical orifice. The temperature of 280°K
corresponds to 458F, a typical mean temperature for the network. The
temperature term is generally very close to unity. (It differs from 1.00
by less than 5% for temperatures between -5%"F to 96BF.) the critical
-------
IMPROVE Sampler Manual
Version 2, 1988
orifice is chosen to give a flow rate of 21.7 1/min (fine modules) or 19.8
1/min (PM10) for a typical filter. The constant Q0 is measured during the
system calibration at the site.
The constant Q. is designed to give the appropriate flow rate for a given
filter type. The desired value depends solely on the average temperature
of the air at the critical orifice and is independent of the altitude of
the site. If primes indicate the values at another site (such as at
Davis), then
Q0 - V (r) ,
In order to account for variations from the ideal equation, the equation
for Qx can be written in the more general form
where at and b are determined by varying fiP in the calibration protocol
and measuring Q1 •
The second measurement of the flow rate uses a magnehelic gauge to measure
the pressure drop across a fixed orifice that is large compared to the
critical orifice. For the fine modules, the fixed orifice is provided by
the cyclone, while for the PM10 module, it is provided by an orifice built
into the system just before the critical orifice. If 5P is the pressure
drop across the orifice and the pressure at the front of the orifice is at
atmosphere then the flow rate is
Q2- C(5P)b (|)4
where C is a constant depending on the geometry of the orifice and b is 4
if the flow is laminar. Note that the pressure drop across this orifice of
SP .is less than 1% of the AP across the filter. Assuming the sampler is to
remain at the same altitude, the equation can be rewritten in terms of two
parameters
Again the parameters a2 and b2 are measured during the system calibration.
The value of a2 is around 40 and b2 slightly less than 0.5.
For the PM10 orifice, the equations are slightly different, because the
pressure at the front of the measuring orifice is reduced by the drop
across the filter, so
Q2-C(«P)b (^(l-^)4
Eliminating AP, the pressure drop across the measuring orifice can be
written in a parameterized form as
-------
IMPROVE Sampler Manual
Version 2, 1988
The value of a2 is around 40 and b2 slightly less than 1.
The average flow rate for the sampling period will be a weighted average of
Qx and Q for the initial and final readings. The flow rate will be
corrected for changes in temperature using the mean temperature for the
week, calculated from the minimum and maximum.
6.3 Flow Rate Audit
The flow rate audit is a check of the flow rate using an audit device
placed between a filter cassette and a solenoid, for each of the four
filter modules. The audit may be conducted by a third party auditor, using
any flow rate meter that has a low pressure drop, by field personnel using
an orifice meter supplied by Davis and mailed to and from the site, or by
Davis personnel using either an orifice meter or a dry test meter. A mass
flow meter is an acceptable device. The orifice meter will be calibrated
at Davis using a spirometer and dry test meter. The system magnehelic and
vacuum gauge will be read simultaneously in order to compare the flow rate
measurements. There are two purposes to the audit: (1) to compare the
flow rate measured by the module with the audit flow rate, in order to
determine the accuracy in the volume of air collected, and (2) to compare
the audit flow rate with the nominal flow rate (21.7 or 18.9 1/min), in
order to determine the accuracy in the particle sizing of the cyclone or
PH10 inlet. For mail audits the actual flow rates will be calculated at
Davis using the information on the logsheet.
The equipment needed for the audit will be a flow rate meter with 1/2 inch
standard compression fittings at either end (male fitting at upstream end
and female fitting at downstream end), and four filter cassettes with hoses
containing standard filters. If the standard filter cassettes are not
available, it is possible to use the filters at the site provided for the
normal sampling. A logsheet will be provided for the audit, as shown in
Figure 11. Record the information at the top of the sheet.
The following procedure is to be used for the audit. First, using the
override button for channel 1 on the time clock (see section 5.3), turn on
all the pumps and open all first solenoids.
For each of the four modules do the following steps.
— Attach the audit device to solenoid 1 of module.
— Attach the filter cassette with the appropriate filter for that module to
other side of the audit device.
— Record the reading for the audit device and for both module gauges.
Any audit device added to the system will decrease the pressure at the
front of the critical orifice and thus decrease the flow rate. It is
important that the pressure drop produced by this device is small enough to
-26-
-------
IMPROVE Sampler Manual
Version 2, 1988
produce no significant effect. Suppose the pressure drop for the total
system up to the critical orifice without the device is AP and the device
produces an additional SP. The equation for the fractional decrease in
flow rate is
AQ (P-AP) - (P-AP-&P) m , &P
Q (P-AP) P-6P
That is, the drop produced by the device must be small compared to the
pressure at the front of the orifice. The worse case (smallest P-AP) would
occur at high altitudes and large pressure drop across the filter. For the
network this would be with P-AP approximately 15 "Hg - 200 "H20. A
pressure drop of 2 "H,0 would cause a 1% decrease. The standard
calibration devices have a pressure drop of 0.5 "H20 at 22 1/min, causing
an error of 0.25%.
1HPROVI I A H F L t * AUDIT LOO
ntt mm
BVTB
SAMPLER SERIAL NUMBER
(UDITD) BT
•nnPCRATORB C + Z73 • K
M.TITODB
ttJDIT NETCT NUMBER
ttJDIT coerrst •
b -
M.TITUDB f*CTOR (A)
DEHSITI FACTDRl (T/398P * A
flow rate at «ite - deiutty (actor * 10s(«plb
racxjuf A
s-Xvic.
put? v
•sonehellc
"•HjO
VCUUB
small
9*sr
flow tat* by audit devic* -
flow r«t» by tyiten na^nehelle *
flow tat* by trail gauqw •
HDDUUE C
devic*
pump v»euu» _
gyste*
•agnehalie
*v
imall
"ST
flow cat* by audit d*vle* -
flow tat* by cysten ragnctolie -
flow tat* by email 98119* •
HCCULK B pmv vacuum
audit system inwll
d«vlc* nBTJiehellc ^ 9fuj»
H*0 H^
flow rat* by audit devic* -
flow rat* by *y*teit magnehellc -
flow rat* by caall gauge •
HOCULS D pimp vacuui
audl t system small
device BemeHellc 9»uoe
"VH20 EH|
flow rat* by audit devic* •
flow rat* by tyiten BBgnehellc •
flow rat* by mall 9*09* •
Figure 11. Flow rate audit log.
-27-
-------
IMPROVE Sampler Manual
Version 2, 1988
6.4 Flow Rate Calibration
The flow rate calibration is similar to the flow rate audit, except that
the flow rate is adjusted to several values using a valve to imitate a
variable filter. The calibration will be performed by Davis personnel and
includes calculation of the flow rates on site. The only additional
equipment is the valve assembly. The calibration logsheet is shown in
Figure 12.
At the top of the calibration logsheet, record the site name, sampler
serial number (located inside door of controller module), date,
temperature, your name, altitude, and altitude (pressure) factor relative
to Davis. These last two numbers should be read from Table 2 or Table 3.
Using the temperature (°K) and the altitude factor, calculate the air
density factor relative to Davis.
C - f)1* * (altitude factor)
and record on the sheet. This factor is to be used in several places to
convert flow rates from Davis values to site values. Record the
calibration coefficients for the meter on the sheet. The coefficients
relate the flow rate at the site and the measured pressure from the
calibration magnehelic:
Q - C 10a (Sp)b
where C is the density factor.
Using the override button for channel 1 on the time clock (see section
5.3), turn on all the pumps and open all first solenoids.
Using the coefficients for the calibration orifice/magnehelic calculate the
APc's for the pre-set flow rates on the log using the relationship
SPC -
Record the desired meter settings on the calibration log sheet.
-28-
-------
IMPROVE Sampler Manual
Version 2, 1988
IMPROVE SAMPLER CALIBRATION LOG
SITE NAME
DATE
SAMPLER SERIAL NUMBER
CALIBRATED BY
TEMPERATURE
ALTITUDE
C + 273 -
CALIB METER NUMBER
CALIB COEFFS! a -
ALTITUDE FACTOR (A)
b -
DENSITY FACTOR: (T/296P * A
flow rate at site - density factor * 10a(«p)b
MODULE A pump vacuum
flow rate
1/mln
max
21.7
19.5
16.3
calib
magnehelic
H2°
system
magnehelic
H-0
small
gauge
*Hg
magnehelic (log fit) r -
log flow- 4 *log(8P)
small gauge (linear fit) r -
flow- - *(AP)
MODULE C pump vacuum
flow rate
1/min
max
21.7
19.5
16.3
calib
magnehelic
"H2°
system
magnehelic
H2°
small
^^
magnehelic (log fit) r -
log flow- + *log(SP)
small gauge (linear fit) r -
flow- - MAP)
max - readings with no filter
MODULE B pump vacuum
flow rate
1/mln
max
21.7
19.5
16.3
callb
magnehelic
system
magnehelic
H2°
small
V9
magnehelic (log fit) r -
log flow- + *log(«P)
small gauge (linear fit) r -
flow* - MAP)
MODULE D pump vacuum
flow rate
1/min
max
18.9
17.0
14.2
calib
magnehelic
H2°
system
magnehelic
H2°
small
gauge
*Hg
magnehelic (log fit) r -
log flow- + *log(8P)
small gauge (linear fit) r -
flow- - MAP)
Figure 12. Flow rate calibration log.
-29-
-------
IMPROVE Sampler Manual
Version 2, 1988
Table 2. Density factors at IMPROVE sites
(factor relative to Davis)
factor - I P(Davis) / P(site) ] *
site
alt P factor
feet "Hg
site
alt P factor
feet "Hg
Acadia
Arches
Big Bend
Bridger
Bryce Canyon
Canyonlands
Chiricahua
Crater Lake
Denali
Glacier
Grand Canyon
Great Smoky
470
5500
3460
8000
7950
5950
5400
6479
2100
3200
7100
2700
29.4
24.4
26.4
22.2
22.2
24.0
24.4
23.5
27.7
26.6
22.9
27.1
1.007
1.106
1.063
1.159
1.159
1.116
1.104
1.126
1.037
1.059
1.140
1.049
Guadalupe Mtns
Jarbidge
Mesa Verde
Mount Rainier
Petrified For
Pinnacles
Rocky Mountain
San Gorgonio
Shenandoah
Weminuche
Yellowstone
Yosenrite
5446
6200
7200
1400
5500
1040
7910
5618
3515
9140
7750
5250
24.4
23.7
22.8
28.4
24.4
28.8
22.3
24.3
26.3
21.2
22.4
24.6
1.105
1.121
1.142
1.025
1.106
1.018
1.157
1.108
1.065
1.186
1.154
1.101
Table 3. Density factor vs. altitude
(factor relative to Davis)
factor - { P(Davis) / P(site) ]
alt P factor
feet "Hg
0
200
600
800
1000
1200
1400
1600
1800
2000
2200
2400
2600
2800
3000
3200
3400
29.9
29.7
29.2
29.0
28.8
28.6
28.4
28.2
28.0
27.8
27.6
27.4
27.2
27.0
26.8
26.6
26.4
0.998
1.002
1.010
1.013
1.017
1.021
1.025
1.028
1.032
1.036
1.039
1.043
1.047
1.051
1.055
1.059
1.063
alt
feet
3600
3800
4000
4200
4400
4600
4800
5000
5200
5400
5600
5800
6000
6200
6400
6600
6800
P
"Hg
26.2
26.0
25.8
25.6
25.4
25.2
25.0
24.8
24.6
24.4
24.3
24.1
23.9
23.7
23.5
23.4
23.2
factor
1.067
1.071
1.075
1.079
1.083
1.088
1.092
1.096
1.100
1.104
1.108
1.113
1.117
1.121
1.125
1.130
1.134
alt P factor
feet "Hg
7000
7200
7400
7600
7800
8000
8200
8400
8600
8800
9000
9200
9400
9600
9800
23.0
22.8
22.7
22.5
22.4
22.2
22.0
21.8
21.7
21.5
21.3
21.1
21.0
20.8
20.7
1.138
1.142
1.146
1.151
1.155
1.159
1.164
1.168
1.173
1.178
1.183
1.187
1.192
1.197
1.201
-------
IMPROVE Sampler Manual
Version 2, 1988
For each of the three fine modules do the following steps.
—Attach the calibration meter to solenoid 1 of module.
—Attach the filter cassette with the appropriate filter for that module to
other side of the calibration meter, in the same manner as for the audit.
—Calculate the flow rate using the calibration magnehelic using the
relationship for the calibration magnehelic system
Q - C 10a («p)b
—If the flow differs from 21.7 or 18.9 by more than a few percent, replace
the critical orifice until a better value is obtained.
—Replace the filter cassette with the valve assembly (without filter)
designed to act as a variable filter.
—For each of the four points adjust the valve to give the desired SPc on
the calibration magnehelic and record both module gauges. If desired do
other points.
—For modules ABC, perform a linear regression fit for the magnehelic using
the log relationship
log A - a2 + b2 log (SP).
—For module D, perform a regression fit for the magnehelic gauge using the
linear relationship
Q - a2 + b2 5P
Check that r is greater than 0.990. Record the coefficients. For
modules ABC, b2 is slightly less than 0.5 and a is around 1.5. For
module D, the coefficient a2 should be much smaller than b2.
—Perform a regression fit for the vacuum gauge using the relationship
Q - at + bx AP
Check that -r is greater than 0.990. Record the coefficients. The
coef f icient "bj will be negative, and a: is much larger than bt.
—For modules ABC, mark the 21.7 readings on both gauges in green ink and
the 23.9 and 19.5 readings on both gauges in red ink. For module D, mark
the respective readings are 18.9, 20.8 and 17.0.
-31-
-------
IMPROVE Sampler Manual
Version 2, 1988
6.5 Electrical Circuitry
The electrical schematic for a filter module is shown in Figure 13. The
connection with the control module is through a six wire cable. The
solenoid valve (SV) and elapsed time indicator (ETI) will be activated if
either the appropriate clock output is on or if the appropriate toggle
switch is turned on. When the toggle switch is released if will always
return to the normally open position, giving control to the time clock.
I I Controller terminol
/\ Filter module terminal
TS Toggle switch
Figure 13. Electrical schematic for filter module.
ETI - elasped time indicator, SV - solenoid valves.
the wire on the 6-wire cable.
TS - toggle switch,
The colors refer to
-32-
-------
IMPROVE Sampler Manual
Version 2, 1988
6.6 Nitrate Denuder
The nitrate denuder is installed in the inlet tube of module B. As shown
in Figure 14, it consists of a series of four concentric aluminum cylinders
coated with NaC0. The total surface area is 0.3m3. The nitrate gases
(N02 , N03 ) will diffuse to a wall and be captured, while the particles pass
,3.
ll diff
through, because of the much smaller diffusion rate for particles. At -a
regular interval (perhaps a year) the entire cylinder must be replaced and
returned to Davis for recoating. Normally this change will be performed by
Davis personnel.
The denuder will fit inside the inlet tube. The operator need only open
the cap at the base of the inlet, pull out the old denuder," replace it with
a new denuder and recap the inlet. More detailed procedures will be
available when the system is developed.
To replace the denuder, it is necessary to remove the cap at the base of
the inlet and remove the screw-pin holding the denuder in the inlet. A
handle will be screwed into the base of the denuder (from below) and the
denuder pulled down. The new denuder will be pushed up using the same
handle and the screw-pin inserted. The handle will then be removed and the
cap replaced. The spent denuder will be recharged at Davis.
Tutinj f
alum!
M.l«: Inner
fbread
Figure 14. Nitrate denuder.
-33-
-------
-------
IMPROVE Sampler Manual
Version 2, 1988
7. PUMP HOUSE
7.1 General Description
The pump house is an all-weather house measuring 3'x2'x4'. Inside there
are four pumps with surge tanks, a box of 4 switched 120V outlets, a box of
4 unswitched 120V outlets, a fan with thermal switch, a lamp socket with
manual switch and heat tape with thermal switch (where necessary). The
wall of the pump house is shown in Figure 15. Attached to the outside of
the house is a weather-tight electrical box contain four power relays.
The external 120V line is attached to the terminal strip inside the relay
box. 120V lines are then carried to the control module, to the unswitched
outlet box and through the power relays to the switched outlet box. A
6-wire cable from the control module connects to the control terminals of
the power relays.
7.2 Pump Specifications
The pump is a 0.5 HP oilless diaphragm vacuum pump, Cast model DAA-V132-GB.
The pump can draw 38 1/min at 15 "Hg and 17 1/min at 20 "Hg.
unswitched
120V outlet
switched
120V outlet
O pass-through
to relay box
Figure 15. Layout of wall of pump house.
-34-
-------
IMPROVE Sampler Manual
Version 2, 1988
7.3 Electrical Circuitry
The electrical schematics for the pump house are shown in Figure 16. The
output from the control module activates a 20A power relay which provides
120V to a switch outlet used by one of the pumps. For safety, both the hot
and neutral 120V lines are switched. In addition to the four switched
outlets, the pump house has four unswitched outlets (not shown).
CONTROL
CIRCUIT
red
orange
green ^^
blue
61 U
Controller lerminals
white
POWER
CIRCUIT
no v
[[thermal switch *Tfo"n)
i, thermal switch he?te/
II
umanual
n
II CR"A
U
i, CR-B
II
1 1 CR ~C
II
" V V V ~
switch >~--X
.CR-AH
ii
,iCR-Bz
II '
pCR-Cz
II '
,,CR-C3
II
Figure 16. Electrical schematic for pump house. CR - control relays
-------
IMPROVE Sampler Manual
Version 2, 1988
The relays are located in a weathertight box attached to the outside of the
pump house. Figure 17 shows a schematic for the relay box. The power
cable enters the box from the bottom and is attached to the terminal strip
as shown.
PUMP OUTLETS
(inside of pump house)
POWER
RELAYS IN CONTACTOR BOX
Figure 17. Electrical schematic for relay box
-36-
-------
IMPROVE Sampler Manual
Version 2, 1988
8. SAMPLER STAND
The five fiberglass enclosures must be mounted on a stand. Because of the
difficulty in transportation, and because individual sites may have special
needs, the stand is generally built locally. At some sites the sampler is
located outdoors, so that a freestanding wooden stand is appropriate. At
some sites where an air quality building already exists, the stand consists
of support boards attached to the walls of the building.
The stand shown in Figure 18, often built of redwood, is used at most sites
that do not have an air quality building. To insure the stability of the
unit, the legs are fastened to concrete piers and imbedded in the ground.
In situations of uneven ground, it is desireable to have legs of different
length, in order to keep the horizontal beams level. Holes with 12 inch
spacing must be drilled in the horizontal beams in order to mount the
enclosures.
The stand is also to have sunshields above the enclosures. A short shield,
2 feet long, goes above the control module. A long shield, 104 1/2 inches
long, goes above the four filter modules, with holes cut for the inlets.
The shields are attached to the stand using angle brackets. It is
important to orient the stand so that the fronts of the enclosures do not
face the afternoon sun (west).
If the sampler is mounted inside a shelter, it is usually necessary to use
longer inlet stacks. The four holes in the support board for each module
must be 12 inches apart horizontally and 18.75 inches apart vertically.
Adjacent modules should not have the mounting holes closer than 8 inches
apart.
-37-
-------
IMPROVE Sampler Manual
Version 2, 1988
Provide ? iddttlonil redwood el«nkt to t>* ui«i)
M sunshlotdt (to be «tt«ched by br*ek«tfl)t
I" « 17" « J4" «nd 1" > 17" i 104V
END VIEW
Figure 18. Outdoor stand for IMPROVE sampler
-38-
-------
IMPROVE Sampler Manual
Version 2, 1988
9. Acknowledgments
The following members of the Air Quality Group or Crocker Nuclear
Laboratory were responsible for the development of the IMPROVE sampler.
Sampler design and evaluation; C. Goodart, R.A. Eldred, T.A. Cahill,
P.J. Feeney, L.K. Wilkinson, P. Beveridge, T. Essert (electrical), 0. Raabe
(flow rate, denuder), S. Teague (flow rate, denuder), B. Pedersen
(denuder), K. Bowers (S02 collection)
Sampler construction and installation: R.A. Eldred, C. Goodart,
P. Beveridge, I. Wittmeyer, P. Wakabayashi, D. Everitt, J. Cordova,
B. Matsumura, S. Eldred, P. Dyer, M. VandeWater, E. Steen,
D. Cecil (purchasing).
Special field studies (WHITEX, SCAQS): L.K. Wilkinson, I. Wittmeyer,
P. Beveridge, M. Surovik, P. Wakabayashi, D. Everitt, J. Cordova,
B. Matsumura, H. Miyake, B. Perley, S. Eldred, T.A. Cahill, R.A. Eldred,
T. Tanada, D. Orr, C. Semantis, J. Cooper, E. Steen, H. Miyake,
K. Mitchell, 0. Beckmann, F. terVeer, W. Reeves, B. Nicolet, K. Bowers,
C. Cahill
Documentation: R.A. Eldred, J. Hancock, C. Goodart, C. Baro.
The controller modules were fabricated by ENDECO Controls of Rio Linda.
Most of the machine work was done by Streetman Precision of Cameron Park
and G's Machine of Placervilie.
We would like to thank the UC Davis purchasing department for their
considerable help in reducing costs, especially Mari Harrod.
We would like to thank Walter John of Air Industrial Hygiene laboratory for
the plans for the cyclone.
We would like to thank Marc Pitchford, EPA-Las Vegas and a member of the
IMPROVE committee for the many helpful discussions concerning design.
-------
IMPROVE SOP APPEND JULY 1989
APPENDIX 3: Gravimetric Mass Startup Procedures
A. Cleaning and Calibration of the Electrobalance
B. IMPROVE Gravimetric Controls
-------
IMPROVE SOP APPEND JULY 1989
APPENDIX 3: Gravimetric Mass Startup Procedures
A. CLEANING AND CALIBRATION OF THE ELECTROBALANCE
The first work action at the beginning of each day and immediately after lunch
period is to clean and calibrate the electrobalance. It will also be
recalibrated if the balance fails a "zero" test that is performed
periodically. The procedures described here are outlined in the Cahn 31
Electrobalance Instruction Manual in Section 4, Operations.
A Balance Log is maintained for each balance. All of the calibration results
are to be recorded in the Balance Log. In addition, all significant events
concerning the balance and any balance maintenance other than routine are to
be recorded in the Balance Log.
The steps for cleaning and calibration are as follows.
1. Clean the metal and plastic forceps with ethanol and a Kimwipe.
2. Carefully remove each balance pan by inserting closed forceps below
the wire yoke of the pan bail and lifting the hook from the eye. To
protect the hangdown and hangdown loop, do not grasp the bail or hook
assembly with the forceps. Gently rest the pans on a. fresh clean
Kimwipe.
3. Remove the antistatic ionizing strip from the weighing cavity.
A. Clean and deionize the inside of the balance cavity using a
cotton-tipped applicator wetted with ammonia. Gently brush the
exposed surfaces of the cavity, taking care not to disturb the
hangdowns extending from the top surface. Gently close the glass
slide door and gently brush the outer surface with the ammonia tip to
control static change.
5. Clean the top surfaces of the balance pans with a cotton-tipped
applicator wetted with ethanol. The entire surface lust be clean and
dry. Do not use any solvent other than ethanol on the pans.
6. Clean the top surface and the strips of the antistatic ionizing units
by gently rubbing with a cotton-tipped applicator wetted with
ethanol.
7. Replace the clean ionizing unit in the center back of the balance
cavity.
8. Gently return the balance pans to their hang down loops. Use the
bail- lifting procedure previously described. Do not put any stress
on the bail or hang down loop.
-------
IMPROVE SOP APPEND JULY 1989
9. Record the time and the balance temperature from the top surface
thermometer. Read the mass when there is nothing in the pan; this
is the "zero" mass. It should be within 0.010 mg of 0.000. This
value may be exceeded if the ethanol has not completely evaporated;
in this case, the reading will drift toward 0 as the ethanol
vaporizes. Record the "zero" mass.
10. Depress the tare button. This forces the "zero" mass to be exactly
0.000. Record this 0.000 value.
11. Momentarily ground yourself by touching the electrostatic mat. Use
plastic forceps to remove the 200.000 mg class 1.1 (Class M)
calibration mass from its container. Gently place it in the center
of the balance pan and allow the mass reading to stabilize and stop
decreasing. This will require from 30 to 60 seconds depending on how
smoothly the mass was released onto the pan. When the reading
stabilizes, record the mass. Next, press the calibrate button,
forcing the balance to indicate 200.000 mg. Record this value.
Gently remove the calibration mass and replace it in its container.
12. Remove the standard 50.000 mg mass from its container using plastic
forceps. Gently place it in the center of the balance pan. After
the reading stabilizes, record the mass in the Balance Log. Record
your initials to indicate your calibration.
13. Allow the balance to return to "zero". Compare the zero value and
the value determined for the 50.000 mg mass to previous values. If
they exceed ±2 micrograms, repeat the procedure. If greater
variations are observed, see the laboratory supervisor.
14. After cleaning and calibration, the electrobalance is available for
routine determination of mass.
15. Clean the work surface of the antistatic mat with a clean brush or a
Kimwipe dampened with water. Do not use ethanol as it will damage
the mat surface.
16. Clean the metal forceps and the antistatic ionizing strips by wiping
them with ethanol and a clean Kimwipe.
17. The next step is to process the control filters, following the
procedures described under IMPROVE Gravimetric Controls.
18. On a random basis, but at least semiannually, the laboratory
supervisor will request a comparison of the normal calibration
standards with a second set of reference standard masses maintained
by the laboratory supervisor. After calibration, measure these
200.000, 50.000, and 20.000 mg standards and report their masses to
the supervisor. These results are used to verify the integrity of
the electrobalance and the standard masses used in daily
calibrations.
-------
IMPROVE SOP APPEND JULY 1989
B. IMPROVE GRAVIMETRIC CONTROLS
The gravimetric control program is used to determine the precision of the
gravimetric analysis and the mass artifact associated with the storage of
teflon filters in cassettes.
*
The procedure for a control filter is .to measure its mass twice, allow it to
remain in an IMPROVE cassette for approximately 35 days, and then remeasure
the mass twice. The filter is then put in a permanent slide mount and
archived in a slide tray. A computer program determines the precision of two
pairs of mass measurements and calculates the mass artifact associated with
the cassette by subtracting the mean initial mass from the mean final mass.
The procedures for the beginning of the day, immediately after cleaning and
calibrating the balance, are as follows.
1. Obtain a clean 25 mm teflon filter from the prepared stock maintained
for SFU samplers. This filter is in a petri dish identified with an
additional "S". This filter is known as the "PRE" filter.
2. Select the next premade Control Identification Tag kept at the front
of the Balance Log (e.g. 1-465) and attach it to the petri dish.
3. Locate the oldest IMPROVE Control cassette, indicated by the lowest
control number and place it in the electrobalance work area. This
filter is known as the "POST" filter.
4. Download the POST cassette and place both the PRE and POST filters on
antistatic strips.
5. Prepare an IMPROVE Control tag for recording all the values required
by the Control database. This tag should have three columns and have
the following form.
Pre "new" control ID RePre
Post "old" control ID RePost
Date Technician entered
6. Measure the PRE filter, record its mass, and place the filter into a
clean petri dish. Affix the identification tag for this filter on
the petri dish.
7. Measure the POST filter, record its mass, and place the filter into a
clean petri dish. Remove the identification tag from the cassette
and affix it on the petri dish.
8. Store both petri dishes, the control tag, and the empty cassette
until required for the afternoon measurements.
-------
IMPROVE SOP APPEND JULY 1989
The procedures for the beginning of the afternoon, immediately after cleaning
and calibrating the electrobalance, are as follows.
1. Remove the PRE and POST filters from their petri dishes and place
them on antistatic strips.
2. Prepare an 18x24 mm slide mount to receive the POST filter. Write
the following information on the white side of the mount using a
felt-tip pen.
a. IMPROVE control number (e.g. 1-430),
b. The corresponding S number (e.g. S-1391),
c. Today's date (e.g. 5/17/89),
d. The number of the next empty position in the 40-position archive
slide tray (e.g. 13).
3. Remeasure the PRE filter, record its mass (REPRE), and mount the PRE
filter in the empty IMPROVE control cassette, following the standard
protocol.
n
Center the drain disk and the 2.2 cm^ mask. Place the filter on the
disk and mask with the smooth side up. Replace the lock ring and
cassette cup and protective red cover. Affix the PRE identification
tags to the cassette and place it at the end of the queue of IMPROVE
control cassettes.
4. Measure mass of the POST filter and record its value (REPOST). Mount
the filter on the slide mount with the smooth side facing the black
half of the mount. Place the mount in the next empty slide tray
position. Replace the tray cover, and return the tray to its storage
place.
5. Input all the control data into the database using the program
IMPCONT. Enter the morning masses, PRE and POST, followed by the
afternoon masses, REPRE and REPOST. The program calculates and
presents the precision of both pairs and the 35-day mass gain.
6. Precision values should not exceed ±5 micrograms. If they do, select
the no-entry function of IMPCONT and correct the gravimetric problem.
Precision of ±5 to 10 micrograms generally indicates a calibration
problem. Greater values indicate incorrect identification problems.
7. When verified, input the data into the database permanently by
selecting the "yes" input. This completes the daily control actions.
8. Every month, the laboratory supervisor will activate a feature of
IMPCONT that calculates and summarizes all precision and mass
artifact values for the selected month. These results are appended
to the data for previous months and printed out in hard copy.
-------
IMPROVE SOP APPEND JULY 1989
APPENDIX 4: LIPM Startup Procedures
CALIBRATION OF THE LASER INTEGRATING PLATE METHOD (LIPM) SYSTEM
1. Do not begin the calibration unless the LIPM system has been on for
at least two hours, allowing the laser to stabilize. The LIPM system
is programmed to be on daily between 0600 and 1800. Verify that the
timer is indicating the correct time. If necessary, set the timer to
the correct time by rotating ahead clockwise.
2. Routine operation of the system requires no adjustments except a
vernier multiplier setting. Verify that the Oriel Detection System
Model 7022 has the following startup settings:
a. Multiplier is set on the xlO~" level
b. The 45 volt switch is ON
c. Ambient supress is OFF
d. Response is set to MEDIUM
e. Zero Set is OFF
f. Verify that the collimator aperture is installed.
3. Blank off the input laser light by moving the Beam Attenuation lever
to "closed." Verify that the detector digital readout is 0.000
(zero). If not, adjust the Zero Set knob until the detector
indicates 0.000.
4. Open the Beam Attenuation lever and push the slide changer in all the
way. Make sure the tray-covering flap is closed. The detector
should indicate 0.000. If the measurement exceeds +0.005, check for
an open cover flap or for a physical malfunction of the changer.
Notify the laboratory supervisor.
5. Pull the slide changer all the way out. Use the vernier multiplier
extension arm to set the detector reading to the protocol calibration
level of 0.750.
6. Mount the LIPM standards tray of 10 standards in the system. These
standards test all parameters of slide orientation and include a
spectrum of actual filters and aerosols.
7. Record the date, your initials, and the values obtained for each of
the standards slides on the LIPM calibration logsheet in the LIPM
protocol folder. Record the values for standards 1 thru 10.
Periodically verify between slides that the detector reference level
remains at 0.750 with the slide changer fully out.
8. Compare the values obtained with previous values reported in the
logbook and the highlighted standard values. If values differ by
more than ±0.003, repeat the procedure. If deviations still exist,
consult with the laboratory supervisor.
9. The LIPM system is now ready for routine analysis of filters.
-------
IMPROVE SOP APPEND JULY 1989
APPENDIX 5: PIXE/PESA Procedures
PIXE RUN DIRECTIONS
1. CONTROL ROOM OPERATIONS
a. Before going into CAVE (close target)
— Hit 'CLOSE' button on the lower left hand side of control panel.
— Hold down 'TARGET' button (in the middle of the control panel)
until 'CLOSE' button turns amber. (Takes 5 seconds)
— Hit 'CLOSE' Button to Cave (on top right hand side of control
panel)
— Remove key labeled North Cave Door No. 3.
— Close pumps (Move toggle from HIVAC to middle position)
b. After coming out of CAVE (open target)
— TURN ON roughing pump (toggle to rough). When vacuum reaches 80
microns, turn toggle to HIVAC.
— Put back key
— Hold down 'OPEN' button to cave, in the center section of the
control panel next to the 'TARGET' button. (Takes 30 seconds)
— Hit 'OPEN' button on upper right hand side of panel until button
lights up
— Hit 'OPEN' button in lower left hand side of panel after you check
with operator to make sure vacuum pressure is low enough.
2. TO MOVE TRAY OR STRIP FRAME AT DISPLAY PANEL
— Switch to manual
— If slide is still showing on TV screen (and "slide in" light is
on), hit black button next to 'Slide Drive' and slide will move
back into tray, (and "slide out" light will be on).
— Hit toggle switch next to 'Tray Drive' toward 'F' (forward) or
'R' (reverse), depending on the direction you wish to move to
tray or strip.
— Hit black button next to tray drive and tray or slide will move
one position at a time. (Remember, only move the tray when the
slide is "out".)
— Hit black button next to 'Slide Drive' and a slide should be
showing on the TV screen. ("Slide in" light will be on)
— Switch back to automatic and hit 'Enter Auto' on the display panel.
— Enter master tray #, hit return, enter position # hit return, and
analysis will begin again.
3. TO CHANGE TRAYS OR STRIP FRAMES
1) Press the green button on the metal box located on the beam line
before the slide changer. This is just to make sure the gate valve
is really closed.
2) Turn blue handle on slide changer slowly to let air into tube.
3) a) FOR TRAYS
— Open bottom of tube and pull out the 3 trays slowly, making sure
the slides do not fall out of trays.
-------
— Replace bottom
— Place an "X" on tray boxes that were just analyzed
— Open top of tube and slide new trays in with position # facing up.
Make sure trays attach in tube and are not at an angle- this may
cause the slide changer to jam.
— Replace top
b) FOR STRIPS
— Move strip back to position #1 by holding both black buttons next
to 'DRIVE' and 'DIRECTION' on side of slide changer
— Remove plastic cover over stripper
— Take out strip, put an 'X' on white side of frame and put into
frame box
— Put next strip to be analyzed into stripper with the blue side of
the frame facing" down and the green dot on the left hand side.
— Replace plastic cover-don't make screws too tight
4) Use red button in back of slide changer to move 1st slide into
place- you know you are in the right position when the #1 slot in the
first tray is covered by the slide changer.
5) Turn blue handle over slide changer to seal vacuum
6) Push white button in back of cave and close the cave
-------
IMPROVE SOP APPEND JULY 1989
APPENDIX 6: Ion Contractor Procedures (RTI)
RESEARCH TRIANGLE INSTITUTE
STANDARD OPERATING PROCEDURE:
ION ANALYSIS OF FILTERS
for
THE NATIONAL PARK SERVICE
Prepared by
Research Triangle Institute
Research Triangle Park, North Carolina 27709
NPS Contract No.: CX-0001-6-0007
Prepared for
U.S. Department of the Interior
National P?rk Service
Washington, D.C. 20013
POST OFFICE BOX 1219.4 RESEARCH TRIANGLE PARK, NORTH CAROLINA 27709-2194
-------
STANDARD OPERATING PROCEDURE
FOR NPS FILTER ANALYSIS
The procedures followed by the Research Triangle Institute (RTI) in the
analysis of National Park Service (NPS) samples under Contract No. CX-0001-6-
0007 are summarized below:
1) RTI receives filters in lots of approximately 100 to 200
each. Sample information is entered onto sample custody
logsheets upon receipt.
2) Upon receipt, the samples are stored in a locked labora-
tory. After the samples have been desorbed, the extract
is stored in a refrigerator in a locked sample custody
room.
3) Within 30 days after sample receipt, Cl", N02, NOg, and
S0$~ analyses are performed on each sample by ion
chromatography.
4) In addition to analyzing the field samples, RTI partici-
pates in EPA-sponsored analytical performance audits which
include the analysis of reference precipitation samples.
5) The data are reported, by filter lot, to the National Park
Service in hard copy and on floppy disk in ASCII format.
Details of these procedures are presented in the following sections.
1.0 SAMPLE STORAGE AND TRANSPORTATION PROCEDURES
Sample collection and transportation of the samples to the RTI Ion Ana-
lysis Laboratory are the responsibility of the particulate monitoring coordin-
ation contractor (PMCC)/NPS.
Samples are shipped to RTI in lots of approximately 100 to 200.
2.0 SAMPLE CHECK-IN AND HANDLING IN THE LABORATORY
Samples are mailed to RTI in secure containers. Upon receipt, RTI does
the following:.
1) Open the shipping box and remove the samples.
2) Record the PMCC/NPS Sample ID, date of receipt, and com-
ments (broken package, contamination, etc.) on the Sample
Log Form (Figure 1).
-------
Date Received
INKS Fi iter Samples
Lot
Sample ID
Extraction
Date
Comments
Analysis
Date
1C
Rerun
Figure 1. Sample Log Form.
-------
3) Store the samples 1n a locked laboratory for future de-
sorption.
3.0 SAMPLE ANALYSIS METHODOLOGY
The analysis procedures for the filter samples are given below. The Ion
chromatographic procedures Initially were taken from "Operations and Mainten-
ance Manual for Precipitation Chemistry Measurement Systems" prepared by Rock-
well International. Some improvements have been made..
3.1 Filter Extraction
Using tweezers, place each filter in a Nalgene Monovette and add 15 mL of
extraction solution (0.0017M NaHC03/0.0018M Na2C03) using a repipet. Push up
the Monovette plunger so that no air space remains above the filter. Expose
the Monovette and solution to ultrasonic energy for 30 minutes and then allow
to sit overnight. This process releases greater than 97 percent of the chlo-
ride, nitrite, nitrate, and sulfate into the solution. Record the date of
extraction on the Sample Log Form.
3.2 Determination of Chloride, Nitrite, Nitrate, and Sulfate Using Ion
Chromatography
3.2.1 Scope and Application —
This method covers the determination of chloride, nitrite, nitrate, and
sulfate in filter extracts.
3.2.2 Summary of Method —
The anions are separated when passed through a resin consisting of poly-
mer beads coated with quaternary ammonium active sites. The separation is due
to the different affinities of the anions for ^he active resin sites.
After separation, the anions pass through a suppressor column which ex-
changes all cations for H+ ions. Species are detected as their acids by a
conductivity meter. (An eluent which yields a low conducting acid is used.)
3.2.3 Interferences —
Large amounts of anions eluting close to those anions of interest will
result in an interference. No interferences have been observed in NPS filter
samples analyzed to date.
-------
3.2.4 Apparatus --
Ion chromatograph (Dionex Manual Model 21201 or Model 14) with anion
guard column, anion fast run separator column, and anion micromembrane sup-
pressor column.
Four-liter collapsible bags.
Pi pets - an assortment of sizes.
Volumetric flasks - an assortment of sizes.
Disposable syringes - 5 ml capacity
Disposable filters - Acrodisc 0.45 /jm, or equivalent
IBM/PC-based Dynamic Solutions Chromatographic Data Acquisition System.
3.2.5 Reagents —
Use ACS reagent grade chemicals for the preparation of all solutions.
Dry chemicals used for the preparation of calibration standards at 105°C for 2
hours and cool in a dessicator immediately before use.
Eluent, 0.0017M NaHCC»3/0.0018M NaeCOs: Dissolve 2.8562 g NaHCC-3 and
3.8156 g Na2C03 in 20 liters deionized water.
Regenerant, 0..025N ^SO^: Add 500 mi IN ^$04 to a Nalgene Carboy and
dilute to 20L with deionized water.
Mixed Stock Solution, 1000 mg/l NO?, NO-j, and SO^2, and 200 mg/L C1-;
Dissolve 1.4998 g NaN02, 1.6305 g KNC-3, 1.8142 g K2S04, and 0.3297 g NaCl in 1
liter deionized water.
Standard Solution A: Dilute 10 ml mixed stock solution to 100 mL with
eluent (100 mg/L NOj NO}, and SO??, and 20 mg/L C1-).
Standard Solution 8: Dilute 10 ml standard solution A to 100 ml with
eluent (10 mg/L NC>2, NOj, S0?2, and 2 mg/L C1-).
Using standard solutions A and B, prepare standards with eluent in 100 mL
volumetric flasks as shown in Table 1. Preparation of standards in eluent
eliminates the water dip which interferes with chloride quantisation. Prepare
fresh standards weekly.
3.2.6 Procedure —
1) Begin the flow of regenerant through the anion micromem-
brane suppressor column.
2) Set up the range for maximum sensitivity (usually 10 /imho
full scale for NPS samples).
-------
TABLE 1. PREPARATION OF ANION CALIBRATION STANDARDS
Standard
1
2
3
4
5
6
7
8
N02, NOg, S(
mg/L
10
5
3
2
10
5
2
1
D$ Cl- mL
mg/L
STANDARD SOLUTION A
2.0
1.0
0.6
0.4
STANDARD SOLUTION B
0.2
0.1
0.04
0.02
mL Standard
Solution/100 mL
10.0
5.0
3.0
2.0
1.0
0.5
0.2
0.1
NOTE: Higher concentration standards can be prepared from Standard
Solution A or the mixed stock solution if needed.
TABLE 2. RECOMMENDATIONS FOR OPTIMUM INSTRUMENT SENSITIVITY
DIONEX
1. 0.0017M NaHC03/0.0018M Na2C03 eluent with fast run separator and sup-
pressor.
2. Detector output range 10 pmho full scale.
3. 100 mL injection loop
4. Flow rate 2.3 mL/min.
-------
3) Begin to pump the eluent through the columns and allow
baseline to stabilize.
4) Arrange calibration standards and samples to be analyzed.
If the conditions suggested in Table 2 are used, an aver-
age of 5 samples can be analyzed per hour. Analyze a
quality control sample first, using a single standard, to
verify that the instrument is operating properly. If the
observed value of any anion differs by more than 10% from
the known value, find and correct the problem before ana-
lyzing any samples.
5) Using a disposable Nalgene syringe fitted with a 0.45 /jm
disposable filter, begin to inject the field samples.
Record on the Analysis Logsheet (Figure 2) the computer
directory name for storing the chromatograms, the NFS
sample ID, and the filename assigned to the chromatogram
for that sample. Record the date of analysis and the 1C
Model Number on the Sample Log Form. Analyze one randomly
selected sample in duplicate. If any sample produces an
anion peak that is offscale, indicate this on the Analysis
Logsheet and the Sample Log Form. Place the sample in a
separate rack for later analysis using a higher detector
output range (usually 100 /jmho full scale).
6) When approximately half of the day's samples have been
"run, perform the daily calibration beginning with the
standard of highest concentration. Record the standard ID
and filename on the Analysis Logsheet.
7) Complete the analysis of the field samples, including an
EPA Quality Assurance sample and a quality control sample.
3.2.7 Calculations Using A Linear Least Squares Fit --
Peak heights are entered into the computer where linear least squares
calculations are performed. The linear least squares fit yields the following
parameters: slope (s), intercept (I), and correlation coefficient (r). The
slope and intercept define a relationship between the concentration and the
instrument response of the form.
yi = sxi + I (1)
where:
yi is the predicted instrument response
Xj is the concentration of standard i.
s is the response slope
I is the intercept
-------
Directory: MAX\DA1
Method name:
)ata Copied (date) :
Mskette No.:
)ata Deleted (date)
Sample ID
"A \ Date:
Analyst:
1C Model No,.:
Project No.:
•
Fi 1 ename
Comments
Figure 2. Analysis Logsheet.
-------
Rearrangement of Equation 1 yields the concentration corresponding to an
instrumental measurement:
xj * (yj - I)/s (2)
where:
Xj is the calculated concentration for a sample
yj is the actual instrument response for a sample
s is the calculated slope from the calibration above
I is the calculated intercept from the calibration above
3.2.8 Quality Control —
Compare the regression parameters for the standard curve with those ob-
tained in the past. If they exceed the control limits, stop analysis and look
for the problem.
Analyze a quality control sample against a single calibration standard at
the beginning of every analytical run. Compare the results with those ob-
tained in the past. If the observed concentration of any anion differs from
the known value by greater than 10%, stop the analysis until the problem is
found. Analyze a duplicate sample and an EPA Quality Assurance Sample each
day.
When a new stock standard solution is prepared, dilute calibration stan-
dards from both the old and the new stock. Analyze the old and new standards
and compare the calibration curves.
3.2.9 Troubleshooting —
Refer to the Dionex Model 212CH or Model 14 Operators' Manual for any
instrumental problems.
4.0 QUALITY ASSURANCE PROCEDURES
The role of any analytical laboratory is to provide qualitative and quan-
titative data which accurately describe the characteristics and/or concentra-
tions of the constituents in the samples submitted. The laboratory data must
be backed up by an adequate program to document the proper control of all the
factors which affect the final result. RTI is committed to the implementation
-------
of a thorough and dependable quality assurance/quality control program which
is understood and followed by all operating personnel and supported by manage-
ment.
4.1 Project Organization and Responsibility
QA project organization is shown in Figure 3. Dr. W. F. Gutknecht, Pro-
ject Manager, provides overall supervision of the project. Dr. E. D. Estes,
Project Leader, is responsible for the analyses conducted during the study and
for data reporting. Dr. W. C. Eaton serves as the program's Quality Assurance
Officer. Ann R. Turner is the sample custodian and analyst.
4.2 QA Objectives for Measurement Data in Terms of Precision, Accuracy^
Completeness, Representativeness, and Comparability
The filter samples collected in this study are analyzed for the parame-
ters summarized in Table 3. The methods used for these measurements also are
summarized in the table.
4.2.1 Precision and Accuracy --
Precision and accuracy objectives for each of the analyses are listed in
Table 1.
4.2.2 Completeness --
Analytical results will be obtained for at least 98 percent of the fil-
ters.
4.2.3 Representativeness --
All filter samples are collected by the PMCC/NPS which is solely respons-
ible for the representativeness of the samples.
4.2.4 Comparability --
All analyses for a given parameter are reported in the same units and are
directly comparable.
4.3. Sampling Procedures
Sampling is the responsibility of PMCC/NPS which is expected to provide
representative filter samples.
-------
NPS
Project Officer
RESEARCH TRIANGLE INSTITUTE
J. B. Tommerdahl
Vice President
Environmental Sciences and
Engineering
C. E. Decker
Director
Center for Environmental
Measurements
PROJECT OFFICE
W. F. Gutknecht
Project Manager
E. D. Estes
Project Leader
KeyTechnical Personnel
ION ANALYSIS
E. D. Estes
A. R. Turner
QA Office
W.. C. Eaton
Quality Assurance
Officer
Figure 3. Project Management Structure
-------
TABLE 3. PHYSICAL/CHEMICAL PARAMETERS AND METHODS FOR ANALYSIS
Unit of
Parameter Report
CI-
NQ?
N03
S0,2
mg/L
mg/L
mg/L
mg/L
Recommended Range of
Methoda Parameter
1C
1C
1C
1C
0.03
0.02
0.002
0.02
- 7.5
- 1.0
- 27
- 22
Estimated
Precision
i 10%
. + 10%
i 10%
i 10%
Estimated1*
Accuracy
+ 10%
+ 10%
+ 10%
+ 10%
a 1C - 1on chromatography
b These values assume concentrations which equal or exceed 100 times the
minimum detectable limit (MDL). For concentrations at the MDL, measure-
ments are accurate to within + 100%. At concentrations equal to 10 times
the MDL, accuracy is within + 20%.
-------
4.4 Sample Custody
Ann R. Turner is the RTI sample custodian and analyst and is responsible
for maintenance of a laboratory sample custody log. As samples are received
in the laboratory and processed, the following information is entered onto the
Sample Log Form.
1) Date samples received in laboratory
2) PMCC/NPS sample identification number
3) Date of desorption
4) Date of ion analysis and instrument used
4.5 Calibration Procedures and Frequency
Calibration procedures and frequency for individual measurements are in-
cluded in Section 3.0 Sample Analysis Methodology.
4.6 Analytical Procedures
Analytical procedures are found in Section 3.0 Sample Analysis Method-
ology.
4.7 Data Reduction, Validation, and Reporting
Data reduction schemes are contained in the methods presented in Section
3.0.
Data entries are reviewed by the Quality Assurance Officer who checks
calculations and verify completeness of data. Relative standard deviations
and relative errors are calculated from analyses of replicates and quality
control/quality assurance samples, respectively. If precision and accuracy
fall outside the data quality objectives summarized in Table 3, analyses are
repeated.
Laboratory notebooks are checked and signed by the Quality Assurance
Officer and then reviewed and signed by the Project Leader. Notebooks and
data sheets are available for review by the Quality Assurance Officer. Ana-
lytical results for each filter lot, including QC/QA data, are compiled by the
Project Leader and presented to the Quality Assurance Officer. Upon approval
of the Quality Assurance Officer, the data are submitted to the NPS.
-------
4.8 Internal Quality Control Checks
Each day of analysis, a quality control sample, a duplicate sample, and a
quality assurance sample are analyzed. Filter blanks are analyzed as they are
supplied by the PMCC.
4.8.1 Filter Blank ~
This is a blank filter supplied by the PMCC which is subjected to the
same preparation procedure as are the field samples being analyzed. This
sample will be used to check for contamination which may occur in sample prep-
aration or analysis.
4.8.2 Duplicate Sample --
This is a sample chosen at random from a day's run. This sample provides
information on analytical precision.
4.8.3 Quality Control Sample --
This sample is prepared by the Project Leader from stock solutions inde-
pendent of those used to prepare the calibration standards. It is analyzed at
the beginning of each run to ensure that the chromatograph is operating pro-
perly. If the observed value of any anion deviates from the known value by
more than 10%, no samples are analyzed until th-e problem is corrected.
4.8.4 Quality Assurance Sample —
This sample is an EPA Quality Assurance sample with known values of the
constituents. The data are reported to the Project Leader who calculates
percentage recovery and reports the data to the analyst and to the Project
Manager. This sample provides an assessment of data quality independent of
analyst's judgement.
4.9 Preventive Maintenance
Analytical instrumentation used in this project will be carried through
preventive maintenance procedures and schedules as recommended by the manufac-
turer.
Dionex Ion Chromatographs require a minimum of maintenance if operated at
pressures less than 800 psi. During the following periods, perform the main-
tenance listed to eliminate unnecessary troubleshooting.
-------
Dally (all operating days):
- Check for leaks at all valves and column fittings (at
normal operating pressure).
- Cycle the injection valve by repeated switching between
INJECT and LOAD and rinse with DI H20.
- Check the meter Zero and Cal adjustments.
- Wipe-up liquid spills and salted-out chemicals.
- Check the drip trays.
Weekly:
- Compare standard chromatograms to check that no signif-
icant changes in column efficiency have occurred.
- Check all air and liquid lines for crimping or dis-
color.
Monthly:
- Check column resolution by measuring percent resolution
of the N03 and 50$ peaks.
- Oil each pump with 2 drops SAE No. 10 oil.
5.0 DATA REPORTING PROCEDURES
The data from the analysis of each lot of filters are reported to the
National Park Service both in hard copy and on floppy disk in ASCII format.
The data are grouped according to the day of analysis and the instrument
(Model 2120i or 14) on which they were run. With each filter lot, quality
control (QC) and quality assurance (QA) data are reported for each instrument,
for each day that samples were analyzed. Reporting of QC/QA data in this
manner facilitates the detection and correction of any instrument-specific
problems. Data from the analysis of daily duplicate samples are compiled and
periodically reported to the NPS so that precision estimates can be made.
-------
IMPROVE SOP APPEND JULY 1989
APPENDIX 7: Carbon Contractor Procedures (DRI)
Titles Thermal/Optical Carbon Analysis of Aerosol Page 1 of
Filter Samples
Numbers 2-204.1 Datet 5/2/89
1.0 GENERAL DISCUSSION
1.1 Purpose of Procedure
This standard operating procedure is intendeds
- to provide a basic understanding of the principles behind carbon
analyzer operation;
- to describe routine determination of organic, elemental, and
carbonate carbon from ambient and source filter samples using the
OGC/DRI thermal/optical reflectance carbon analyzer;
- to detail the concerns and procedures which will insure a
state-of-the-art carbon analysis measurement process.
This procedure will be followed by all analysts at the Environmental
Analysis Facility of the Energy and Environmental Engineering Center
of the Desert Research Institute.
1.2 Measurement Principle
The OGC/DRI thermal/optical carbon analyzer is based on the preferential
oxidation of organic and elemental carbon compounds at different
temperatures. It relies on the fact that organic compounds can be
volatilized from the sample deposit in a helium (He) atmosphere at Jow
temperatures while elemental carbon is not oxidized and removed. The
analyzer operates by 1) liberating carbon compounds under different
temperature and oxidation environments from a small punch taken from
a quart?: fibnr filter; 2) converting these compounds to carbon dioxide?
(C02) by passing the volatilized compounds through an oxidizer (heated
manganese dioxide, MnO2); 3) reduction of C02 to methane (CH4)
by passing the flow through a methanator (hydrogen-enriched nickel
catalyst); and 4) quantification of CH4 equivalents by a flame
ionization detector (FID).
The principal function of the optical (laser reflectance) component of
the analyzer is correction for pyrolysis of organic carbon compounds
to elemental carbon. Without this correction, the organic carbon
fraction of the sample would be underreported and the elemental carbon
fraction would include some pyrolyzed organic carbon. The correction
for pyrolysis is made by continuously monitoring the filter reflectance
(via a helium-neon laser and photodetector) throughout an analysis
cycle. This reflectance, largely dominated by the presence of black
elemental carbon, decreases as pyrolysis takes place and increases as
elemental carbon is liberated during the latter part of the analysis.
By monitoring the reflectance, the portion of the elemental carbon
peak corresponding to pyrolyzed organic carbon can be accurately
assigned to the organic fraction. The correction for pyrolytic
conversion of organic to elemental carbon is essential for an unbiased
measurement of both carbon fractions, as discussed in Johnson et al.
-------
(1981).
Carbonate carbon may be determined by measuring the C02 evolved
upon acidification of the sample punch before the normal carbon analysis
procedure.
1.3 Measurement Interferences and Their Minimization
Carbonate carbon presents significant interference in carbon analysis
if it consititutes more than 5% of total carbon in the ambient or source
sample, as it is measured as both organic and elemental carbon during
thermal/optical carbon analysis. Acid pretreatment of the filter samples
can eliminate the carbonate interference.
The presence of certain minerals in some soils can affect the laser
correction for pyrolysis. These minerals change color as the sample
punch is heated, generally resulting in a sample which is darker. For
samples which contain large fractions of resuspended soils, the split
between organic and elemental carbon may have to be estimated manually.
Some minerals, again predominantly in soil samples or soil dominated
samples, may affect the laser reflectance by temporarily changing color
or changing the surface texture of the deposit residue. Unlike the
effect described above, these changes are reversible and highly
temperature dependent.
Some colored organic compounds can affect the laser correction
as well, causing increased reflectance as these compounds are removed.
This effect is readily ascertained by examining the laser response
during the organic portion of the analysis. Again, the split between
organic and elemental carbon may have to be estimated manually if the
effect is large.
Finally, the presence of certain elements (Na, K, Pb, Mn, V, Cu, Ni,
Co, and Cr) existing either as contaminents on the filter or as part of
the deposit material has been shown to catalyze the removal of
elemental carbon at lower temperatures (Lin and Friedlander, 1988).
Such catalysis would affect the distribution of carbon peaks during the
analysis.
1.4 Ranges and Typical Values of Measurements
A wide range of aerosol concentrations can be measured with this
method, the limiting factor being the concentration of the carbon
compounds on the filter on a ug/cm2 basis. Dirtier environments
may be sampled and still be analyzed within the range of the carbon
analyzer by increasing the filter deposit area or by decreasing the
. flow through the filter media.
The carbon analyzer can effectively measure between. 0.2 and 750 ug
carbon/cm2. The upper range is somewhat arbitrary, depending on the
particular compounds on the filter and the temperatures at which they
evolve. This upper range may be extended by taking special precautions,
such as reducing the punch size or by special temperature programming,
to avoid an over-range FID signal.
Typical carbon values range between 10 and 100 ug carbon/cm2 for ambient
samples.
-------
The lower quantifiable limits (LQLs) of carbon combustion methods
depend upon the variable carbon content of the blank quartz filters as
well as the analysis method. For better LQLs , the unexposed filters
should be pre-fired in an oven at high temperatures for several hours
to remove any residual carbon contamination (Fung, 1986; Huntzicker,
1986; Rau, 1986). All quartz filters originating from DRI are pre-fired
for a minimum of four hours at 900 C and are tested for blank levels
before use. For well-cleaned quartz filters, the standard deviation of
the blanks for organic and elemental carbon is on the order of 0.5 and
0.2 ug/cm2, respectively (Fung, 1986). Typical pre-fired blank levels
at DRI are 0.5 - 1.0 ug organic carbon/cm2 and 0.0 - 0.2 ug elemental
carbon/cm2 . Because even pre-fired filters can absorb organic vapors
during shipping and storage, the LQL of analysis on a particular set of
filters depends on the number of field blanks analyzed and the
variability in the results front those blanks.
Acid-evolved carbonate levels in pre-fired quartz filters have been
shown in several informal tests at DRI to be quite variable over
time. Part of this phenomenon is apparently due to the reaction of
ambient C02 with alkaline sites on the quartz fibers . Acceptance
testing for carbonate is not routinely performed at DRI .
The precision of this analysis has been reported to range from 2 to
4% (Johnson, 1981). For analysis of actual ambient and source filters,
homogeniety of the deposit is most important for reproducible results.
For evenly loaded filters, precision is generally 5% or less; for poorly
loaded filters, replicates may deviate by as much as 30%. The
precision of carbonate analysis results is approximately 10%.
The precision of the laser-dependent split between organic and elemental
carbon fractions depends upon how rapidly the laser is increasing at
the time of the split and whether the split is made in the middle of
a carbon peak or not. Typically, relative laser split times are
reproducible within 10 seconds and deviations in calculated splits are
less than 5% of the total measured carbon.
The accuracy of the thermal/optical reflectance method for total carbon
determined by analyzing a known amount of carbon is between 2 to 6%
(Rau, 1986). Accuracy of the organic/elemental carbon split is between
5 and
1.6 Personnel Responsibilities
All analysts in the laboratory should read and understand the entire
standard operating procedure prior to performing carbon analysis, which
includes routine system calibration, actual analysis, and immediate
review of the data as it is produced to correct system problems.
It is the responsibility of the laboratory manager or supervisor
to ensure the carbon analyses procedures are properly followed, to
examine and document all replicate, standard, and blank performance test
data, to designate samples for reanalysis, to arrange for maintenance
and repair, to maintain the supplies and gases necessary to insure
uninterrupted analysis, and to deliver the analysis results in dBase
format to the project manager within the specified time period.
The quality assurance (QA) officer of DRI ' s Energy and Environmental
-------
Engineering Center is responsible to determine the extent and methods of
quality assurance to be applied to each project, to estimate the level of
effort involved in this quality assurance, to update this procedure
periodically, and to ascertain that these tasks are budgeted and carried
out as part of the performance on each contract.
1.7 Definitions
The following terms are used in this document:
Calibration injection - the injection of calibration gases (methane in
helium [CH4/He] or carbon dioxide in helium [C02/He]) into the sample
stream to check instrument performance.
Calibration peak - the FID peak resulting from the automatic injection
of methane calibration gas (CH4/He) at the end of each analysis run.
All integrated peak areas are divided by the calibration peak area
and multiplied by an instrument-specific calibration factor to obtain
ug carbon.
Elemental carbon - carbon evolved from the filter punch in a helium/
oxygen (He/02) atmosphere at 550, 700, and 800 C minus pyrolyzed
organic carbon.
Laser split - the time at which the laser-measured reflectance of the
filter punch reaches its initial value, indicating that all pyrolyzed
organic carbon has been removed and original elemental carbon is
beginning to evolve.
Lower split time - the time at which the laser-measured reflectance of
the filter punch reaches its initial value minus the precision of the
laser signal (currently defined as 10 counts).
Organic carbon - carbon evolved from the filter punch in a He
atmosphere at 120, 250, 450, and 550 C plus pyrolyzed organic carbon.
Pyrolysis - the conversion of organic carbon compounds to elemental
carbon due to incomplete combustion/oxidation; may be envisioned
as "charring".
Pyrolyzed carbon - the carbon evolved from the time that the carrier
gas flow is changed from He to He/02 at 550 C to the time that the
laser-measured filter reflectance reaches its initial value.
Upper split time - the time at which the laser-measured reflectance of
the filter punch reaches its initial value plus the precision of the
laser signal (currently defined as 10 counts).
1.8 Related Procedures
SOP's related to carbon analysis activities which should be reviewed
in conjunction with this document are:
DRI SOP t6-001.1 Shipping and Mailing Procedures.
DRI SOP #6-009.1 Field and Laboratory Safety Procedures.
any SOP's dealing with filter handling and shipping in conjunction
-------
DRI SOP #4-001.1 Creation, Revision, Distribution, and Archiving
of Standard Operating Procedures.
DRI SOP # Pre-Firing of Quartz Filters for Carbon Analysis
The maintenance and troubleshooting guide for the DRI/OGC carbon
analyzer.
The appropriate MS-DOS or PC-DOS manual for the computer used with
the carbon analyzer.
2.0 Apparatus, Instrumentation, Reagents, and Forms
2.1 Apparatus and Instrumentation
2.1.1 Description
The components of the DRI/OGC thermal/optical carbon analyzer are
depicted in Figures 1 and 2; the complete gas flow schematic is shown in
Figure 3. The programmable combustion oven is the heart of the carbon
analyzer and includes loading, combustion,, and oxidation zones in
a single quartz "oven" as depicted in Figure 4.
In addition to the DRI/OGC thermal/optical analyzer connected to a
IBM-PC or compatible computer, the following items are needed for
routine carbon analysis:
- Punchs 0.503 cm2 area for removing small sample punches from
quartz filters. This punch needs to be kept clean and
sharp. If the punch is sharpened, the punch area must
be reverified.
- Syringes: a gas-tight 1000 or 2500 ul syringe for calibration
injections; 25 or 50 ul syringe for carbonate analysis
and for analyzer calibration.
- Quartz filters: Pallflex 2500QAT-UP or equivalent.
- Tweezers.
- Glass petri dish.
- Log book/notebook.
- Transparent tape.
- Kimwipes.
- Small styrofoam cooler.
- Blue ice.
- A copy of Carbon.EXE (the analysis program), version P2.1 or
later, and Carbon.DAT (the analysis parameter file), version
D2.0 or later.
2.1.2 Instrument Characterization
-------
Vl
o
M
a)
5
o
•g
lit
u
u
CU
o
10
M
-------
§•
g
1
o
a
c
o
o
3
J
•5 <*•
* O
0 &
I x
. °
JJ 1
• o 3
: Is 4
z u > C
» 1 NISI
s
T3
1
j
OTD
||j|
? ? 1 ;
1335
a i fl >
^
-------
o
c
o
^-1
0)
CL 0)
Q
1°
L J
_
asuopayay
c
en1 o
I*
o .*_/
0)
.4_*
D
a.
S
o
u
o
L.
u
u
r
0)
X
(U
'X
OT
Q.
O
O
Q-
_O
Q.
W
Q
L .
(O
VI
o
2
0)
en
O
I—
O
-*— '
B
a
VI
00
to
a.
o
-------
o
c
o
•H
u
w
3
O
U
1-4
0)
M
0.
o
OS
Q
O
o
-------
_,—^ is program-vanven ana aata is stored
automatically to disk via an IBM-PC compatible computer. Response
times and signal lag times are built into the parameter file which
is loaded when the analysis program begins. The program is event-
driven; that is, when the FID signal returns to its baseline
after a minimum of 80 seconds at one condition, the program will advance
to the next temperature or carrier gas mixture. A maximum time limit
per condition is also established to prevent a slight baseline drift
from holding the analyzer in one condition indefinitely.
This method requires no sample pretreatment, requires between 15 and
70 minutes per sample, and destroys the sample.
Operator concerns for correct routine operation of the instrument
should be the following (refer to section 4 for more details):
- Remember to push to sample in when the tone sounds; DO NOT leave
the room until the analysis begins.
- Insure that the thermocouple is physically decoupled from the
sample boat after pushing in the sample to prevent oven temperature
from influencing the laser reflectance signal..
- Insure that the sample port is tight after loading a sample punch.
- Check the graphical printout after each analysis run to insure that
the FID, temperature, and laser signals are behaving as expected.
- The analyzer's quartz oven is susceptible to breakage, especially
at the sample port. Care should be taken to avoid exerting
tangential pressure on the oven when manipulating the sample port
fitting.
2.1.3 Maintenance
Regular maintenance for the analyzer involves daily checking of
compressed gas supplies, cleaning the punch and tweezers between each
sample, and backing up data files on a regular basis. Checks
of laser adjustments (physical and electrical) are made at least
monthly; analyzer calibrations are performed every six months. All
calibrations and repairs must be recorded in the log book.
Refer to the maintenance and troubleshooting guide for additional
information.
2.1.4 Spare Parts
It is strongly recommended that the following spare parts be kept on
hand to insure minimal interruptions in analysis:
- Quartz rods: 3 mm nomimal diameter, Homosil optical quality rod,
available from GM Associates (Oakland) cut to 9 3/4" lengths and
polished on both ends.
- Quartz ovens: specially built ovens by the Oregon Graduate Center
glass blower.
- Quartz boats: made in-house from scraps of broken ovens.
- Thermocouple rods: 18" length by 1/8" OD, type K ground isolated
-------
with J14 stainless steel sheaths (March! Associates, #SDH175).
- FID flame tips: for Gow-Mac #12-800 FIDs (Gow-Mac, #132-117).
- Fuses: 15 A, MDL 15, slow-blow.
- Punches: smaller than the 0.503 cm2 punch normally used, for
excessively heavily loaded samples.
- Septa: 1/4" and 1/8", for injection ports.
- Replacement needles for syringes.
- Replacement scrubber tube: Supelco oxygen scrubber (Supelco,
#2-2396).
- Stainless steel wire: for forming "ears" to hold the sample boat
in position and for wrapping the "ears" onto the thermocouple
push rod (Rocky Mountain Orthodontics, #RMO E-19, 0.914 mm).
- Quartz wool: for repacking the oxygen oven (Alltech Associates,
#4033).
- Teflon ferrules: Parker or Swagelok style, 1/2" ID, for the sample
port fitting.
- Teflon ferrules: 1/2" OD by 1/8" ID, for the thermocouple rod
at the back of the oven.
- Heating element for oven: custom made 650 W coiled heater (Marchi
Associates #SDH175).
- FID batter: 300 VDC (EDCO, EverReady #495).
- Printer paper.
- Printer ribbons.
- Computer disks, double side, double density.
2.2 Reagents
The following chemicals should be reagent grade or better:
- Potassium hydrogen phthalate fKHP), for calibration use (Fisher,
SP-243).
- Sucrose, for calibration use (EM Science, ISX1075-1).
- Manganese dioxide (Mn02), crystalline, as an oxidizer in the
oxygen oven (Nurnberg Scientific, #RM200).
- Hydrochloric acid (HC1), 0.4 molar solution, for use in cleaning
punch and quartz ovens, and for use in carbonate analysis.
- Distilled dionized water (DDW): total carbon background should be
6 ppm or less.
2.3 Gases
-------
The following compressed gases should be zero grade or better:
- Helium for a carrier gas, regulated to 35 psi with a metal
diaphragm regulator. The higher pressure is required due to
the pressure drop across the Supelco oxygen scrubber.
- 5% methane by volume in helium for calibration injections and
calibration peaks; regulated to 10 psi by a metal diaphragm
regulator.
- 5% carbon dioxide by volume in helium for calibration injections;
regulated to 10 psi by a metal diaphragm regulator.
- 10% oxygen by volume in helium as a carrier gas, regulated to
10 psi by a metal diaphragm regulator.
In addition, the following gases are required:
- Hydrogen for the FID flame, regulated to 10 psi with a metal
diaphragm regulator.
- Compressed air to supply oxygen to the FID, regulated to 10 psi
by a metal diaphragm regulator.
At least one backup cylinder per gas type should be kept on hand
at all times. The calibration gases typically last for one year.
The hydrogen, helium, and 02/He mixture are typically replaced
every three to five weeks. The compressed air is replaced every
4 to 5 days. All gases are replaced when the cylinder pressure
drops below 500 psi.
2.4 Forms and Paperwork
All samples are logged into the "Air Analysis Logbook" upon receipt at
the laboratory. Refer to Figure 5 for the format of this logbook.
A sample analysis list will be prepared by the laboratory manager
indicating which samples will be analyzed and any special instructions.
Samples designated for carbon analysis are logged into the "Carbon
Analysis Logbook" prior to analysis; Figure 6 provides a sample of
entries in this logbook.
As individual samples are analyzed, entries are made in the "Carbon
Analyzer Logbook", as shown in Figure 7. As each analysis run is
completed, the sample analysis list is marked with the date and
analyzer number, as in Figure 8.
3.0 Calibration Standards
3.1 Preparation, Ranges, and Traceability of Standards
Four standards are used in calibrating the carbon analyzers: 5% nominal
CH4 in He, 5% nominal C02 in He, KHP, and sucrose. Only the calibration
gases are used on a daily basis as analyzer performance monitors. KHP
and sucrose are used in conjunction with the two gases semiannually to
establish the calibration curve of each analyzer.
The calibration gases are assayed for exact concentrations by the gas
supplier; the assay value is obtained from the tag on the cylinders and
is typically determined by gas chromotography (GC) or gravimetry.
-------
G
Cj r-
^- rv
u
a
00
O
•3
1O
j-
ti
U(
V
•x
.r
"A
-------
I I
—w
\
°
.0
o
I ' I
T
«M+
1' i '
-/iH '-!"
_
_ _
'~l" t~M i' ,
:-j-rrrrf
~r
\-vp\-1
Vv\n\r
4±j
-!f i '
_LJ_U_LJ_
t
o 10
b
i!f jr
i i
,
MM
! i i
£>;_ !_ (T)
».{
-i—)_..
i i
Figure 6. Example of Carbon Analysis Logbook Formac
-------
-^H~
7 i'"i M M F IT 7
7—•>— —; -i • - -I -7—1—•—,•
i
nni
3
I I ! i !
-D
j i-44-|..j.|..
o
b
o
o
^3
m
c
<
o
a)
r-1
Q.
-------
IMPROVE: Lot G
>*******************
ate: 4/15/89
rom: L.Pritchett
'o : J.Chow
J.Watson
C.Frazier .
Carbon Room
'otal nvunber of samples: 418
ipecies to be analyzed:
OC/EC by carbon analyzer
.nstmctions:
1. This is the sixth analysis list for IMPROVE carbon analysis. We
are processing filters in order of lot number, starting with the
oldest lots first.
2. Note that we are using analysis IDs on these samples, given that
the actual filter IDs are too complicated and too long for the
current carbon program. BE VERY CAREFUL THAT THE FILTER ID MATCHES
THE ANALYSIS ID ON THIS LIST. IF THERE ARE ANY DISCREPANCIES, SEE
LYLE BEFORE PROCEEDING WITH THE ANALYSIS.
3. Carbon analysis on lot G will begin the week of April 17, 1989,
immediately following the Santa Barbara, Phoenix Pilot Study, and
IMPROVE Lot F samples.
4. Deposit area for the quartz filters is 3.8 cm2.
5. dBase III files will be named:
carbon data : IMCarG.DBF
Sample ID
Filter ID
OC/EC
P00701
P00702
P00703
P00704
P00705
P00706
P00707
P00708
P00709
P00710
YOSE020288C1
YOSE020288C2
SAG0020288C1
ARCH020288C2
ARCH020288C1
REDW020288C1
PORE012688C2
GRCA020288C1
MORA020288C2
BRCA020288C1
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Figure 8. Example of Carbon Analysis List
-------
The KHP is dried at 110 C for two hours before dispensing. Transfer
0.3826 g of KHP into a glass 100 ml volumetric flask. Dilute to
volume with 0.4 ml concentrated hydrochloric acid (HC1) and dionized
distilled water (DDW). Mix the KHP thoroughly. Store this solution
in a refrigerator until it is used for calibration purposes. This
solution is good for about 40 days. Label the flask with the chemical
name, the date of preparation, the name of the chemist preparing the
solution, and the exact concentration. The concentration, nominally
1800 ppm carbon, is calculated by
/actual g KHPj/8*12 g Carbon \f 10"-3 ml j/10"6 ug J ug Carbon
\ 100 ml J ^204.23 g KHpy\ ul J\ g J ul solution
The nominal 1800 ppm sucrose solution is prepared by transferring
0.428 g of sucrose into a glass 100 ml volumetric flask. Dilute to
volume with DDW. Mix the sucrose thoroughly. Store this solution
in a refrigerator until it is used for calibration purposes. This
solution is good for about 40 days. Label the flask with the chemical
name, the date of preparation, the name of the chemist preparing the
solution, and the exact concentration. The concentration is calculated
by
actual g Suc\/12*12 g Carbon j /10~-3 ml \ /10~6 ug\ ug Carbon
100 ml /I 342.31 g KHP / I ul /( g / ul solution
To prepare a blank solution, add 0.4 ml of concentrated HC1 to a glass
100 ml volumetric flask and dilute to volume with DDW. This acidified
DDW is made fresh each time a 1800 ppm KHP stock solution is prepared.
No primary standards currently exist for carbon analysis. Ideally,
such standards should include a range of organic compounds from low
to high molecular weights and with varying degrees of susceptibility
to pyrolysis, as well as elemental carbon and carbonate compounds.
Currently, KHP, sucrose, and the two calibration gases are used at DRI
for calibration and system audit purposes.
3.2 Use of Standards
The calibration slopes derived from the two gases and the KHP- and
sucrose-spiked filter punches are averaged together to yield a
single calibration slope for a given analyzer. This slope represents
the response of the entire analyzer to generic carbon compounds
and includes the efficiencies of the oxidation and methanator zones
and the sensitivity of the FID. Note that the current calibration
procedure is based only on the total carbon; currently no routine
procedure exists to check the accuracy of the OC/EC split.
3.3 Typical Accuracy of Calibration Standards
The accuracy of the calibration standards is primarily limited by the
accuracy of the calibration gas assays and by the care taken during
preparation of the KHP and sucrose solutions. The calibration slopes
determined by these four compounds historically differ by less than 5%
-------
on a given analyzer if sufficient care is taken during the calibration
procedure (Section 5.1).
4.0 Procedures
4.1 General Flow Diagram
The typical flow of samples and data for carbon analysis is depicted in
Figure 9.
4.2 Analyzer Start-Up
The following steps outline analyzer start-up:
- Check all gas cylinders' pressures; cylinders with gas pressures less
than 500 psi should be replaced before beginning the day's analysis.
- Check that all gas delivery pressures are correct:
Hydrogen — 10 psi
Helium — 35 psi
Compressed air — 10 psi
02/He mix — 10 psi
CH4/He mix — 10 psi
C02/He mix ~ 10 psi
- Check that all FIDs are lit by holding a pair of tweezers over
the FID exhaust stack and watching for condensation. If the FID
is not lit (as immediately after the hydrogen or compressed air
cylinders are changed), relight the flame by turning the H2
rotameter to 50 and holding a butane lighter or match over the
FID stack. A light pop indicates that the flame is lit. Verify
that the flame remains lit by the tweezer test. Often the flame
will not stay lit the first time, especially after the hydrogen
cylinder is changed and air gets into the gas lines. If the FID
is cold, allow at least 30 minutes at the high gas flow to pass
before turning the H2 rotameter to its correct setting.
- Check and readjust if necessary all gas flows at the analyzer. The
correct readings are posted on each rotameter. Read through the
center of the ball. If drastic adjustments are required on one
analyzer, recheck that flows on the other two analyzers have
not been affected.
Turn on the computer monitor. Note: the computers are generally
left on at all times; only the monitors are turned off at night
to avoid "phosphor burn".
If the computers have not been used for more than one day, reset
the date by typing "DATE" and answering the question or by
rebooting the computer ( ).
If the computers on carbon analyzers #2 or 13 are rebooted, turbo
mode must be set. For the computer on analyzer #2, press
<+>, using the plus key on the far right side of the
keyboard. For the computer on analyzer #3, press <->,
using the minus key on the far right side of the keyboard. In
both cases, the screen cursor will change size as turbo mode is
initiated.
- At the C> prompt, type "CARBON" to begin the carbon program.
-------
Insure that the sample port fitting is tight and that the thermo-
couple push rod is reasonably snug at the back fitting. If the
push rod is loose, tighten the rear fitting NO MORE than 1/16 of
a turn. Do not overtighten this fitting: a push rod that is too
tight is not only difficult to operate smoothly, but will cause
excessive wear of the Teflon ferrule.
From the opening menu, select option 4; see Figure 10. After
insuring that the thermocouple push rod is pushed into the
combustion zone, type "Y" to begin baking the oven. The oven will
be baked at 800 C for 10 minutes to insure that the system is clean
before beginning analysis. This option is self-timed and will turn
off the oven after 10 minutes has elapsed.
After the baking cycle is complete and while the oven is cooling,
backup the previous day's analysis data by ending the carbon
program (press ), changing to the appropriate directory, and
typing "BACKUP C: A: /M". The system will prompt for a formatted
disk to be placed in drive A:. NOTE: if data must be archived from
multiple directories, begin backing up the data with the above
command; after changing to the second and subsequent directories,
type "BACKUP C: A: /M /A". If the "/A" is left off the command,
subsequent backup attempts will erase the first set of data. After
all backups are complete, label the disk in the following format:
\directoryname Analyzer* BACKUP
dateofanalysis disk n of n
For example:
\IMPROVE\LOTC C/A fl BACKUP
880714 disk 1 of 1
Wipe the sample tweezers, petri dishes, and sample punch with
clean KimWipes, taking care not to contact the cleaned surfaces
with fingers or other dirty items.
Begin the daily entry in the carbon analyzer logbook. Entries
should follow the format in Figure 7.
Insure that the printers have enough paper for the day and that the
ribbon is producing legible printing.
After the ovens have cooled to less than 100 C, perform a leak test
on the system by flipping off the "From Oven" toggle valve. After
the He-1 and He-2 rotameters settle to zero (if they don't reach
zero in 2 minutes, see leak fixing procedures below), flip off the
"To Oven" toggle valve. This process pressurizes the oven and
connecting tubing and then isolates the oven. After 30 seconds,
flip on the "To Oven" toggle valve. If the He-1 rotameter float
jumps more than 5 units, the system has an unacceptable leak.
Correct the leak by checking the following items:
- Check that the sample port fitting is tight.
- Check that the push rod is snug.
- If the system still does not leak check, disassemble the
sample port fitting, wipe all threads and ferrules clean
-------
.7
j
7
** CARBON ANALYST; PROGRAM **
Desert Research Institute/EEEC/LCP
Ver. P2. I 02/10/8*
/./MMMMMMMMMMMMMMMMMMMMMMMMMMMMS
J Current Directory: C:\ODEQ .?
TMMMMMMMMMMMMMMMMMMMMMMMMMMMM,S
Program options:
1 ... Normal DC/EC run
2 ... CO3 + OC/EC run
3 ... Calibration injection run
A ... Oven baking cycle
5 ... Manual mode
6 ... Recall previous data
7 ... Chanqe current directory
8 ... Print daily summary files
•'Esc>. .. End program
Input option:
Figure 10. Carbon Program Main Menu
-------
with a clean, dry KimWipe, reassemble, and retry.
- If the system continues to leak, check the integrity of
all tubing and of the quartz oven. Refer to the carbon
analyzer troubleshooting manual for additional tips and
procedures.
When the system leak checks satisfactorily, from the main menu
of the Carbon program select option 5. This will result in the
screen shown in Figure 11. While watching the He-1 and Cal
Gas rotameters, select option 6 (calibration gas load) . The
He-1 rotameter should not change from zero, and the Cal Gas
rotameter should momentarily dip down. While watching
the same rotameters, select option 5 (calibration gas inject).
The He-1 rotameter should jump up momentarily and the Cal Gas
reading should jump slightly. Behavior different than this
indicates a leak in the calibration gas injection system which
must be corrected before beginning any analyses. Refer to the
carbon analyzer troubleshooting manual for additional information.
Because calibration gas has been injected into the system by the
above step, the system must be purged before continuing. Open
the "From Oven" toggle valve to restore flow through the oven and
wait at least two minutes to insure all calibration gas has
progressed through the system.
Press to return the Carbon program to the main menu. Select
option 3 to begin the morning calibration injection. Select
He/02 carrier gas (option 2). Select either C02 or CH4
calibration gas type as the same gas used the previous afternoon.
For any given day, one gas will be used in the morning and the
other in the afternoon. By using the same gas in the morning as
was used the previous afternoon, the calibration gas used in the
morning will be rotated on a regular schedule.
The computer will create a sample ID based on the gas type, current
date, and run number. This ID should be entered in the analyzer
logbook (see Figures 12 and 7). Press in response to
the purge option to begin the calibration run.
Insure that the printer is on-line.
- When the elapsed time reaches 90 seconds (Figure 13), flush the
1000 or 2500 ul syringe with the appropriate calibration gas three
times. A low pitch warning tone will sound at 114 seconds (the
number of beeps corresponds to the carbon analyzer number). When
the analysis start tone sounds at 120 seconds, inject 1000 ul of
the calibration gas into the injection port before the oven. The
rest of the analysis is automatic.
- When the analysis is complete, a tabular and graphical printout
similar to Figures 14 and 15 will be generated. From the
tabular printout locate the calibration peak counts and the
calculated ug C/filter. Record these values in the logbook as
in Figure 7. The calibration peak counts should be in the
following ranges:
Analyzer #1 : 25000 counts to 27000 counts
Analyzer #2 : 23000 counts to 25000 counts
Analyzer #3 : 21000 counts to 23000 counts
-------
/.'MMMMMMMMMMMMMMMMMMMMMMMMMMMMM6'
J Current Directory: C:\CftLIB J
TMMMMMMMMMMMMMMMMMMMMMMMMMMMMM v
Carrier gas options:
1 ... He on 1x
2 ... He/02
Input option: 1
DODODDDDDDDDDODODODDDCfODDDOOODDDDDODDOODDDDDODDODDOODDODODDDOODDDDOODOOOCDODOi
Calibration gas options:
1 ... methane
2 ... C02
Input option: 1
DDDODDDODDDDODDODDDC€>DODOODOOOOODOOOOOODDDDDDDDODDDDDDOOODDDDOOODDOODDDODOODOl
Data will be written to MI0224-2
Does the oven need to be purged'"' N
Ready to begin run; press any key when ready ...
Figure 11. Carbon Program Before Starting Calibration Run
-------
C.O
I
§
o
o
Q)
on
o
CT-
-------
Analyzer
CARBON CALCULATION RE'rULT'3
Carbon. FAS: P2. 1 02/10/8'=
Carbon.DAT: 02.1 02/lO/ij
Analysis ID
'Sample ID
Carrier gas
Anal ysis
Calculation
Anal program ver
Calib. slope
Calib. intercept
Baseline window
MI 022 A- 2
MI0224
He 1 i urn on 1 y
02/24/8'3 14:41
02/24/8* 14:52
P2. 1 C02/10/8*.>
Parm
23.64 ug C/peak-to-calibration
0 . 00 ug C
1 counts
Sample transit : 2b sec
Laser
Calib
file ver :
peak ratio
prec i s i on :
transit :
C>2. 1 (.02/1 0/8'=-)
10 counts
•50 sec
Calibration peak area: 26i;i9fe counts
Initial FID baseline : 107 counts
DC Peak ttl
Peak Area
24573 counts
Carbon
21.52 ug C/punch
* •«••** •*•«••*•»••<*•*-»# -i» •»• •*•»
Calculated Carbon
21.5 ug C/cm2
21.5 uq C/filts-r
Figure 13. Example of Tabular Printout
-------
_
•X-
O
01
H
O
CD
I—I
O.
9)
U
3
Ml
C.O
s
£3
<0
-------
UMMMMMMMMMMMMMMMMMMMMMMMMMMMM8
J Current. Directory: C:\QOEQ J
TMMMMMMMMMMMMMMMMMMMMMMMMMMMM .>
Program option : Normal OC/EC run
Enter full sample ID: 'BQ7176
Sample ID is "SQ7176". Is this okay? Y
Enter run/punch number for this sample <1-9): 1
OOOODDDDDDDDDDOODDDDDDDDODOODDDODODODOODODODDOODOODDOODDOODDODODODODODODDOO:
Punch size options:
1 ... 0.503 cm2
2 ... 0.ASA cm2
3 ... 1 cm2
A ... other
Input option:
1 ... 0.503 cm2
2 ... 0.A8A cm2
A ... other-
Input option: 1
1 ... 1 cm2
2 ... 7.8 cm2
3 ... 6. A cm2
A ... 8 . ~ cm 2
5 ... 13.8 cm2
b . . . other-
Input option: 5
Data will be written to 'i07l7t-l
Ready to begin run: press any Key when ready . ..
Figure 15. Carbon Program Before Beginning Regular Analysis
-------
j.n tne roj.xowing ranges:
CH4 gas : 20 to 22.0 ug C
C02 gas : 19.0 to 21 ug C
If results different from these are obtained, rerun the calibration
injection run. If results are still out of the above ranges,
locate and correct the problem; refer to the carbon analyzer
troubleshooting manual.
From the main menu of the Carbon program, select option 7 to
change to the appropriate subdirectory for the samples to be
analyzed. The new subdirectory name, if valid, will be displayed
at the top on the screen.
Based on the analysis list for the day, retrieve the samples to
be analyzed from the sample freezer and place in a styrofoam cooler
with blue ice. Place the cooler in the instrument room.
4.3 Routine Operation
Routine analysis procedures depend on whether or not carbonate carbon
will be determined before OC/EC analysis. The procedures are different
for these two options.
4.3.1 Routine OC/EC Analysis
- Pull the push rod back to the idle zone of the quartz oven
(approximately half way between the sample port and the heating
elements). Allow the boat/push rod to cool until the reading
on the front of the analyzer reaches 50 C or less. Do not
pull the boat into the sample loading zone when the boat is still
hot as the heat will affect the Teflon ferrules of the sample
port fitting.
Insure that the petri dish, tweezers, and punch are thoroughly
wiped clean with a dry KimWipe.
Based on the analysis list, remove the sample to be analyzed from
the styrofoam cooler.
Remove the filter from the PetriSlide or petri dish with tweezers,
taking care to handle the filter only by the edge. Place the filter
on the glass petri dish and remove a sample punch by pushing down
gently on the punch. Rocking the punch slightly will insure that
the punch is complete severed. Try to remove the punch from the
edge of the deposit to avoid wasting the filter, but try to avoid
areas of non-uniform deposits. Leaving the sample punch in the
punch, place the punch under a KimWipe. Replace the filter in the
PetriSlide or petri dish.
Record the filter ID in the analyzer log book (Figure 7).
After the boat has cooled to 50 C or less, loosen the sample port
fitting carefully with a wrench. NOTE: avoid exerting any sideways
pressure on the quartz oven. Try to confine the wrench pressure
to only rotational torque. Loosen the front fitting before
attempting the rear fitting. Slide the sample port fitting forward.
-------
Pull the boat back until it is centered in the sample port, taking
care that the small stainless steel "ear" holding the boat to the
push rod does not catch on the sample port opening and bend. If the
"ear" does bend, carefully bend it back into position with tweezers
or small clean pliers.
Using the tweezers, push the bottom of the punch in the boat forward
so that the top of the punch can be accessed. Remove the punch
and place it on the top of the analyzer.
Pushing the bottom of the sample punch in, remove the sample
from the punch and place it in the sample boat. Generally, the
punch must be inserted sideways into the boat and then turned
so the punch wedges itself facing forward. Push the punch forward
until it is seated against the front of the slot in the boat.
Push the push rod forward until the boat is located in the idle
zone of the quartz oven. Slide the sample port fitting back
until it is centered over the sample port and tighten firmly by
hand. DO NOT tighten with a wrench. As before, avoid exerting
any sideways pressure on the quartz oven.
Select option 1 of the main menu of the Carbon program. Input the
full sample ID. NOTE: the program will automatically place the
computer into Caps Lock mode. After verifying the sample ID,
enter the run number (1-9). The run number must correspond to the
number of punches removed from the filter. Replicate runs are
designated simply by the appropriate punch number (usually "2").
Note: the program creates a file name using the last six characters
of the sample ID plus the run number; if the program finds -another
file with the same name, it will request that a new run number
be input so the existing file will not be overwritten.
Input the appropriate punch size (normally 0.503 cm2) and filter
deposit area. Note that pressing the return key is not necessary.
Also note that if a mistake is made during input of analysis
data pressing will allow the option to be aborted and
restarted.
When all data is correct, press any key except to start the
analysis program. See Figure 16.
Using a small piece of clear tape, attach the previous sample punch
to its thermogram, insuring that the deposit side is up.
Replace the PetriSlide or petri dish containing the filter into
the styrofoam cooler.
The program will purge the oven with He for 2 minutes, after which
data collection will begin. Readings are collected for 2 minutes
to establish baselines. At 114 seconds, a warning tone will sound
(the number of beeps corresponds to the analyzer number). At 120
seconds the analysis start tone will sound. At that time, push
the push rod in until the stop is against the back fitting. While
watching the punch, pull the push rod back 1-2 mm to physically
decouple the pushrod from the boat. If the boat slides back,
immediately push the thermocouple back in and try again. The boat
cannot be physically attached to the push rod during analysis, as
expansion of the thermocouple as the sample is heated will push
the punch closer to the laser rod and cause erroneous laser signals.
-------
o
I-*
o
o o
c. oo i
V5 CT"
UJ *H
UJ '-
z: at
-------
All physical adjustments must be made within 10 seconds: between
130 and 140 seconds analysis time the laser baseline is calculated.
If the sample is not correctly positioned at the end of 10
seconds, press to abort the program, pull the boat back to
the idle zone, and restart the program. Decoupling the boat is
most important for a meaningful laser signal.
The program will proceed automatically from this point without
further operator intervention. At the end of the program,
data is saved to disk, split times are calculated, carbon peaks
are integrated, and tabular and graphical printouts are produced.
When the printer begins, the push rod may be pulled back to the
idle zone to begin cooling.
*
Examine the thermogram for proper laser response, temperature
profiles, realistic carbon peaks, and the presence of the calibra-
tion peak at the end of the analysis. Examine the tabular printout
to insure the calibration peak counts are within specifications
(see Section 4.2). Finally, examine the laser signal at the end
of the run. Drooping of the laser signal as the temperature
is dropping is an indication that the boat was coupled to the
push rod and that the sample should be rerun. If all aspects of
the analysis appears correct. Select the appropriate analysis
flag from the screen that appears at the end of the run
(Figure 17). Mark the analysis date after the sample on the
sample analysis list. If a problem is found, indicate the problem
in the analyzer log book and rerun the sample.
- Repeat the above steps for additional samples.
4.3.2 System Blanks
System blanks are run each Monday. Follow the steps outlined in
Section 4.3.1 with the following exceptions:
Use option 7 from the main menu to change to the \SYSBLK
subdirectory.
- Go through all the steps for a normal analysis, with the exception
that the punch from the previous analysis is not removed. Open .
the sample port, pull the boat .back into the loading zone, and
without touching the existing punch push the boat forward into
the idle zone, seal the sample port, and proceed with the analysis.
Use an ID number derived from the current date: e.g., SB0718.
Calculated carbon concentrations should not be more that 0.2 ug
carbon. Values greater than this warrant a second system blank.
4.3.3 Carbonate Analysis
- Follow the steps under Section 4.3.1 until the sample punch is
loaded into the boat. Pull the boat BACK until the punch is
centered under the acid injection port, taking care that the
"ears" holding the boat to the push rod are not bent in the process.
Select option 2 from the main menu. Enter the sample ID, run
number, punch size, and filter size. Select the purge option and
start the analysis program.
-------
31
O
O
LU
*— <
LU
tst CO
ZOO
O a
-------
- At 90 seconds elapsed analysis time flush the 25 ul syringe with
0.4 M hydrochloric acid (HC1). When the start tone sounds at
120 seconds elapsed time, eject 20 ul HC1 onto the filter punch,
insuring that the needle bevel is turned toward the punch and
that the needle tip is touching the top of the punch.
- When all analyses are underway, flush the syringe with distilled
water to prevent corrosion of the syringe plunger.
- After the carbonate analysis is completed, a tabular summary and
a. copy of the graph will be printed. The progam will automatically
cycle into the normal OC/EC analysis, using the same sample
ID. Push the sample boat into the punch drying area (about 1 cm
from the first coil of the sample oven). If the sample punch has
tipped over during the carbonate analysis, open the sample port,
reorient the punch, close the port, and proceed with drying the
punch. Heat from the oxidation oven will dry the sample in this
position without, prematurely baking carbon from the sample; the
sample temperature should not exceed 42 C. When the punch appears
to be dry (wait at least 5 minutes), start the OC/EC analysis.
4.4 Analyzer Shut-Down
After the final sample for the day is analyzed, shut down the analyzers
by:
Leave the last analyzed punch in the boat and the boat positioned
in the heating zone.
Select option 3 to begin the calibration gas injection routine.
Follow the injection procedures outlined in Section 4.2, with the
exception that a He only atmosphere is used during the afternoon
check.
- When the analysis is complete, record the calibration peak counts
and calculated injection calibration in the logbook. Any values
outside the ranges defined in Section 4.2 should be investigated
and rerun. Because low values from the end-of-day calibration
could potentially invalidate the entire day's runs, any deviations
from the accepted ranges must be noted and the cause defined.
- After a satisfactory injection concentration is obtained, use
option 7 to change to the directories in which analysis results
were saved during the day. Use option 8 to print a single page
summary of the day's analyses in each directory.
Press to end the Carbon program. This is necessary because
when the Carbon program ends, it sets the analyzer valves such
that oxygen is flowing through the Mn02 catalyst, allowing some
regeneration of the catalyst overnight.
Remove the printouts and attach them to a manila folder labeled
with the date and analyzer number. Place on the lab supervisor's
desk for Level I validation.
- Leave the computers and analyzers on overnight unless the potential
for power outages or surges exists. Turn off the monitors overnight
to reduce the possibility of phosphor burn.
- Make a final check of the gas cylinder pressures to insure that
-------
will be available to check them again.
- Put the samples and blue ice in the styrofoam cooler back into
the sample storage freezer and lock the freezer.
If the 25 ul syringe was used for carbonate analysis, thoroughly
rinse the syringe with distilled water and tightly cap all
solutions.
- Lock the carbon analysis room.
4.5 Abbreviated Operational Checklist
Start-Up:
- Check all gas cylinders' pressures and delivery pressures.
- Check that all FIDs are lit by holding a pair of tweezers over
the FID exhaust stack and watching for condensation. Relight if
necessary.
Check and readjust if necessary all gas flows at the analyzer.
- Turn on the computer monitor.
Insure that the date on the computer is current.
- At the C> prompt, type "CARBON" to begin the carbon program.
Insure that the sample port fitting is tight and that the thermo-
couple push rod is reasonably snug at the back fitting.
From the opening menu, select option 4 to bake the oven for 10
minutes.
Backup the previous day's analysis data.
- Wipe the sample tweezers, petri dishes, and sample punch.
Begin the daily entry in the carbon analyzer logbook.
Insure that the printers have enough paper for the day and that the
ribbon is producing legible printing.
After the ovens have cooled to less than 100 C, perform a leak test,
involving isolating the oven and operating the Carle valve.
Purge the system of calibration gas injected by the above step.
Perform the morning calibration injection by selecting option 3 and
He/02 carrier gas (option 1). When the analysis is complete,
record the calibration peak counts and injection concentration.
Insure that these values are within their proper ranges.
Change to the appropriate subdirectory for the samples to be
analyzed.
Retrieve the samples to be analyzed from the sample freezer.
-------
Routine OC/EC Analysis
- Pull the push rod back to the idle zone of the quartz oven
and allow the boat/push rod to cool to 50 C or less.
Insure that the petri dish, tweezers, and punch are wiped clean.
Based on the analysis list, remove the sample to be analyzed from
the styrofoam cooler.
- Remove a sample punch from the filter.
- Record the filter ID in the analyzer log book, along with any
comments on the condition of the deposit or any other conditions
which might affect analysis results.
- After the boat has cooled to 50 C or less, remove the previous and
load the current punch.
Begin the analysis by selecting option 1 from the main menu of the
Carbon program and inputting the sample ID, run number, punch size,
and filter deposit area.
Push the sample into the heated zone at 120 seconds, insuring that
the boat is not physically coupled to the push rod.
- Using a small piece of clear tape, attach the previous sample punch
to its thermogram, insuring that the deposit side is up.
- Replace the PetriSlide or petri dish containing the filter into
the styrofoam cooler.
At the end of the analysis, the push rod may be pulled back to the
idle zone to begin cooling.
Examine the thermogram for proper laser response, temperature
profiles, realistic carbon peaks, and the presence of the calibra-
tion at the end of the analysis. Examine the tabular printout
to insure the calibration peak counts are within specifications
(see Section 4.2). Finally, examine the laser signal at the end
of the run. Rerun any deviants immediately. Indicate successful
analyses on the sample analysis list.
Repeat the above steps for additional samples.
System Blanks (run first each Monday):
Change to the \SYSBLK subdirectory.
Go through all the steps for a normal analysis, with the exception
that the punch from the previous analysis is not removed. Open
the sample port, pull the boat back into the loading zone, and
without touching the existing punch push the boat forward into
the idle zone, seal the sample port, and proceed with the analysis.
- Use an ID number derived from the current date: e.g., SB0718.
Calculated carbon concentrations should not be more that 0.2 ug
carbon. Values greater than this warrant a second system blank.
-------
Carbonate Analysis:
Follow the steps under Routine Analysis until the sample punch is
loaded into the boat. Pull the boat BACK until the punch is
centered under the acid injection port.
Select option 2 from the main menu. Enter the sample ID, run
number, punch size, and filter size. Select the purge option and
start the analysis program.
- At 120 seconds elapsed time, eject 15 ul HC1 onto the filter punch.
Flush the syringe with distilled water between samples.
- Continue the normal OC/EC analysis when the carbonate cycle is
complete.
Analyzer Shut-Down:
- Leave the last analyzed punch in the boat and the boat positioned
in the heating zone.
Select option 3 to begin the calibration gas injection routine.
Follow the injection procedures outlined in the Start Up section
with the exception that a He only atmosphere is used.
When the analysis is complete, record the calibration peak counts
and calculated injection calibration in the logbook. Any values
outside "the ranges defined in Section 4.2 should be investigated
and rerun.
- Print summaries of the day's analyses.
Remove the printouts and attach them to a manila folder labeled
with the date and analyzer number. Place on the lab supervisor's
desk.
- Turn off the computer monitors
- Make a final check of the gas cylinder pressures.
- Put the samples and blue ice in the styrofoam cooler back into
the sample storage freezer and lock the freezer.
If the 25 ul syringe was used for carbonate analysis, thoroughly
rinse the syringe with distilled water and tightly cap all
solutions.
- Lock the carbon analysis room.
5.0 Quantification
5.1 Calibration procedures
The calibration procedures for the carbon analyzers are of two types:
the end-of-run automatic injection and the manual calibration using
KHP and the two calibration gases. The end-of-run calibration
consists of a set quantity of CH4 calibration which is automatically
injected by the Carbon program. All FID readings during the analysis
run are normalized to this peak to minimize the effects of FID
-------
performance and electronic drift over time. The manual calibration,
performed twice a year or when a new calibration gas cylinder is
started, establishes the calibration slope used in converting counts
to ug of carbon, as explained in the next section.
The end-of-run calibration occurs automatically at the end of the
analysis run and requires no operator intervention. The integrated
calibration peak counts should be checked by the operator immediately
after each run to insure that the analyzer is operating satisfactorily.
The manual calibration involves spiking prefired quartz punches with
various amounts of the 1800 ppm KHP and sucrose solutions
(Section 3.1) and injecting various volumes of the C02 and CH4 gases.
A clean blank quartz punch is baked in the oven at 800 C for 10
minutes using option 4 from the main menu of the carbon program.
After the punch has cooled to less than 50 C, the solution is injected
onto the punch using a 20 ul syringe. The following volumes are
used:
5 ul KHP or sucrose solution
10 ul KHP or sucrose solution
15 ul KHP or sucrose solution (do twice)
20 ul KHP or sucrose solution
no injection (as a system blank)
20 ul acidified DDW only (check of background level of DDW)
The sample port is sealed and the punch is pushed to 1 cm from the
sample oven. In this position the punch will be about 39 C due to
the heat from the methanator oven. Allow the punch to dry thoroughly;
the punch will turn from translucent to opaque as it dries. The
punch must be dry to avoid water vapor effects on the FID. The
carbonate option from the main menu is selected and started. This
yields a two-peak thermogram, including the normal calibration peak
at the end of the run. The integrated peak counts for both the
sample peak and the calibration peak are recorded.
The C02 and CH4 calibrations are also run by using the carbonate
option. The following volumes are injected:
100 ml C02 or CH4 gas
250 ml C02 or CH4 gas
500 ml C02 or CH4 gas
1000 ml C02 or CH4 gas
2000 ml C02 or CH4 gas
Again, the integrated peak counts are extracted manually from the
tabular printouts.
Calibration values are plotted as actual ug carbon vs. the ratio of
the integrated sample peak counts to the calibration peak counts.
Obvious outliers are identified and rerun. Linear regression is
performed on each set of calibration data individually. The
calibration slope derived from the C02 injections typically has a
slightly different slope and does not fit as well. The slope is
calculated from
m
-------
and the standard deviation is calculated by
-------
the data is merged into the final database (Section 7.3).
6.0 Quality Control
6.1 Performance Testing
System blanks are performed at the beginning of each week to
insure the system is not introducing bias in the carbon results and
to insure that the laser signal is not temperature dependent.
Contamination is potentially due to:
Operator practices, such as improper cleaning of tweezers and
jpunch.
- Teflon particles on the push rod are getting into the heated
zone of the quartz oven.
- The sample boat is contaminated.
The carrier gases are contaminated.
A temperature-dependent laser signal is potentially due to:
Physical coupling of the push rod to the boat during the run.
- A quartz rod ready for replacement. Microscopic cracks in the
quartz rod will increase internal reflectance of the laser;
as the number of these cracks multiply, the effect of
temperature on these cracks, and thus on the reflectance,
becomes an interference in the laser signal.
As described in Section 5.1, the calibration peak at the end of each
analysis run serves as a regular standard; the integrated area under
the calibration peak serves as a measure of analyzer performance. In
addition, the injection of two calibration gases daily further serves
as standards. Primary standards in the form of spiked filter punches
do not yet exist.
6 . 2 Reproducibility Testing
Replicates of analyzed samples are performed at the rate of one per
group of ten samples. The replicate is selected randomly and run
immediately after a group of ten is completed. The ug/cm2 values
for OC, EC, and TC are compared with the original run. The values
should fall into the following criteria:
Range Criteria
<= 10 ug/cm2 <= +-1.0 ug/cm2
> 10 ug/cm2 <= 10 % of average of the 2 values
Notice that the criteria merge at 10 ug/cm2. Replicates which do
not fall within the above criteria must be investigated for analyzer
or sample anomalies. Analyzer anomalies include poor response (as
reflected in the calibration peak areas) or poor laser splits.
Typical sample anomalies include inhomogeneous deposits or
contamination during analysis. Inconsistent replicates for which
a reason cannot be found must be rerun again.
-------
6.3 Control Charts and Procedures
Three types of control charts are updated at the beginning of each
week. These charts include a month of data and are posted in the
carbon room until the month is complete, after which they are filed
with the raw analysis results.
The first is a plot of calibration peak counts as percent deviation
from a historical mean versus date (Figure 16). Instances where
the calibration peak area deviates by more than 10% from the
historical mean must be investigated and the cause corrected. The
historical mean covers the previous 3 month's results and is updated
quarterly, when the CH4 calibration gas is changed, or when extensive
repairs are performed.
The charts are created by running a program called CarleQA.EXE. This
program requires a data file called CalibSum.HIS, which contains the
historical means for each carbon analyzer. When the historical means
are updated, the values in CalibSum.HIS are altered using an ASCII
word editor.
The second type of control chart is a plot of calibration gas
calculated concentration versus date (Figure 17). Separate charts
are generated for CH4 and C02 gases. Instances where the calibration
gas concentrations deviate from the historical mean by more than 10%
must be investigated. Most frequently, low calibration gas
concentrations are due to poor injection practices, such as failure
to flush the syringe with the gas prior to withdrawing a sample and
slippage of the plunger during injection into the analyzer.
These charts are created by running a program called SyrinjQA.EXE.
Again, this program requires the CalibSum.HIS data file. The
discussion above applies to the calibration gas historical means as
well.
The third type of control chart is a plot of replicate results. These
charts are generated on a project-by-project basis for the purpose
of visual inspection of analysis precision. Because the precision
of the carbon analyzers is strongly influenced by deposit homogeneity
(Section 1.5), comparison of precision results across projects should
be attempted only with caution.
6.4 Analysis Flags
During Level I validation (see Section 6.5), unusual conditions of
the deposit or analysis problems are noted on the analysis printouts.
Errors in pre-analysis data entry (e.g., in filter ID, punch area,
deposit area) are also noted.
Flags are applied to the dBase file created from the analysis results
ASCII file (see Section 6.5). The following flags are used:
bl ... field blank
b2 ... lab blank
fl ... filter damaged, outside of analysis area
il ... inhomogeneous filter deposit
i3 ... deposit falling off
i4 ... abnormal deposit area, possible air leakage during
-------
sampling
15 ... non-white sample punch after analysis
rl ... first replicate on same analyzer
r2 ... second replicate on same analyzer
r3 ... third replicate on same analyzer
r5 ... replicate on different analyzer
v ... sample void
Note that all results flagged with "v" must include a description
of the reason for invalidating the sample in the remarks field.
6.5 Data Validation and Feedback
6.5.1 Daily Validation
Level I validation is performed by manually checking the tabular and
thermogram printouts the day after the analysis is performed. The
following items are checked on the tabular data (Figure 18):
- the filter ID is correct
for calibration injection runs, the gas type is He/02 in the
morning and He only in the afternoon
the analysis date is correct
the punch area is correct; errors in entry require that the
calculated carbon concentrations be recalculated by hand.
the deposit area is correct; errors in entry require that the
calculated carbon concentrations be recalculated by hand.
the calibration peak area is in the correct range (Section 4.2)
the initial and final FID baseline are within 3 counts of each
other; excessive FID baseline drift is cause for reanalysis
the lower laser split time and the upper laser split time are
within 10 seconds of each other. If the times differ by more
than 10 seconds, check that the lower split OC and upper split
OC differ by no more than 5%. OC values which differ by more
than 5%, unless due to a small change in laser signal resulting
from an extremely clean or very dark sample, requires reanalysis.
calculated carbon values for calibration injection runs are
within 10% of the current mean value for that gas type and that
analyser.
filter acceptance runs result in <1.5 ug/cm2 OC, <0.5 ug/cm2 EC,
and <2.0 ug/cm2 TC. Filters which exceed these levels must be
refired.
Items which are found to be okay are underlined in red. Items which
have problems are circled in red.
The thermograms are checked for the following (Figure 19):
the initial FID baseline is flat, indicating that the analyzer
has been thoroughly purged before analysis began.
-------
Analyser ttl
CARBON CALCULATION RESULTS
Carbon. PAS: P2. 1 02 .'10/89
Carbon.DAT: D2. 1 02/1O/
Analysis ID
Sample ID
Anal ys is
Calculation
307176
02/10/89
02/24/89
15:37
15:01
Anal program ver
Calib. slope
C a 1i b. i ntercept
Baseline window
'Sample transit
Punch area
Deposit area
P2.1 <02/lO/89) Farm file ver :
23.64 ug C/peak-to-cal ibration peak ratio
0.00 uq C
1 counts Laser precision:
26 sec Calib transit :
0.503 cm2
13.8U cm2
02. 1 (02/10/8'
10 counts
50 sec
Calibration peak area: 26jiJ_0__coLints
Initial FID baseline : _iiI7_-_ counts
Final FID baseline : 107 counts
Initial laser : 1790 counts
Laser Split Time
Lower split : l_523__3.ec
Regular split: 1532_ sec
Upper split : 153t sec
OC Peak 81
OC Peak #2
OC Peak #3
Q<. Peak #4
Lower pyro'd OC
Reg. pyro'd OC
Upper pyro'd OC
EC Peak ttl
E< P*^k #2
•E' Peak ft 3
Peak Area
801 counts
1225 counts
5lu2 counts
2701 counts
943 counts
990 counts
1034 counts
20t.2 counts
154t- counts
75 counts
vac ac
Lower spl it : 1.4 r:'. U
Regular split: 1.4 I'-' . 1
I'5. 5 263.7
Upper split : 1.4 1*.2
1-3.5 264.8
Laser FID Split Time
1781 counts 1554 sec
1793 counts 1558 sec
1804 counts 1562 sec
Carbon
0.71 uq C /punch
1.09 uq C /punch
4.53 ug C/ punch
2.40 uq C /punch
0 . 84 uq
0 . 88 ug
0.92 uq
1 . 83 ug
1 . 37 uq
'.>.07 uq
EC
4.8
4. 8
65. fc.
4.7
64.6
C /punch
C/ punch
C/ punch
C/ punch
C/ punch
C /punch
TC
23. '3 Ug C/cm2
329.3 uq C/f liter
23. -3 ug C/cm2
32'3.3 ug C/f ilter
23.9 ug C/cm2
329.3 ug C/f liter
QC/TC:
£'. /TO:
QC/EC:
0.80
0.20
4.02
Figure 18. Example Level I Validated Tabular Printout
-------
E
a
u
no
o
w
a)
j:
H
(0
•o
11
>
-------
the final FID baseline prior to the calibration peak is within
3 pixels of the calculated FID baseline; excessive drift is
cause for reanalysis.
the laser signal during the first 2 minutes appears near the
bottom of the graph (no reflectance); an excessively high
initial laser is an indication that the internal reflectance
of the quartz rod is too high, either due to too many internal
cracks or a complete fracture of the rod. High initial lasers
should result in a physical inspection of the analyzer.
the calculated initial laser line matches the laser signal
immediately after the rod La pushed in. A laser line which
is too low is an indication that the sample was not pushed into
the oven in time; a laser signal which exceeds the calculated
initial laser is a symptom of physical coupling between the
sample boat and the push rod, although some automobile emission
samples also show this characteristic; a spike or a number of
jumps in the laser signal indicates that the operator had
difficulty in decoupling the boat from the push rod. All of
these problems are grounds for reanalysis if severe.
- the laser signal should dip below the initial laser line until
oxygen is introduced at 550 C, at which the signal should rise
steeply.
the laser at the end of the analysis is flat; if the laser signal
dips as the oven begins to cool, the boat is physically coupled
to the push rod and the laser signal during the rest of the
analysis is suspect.
the temperature readings reflect stable temperatures at each
level and smooth, quick transitions between levels.
Problems or deviations from normal should be circled in red. If the
sample punch taped to the thermogram is not white, it is also circled.
If examination of the tabular and thermogram printouts result in
a decision that a sample should be reanalyzed, write "Rerun" in
red on the printouts and prepare a reanalysis list. This list should
be posted immediately after the validation is complete, and those
samples should be rerun as soon as they can be conveniently fit into
the current day's analyses.
Evidence of persistent analyzer problems must be resolved, either by
physically examining the analyzer or reviewing the problems with
the analyzer operator.
6.5.2 Validation of Final Data File
The following steps are followed to create and do Level "I 1/2"
validation on carbon data:
Obtain copies of the latest version of the summary file from the
directory corresponding to the desired project. These files are
called CPeaks.n, where n is the carbon analyzer number. These files
may be either restored from the backup files or copied directly from
the carbon analyzer computers. The latter method is recommended,
from the standpoints of ease of use and of a guarantee that the
summary files retrieved are the latest versions. These files are
-------
updated at the end of each analytical run, so the latest version is
necessary to insure that all of the analyses are included.
Copy the files together using the DOS COPY command:
COPY CPEAKS.* TEMP
REN TEMP CPEAKS.ALL
The first command copies all of the available summary files into one
file called TEMP. The second command renames the TEMP file to
CPEAKS.ALL, the designated name for the combined files. Note that
the two commands cannot be combined; i.e., don't try
COPY CPEAKS.* CPEAKS.ALL
This will not work, because at some point the computer will be trying
to copy CPEAKS.ALL into itself.
The ASCII CPEAKS.ALL file is reformatted, sorted, and placed into a
dBase III file by the following:
DBASE
DO INPCAR
The INPCAR program will prompt the user for the ASCII input file name
(CPEAKS.ALL). It will also ask for an output file name. The
dBASE file naming convention calls for a name in the following format:
xxOEnnt.DBF, where xx is the 2 character project identifier
OE stands for organic/elemental carbon
nn is the 2 digit batch number (generally
used to distinguish between different
projects for the same client or between
sampling quarters for an extended project)
t stands for the sample type:
W = woodstove
A = ambient
P = generic point source
D = diesel emissions
G = gasoline emissions
E = mixed vehicle emissions
F = field/ag burn emissions
X = mixed types
The final dBase file name is specified in the analysis list posted in
the carbon room.
After the INPCAR program produces the dBase output file, the program
will alert the operator that it is ready to print the contents of that
file. NOTE: wide carriage printer is strongly recommended.
After the printout is produced, immediately label the top of the
printout with the file name and printout date.
Begin validation by matching the filters listed on the analysis list
with the filters listed on the dBase printout. There must be at least
one entry on the printout for every filter listed on the analysis list.
Flag field and lab blanks as the list is reviewed by placing "bl" or
"b2" in the second column of the printout. Because the dBase printout
is sorted by ID number, replicates and reruns will be grouped together.
-------
Indicate missing data by writing the missing filter ID in the margin
with an arrow drawn to the appropriate place of insertion. Scan the
printout for unusual IDs which may have been mistyped during analysis.
Generally these will appear at the beginning or end of the printout
due to the sorting process. Make sure also that any samples listed
on a rerun list appear on the printout.
Resolve all missing data. If a large amount of data is missing because
of analysis in the incorrect subdirectory, it is generally easier
to retrieve the summary file from that incorrect subdirectory,
trim the unnecessary data from that file using a word processor,
combine the remaining data with the CPEAKS.ALL file, and rerun the
INPCAR program. If this is done, take care that the errant summary
file, also called CPEAKS.n, does not overwrite the CPEAKS.n that
already exists. If only a few data points are missing, it is
generally not too much trouble to simply write the correct values on
the printout and add those values manually to the dBase files at the
same time the flags and other corrections are made.
Scan the deposit areas for incorrect entries. Circle the incorrect
entries to insure that corrected values replace those currently in
the database.
Scan the filter IDs for multiple entries of ID numbers. Under normal
conditions, the only times multiple entries should occur are reruns
and replicates. All multiple entries must be flagged to indicate
the reason for their existence. If no flags appear, draw an underline
to act as a reminder to look for the reason as the files are reviewed.
Scan for missing runs. The most common example is the first run being
aborted or lost for some reason, and the only entry in the dBase file
is the second run. An entry for the first run must be inserted,
flagged as invalid, and labelled as to the reason it was invalid. All
punches taken from the filters MUST be accounted for.
Scan the OC and EC columns looking for unusually high or low values.
At this time make sure that the field blanks and/or lab blanks are
all close to one another. Circle any possible outliers for further
investigation.
Finally, pull the analysis files and go through the analysis summaries
and thermograms one by one. At this time, resolve all circled items
and all missing flags. Determine if analyses flagged by the operator
as "SL" or other such flags are legitimate. If not, draw a line
through the flag to indicate it should be removed. If the sample
should be rerun, add it to a rerun list. If the analysis has some
anomaly but still appears to be legitimate, either flag or add notes
to the comments field as appropriate. Analysis flags are defined in
Section 6.4. All samples flagged as invalid must have an entry in
the comments field to describe why the sample is invalid. The
following notes and comments are commonly used:
Comments Description
"Anomalous laser" Despite good initial laser, laser signal
drifted above initial laser signal
before dropping (typical of auto
emissions)
"Operator error" Used with "v" flag; operator missed
-------
pushing boat in, pushed abort button,
pushed manual advance button at an
inappropriate time, etc.
"Analyzer malfunction" Used with "v" flag; analyzer
malfunction or problem beyond the
control of the operator such as plugged
FID, broken oven heater, etc.
"Poor replicate" Replicate is outside the normal
criteria, but no reason can be found
for the discrepancy.
"Poor initial laser" Used with "v" flag; severe coupling or
boat not pushed in time for calculation
of initial laser signal.
"Sample contaminated" Used with "v" flag} rerun of sample
yields lower values or different peaks.
Typically used with blanks or reruns of
replicates.
All flags generated during the analysis must be either converted to
the flags and/or comments listed above or removed. These flags are
temporary flags only and are not recognized as legitimate analysis
flags at DRI.
After all thermograms have been reviewed and all possible reruns have
been identified, post the rerun list in the carbon room and have the
reruns done as soon as possible. Review the data from the reruns,
looking for inconsistencies. Insure that the reasons for the rerun
have been addressed. Mark the printout with the new values for manual
insertion into the dBase file. Previous runs must be flagged as
invalid or the reruns flagged as replicates.
Finally, all comments, flags, insertions, and other changes made to
the printout are entered into the dBase file. After all changes are
made, generate a new printout. Label the new printout with the file
name and printout date. Forward a copy of the printout and the dBase
file on disk to the person putting the final report together.
-------
IMPROVE SOP APPEND JULY 1989
APPENDIX 8: Maintenance Checklist (Annual Site Visit)
-------
IMPROVE SOP APPEND JULY 1989
Maintenance Checklist.
clean repair replace comments
always if if
defective defective
PUMPHOUSE
Enclosure X
Temperature switch X
Contractor box relays X
Contractor box wiring X
Cooling fans X
Pump: capacitor X
Pump: diaphram X
Pump: piston X
Pump: flapper valve X
Pump: gasket X
Pump: sponge filters X
CONTROLLER
Enclosure X
Cooling fans X
Control clock X
Fuse X
Relays X
Delay switch X
Transformer X
30-Minute bypass switch X
Wiring X
FILTER MODULE A
Enclosure X
Solenoid valves X
Magnehelic X
Vacuum gauge X
Hoses X X
Fittings X
Elapsed time meters X
Toggle switches X
Inlet X
Cyclone X
Electrical connections X
Wiring X
Cyclone 0-rings X X
Nuts and bolts (tighten) X
-------
IMPROVE SOP APPEND JULY 1989
clean repair replace comments
always if if
defective defective
FILTER MODULE B
Enclosure X
Solenoid valves X
Magnehelic X
Vacuum gauge X
Hoses X X
Fittings X
Elapsed time meters X
Toggle switches X
Inlet : X
Denuder (Module B) X
Cyclone X
Electrical connections X
Wiring X
Cyclone 0-rings X X
Nuts and bolts (tighten) X
FILTER MODULE C
Enclosure X
Solenoid valves X
Magnehelic X
Vacuum gauge X
Hoses X X
Fittings X
Elapsed time meters X
Toggle switches X
Inlet X
Cyclone X
Electrical connections X
Wiring • X
Cyclone 0-rings X X
Nuts and bolts (tighten) X
FILTER MODULE D
Enclosure X
Solenoid valves X
Magnehelic X
Vacuum gauge X
Hoses X X
Fittings X
Elapsed time meters X
Toggle switches X
PM10 Inlet X
Cassette manifold X
Electrical connections X
Wiring X
0-rings X X
Nuts and bolts (tighten) X
STAND
Structure X
Sunshield and bench X
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1. REPORT NO.
EPA-450/4-9Q-008a
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
IMPROVE Progress Report
Appendix A
5. REPORT DATE
May 1990
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Marc Pitchford
David Joseph
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Monitoring Systems Laboratory
U. S. Environmental Protection Agency
Las Vegas, Nevada 93478
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
In Section 169A of the Clean Air Act as amended August 1977, Congress declared
as a national goal "the prevention of any future, and the remedying of any existing,
impairment of visibility in mandatory class I Federal areas which impairment results
from manmade air pollution."1 Mandatory class I Federal areas are national parks
greater in size than 6000 acres, wilderness areas greater in size than 5000 acres and
international parks that were in existence on August 7, 1977.2 This section required
the Environmental Protection Agency (EPA) to promulgate regulations requiring States
to develop programs in their State Implementation Plans (SIPs) providing for visi-
bility protection in these areas. EPA promulgated these regulations on December 2,
1980.3
This report summarizes the progress made to date in developing and implementing
the interagency monitoring network which supports the effort, Interagency Monitoring
of Protected Visual Environments (IMPROVE).
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Visibility monitoring
State Implementation Plans (SIP)
Class I Federal Areas
18. DISTRIBUTION STATEMENT
EPA rorm 2220-i rRev. 4-77)
19. SECURITY CLASS (This Report/
21. NO. OF PAGES
20. SECURITY CLASS (Tins page/
22 PRICE
PREVIOUS S1D1TION IS OBSOLETE
-------
------- |