2001 Annual PM2.5 Network Review for Michigan

Michigan Department of Environmental Quality

Air Quality Division

Prepared by Mary Ann Heindorf

June 27, 2001

Table of Contents

Item			Page

1.0 
Purpose………………………………………………………
…..……………...1

	

2.0  Monitoring Objectives	1

3.0  Network Design & Deployment Status	1

	A.  PM2.5 FRM Network Design	2

		FRM Deployment Status	4

	B.  Continuous Network Design	4

		Continuous Monitor Deployment Status	4

	C.  Speciation Network Design	6

		Speciation Deployment Status	7

	

4.0  Proposed Changes	8

	A.  FRM Network Proposed Changes	8

	B.  Continuous Network Proposed Changes	9

	C.  Speciation Network Proposed Changes	9

	D.  PM10 Network Proposed Changes	9

5.0  Fine Particulate Data Summary	10

	A.  Data Completeness	10

	B.  Annual Averages	14

	C.  24-Hour Values	15

6.0  Quality Assurance/Quality Control	15

	A.  Co-located Sites	15	B.  Precision	18

	C.  PEP Audit Results	19

List of Figures

Figure #	Title									Page

Figure 1:	Current PM2.5 Monitoring Network 	2

Figure 2:	Current Continuous PM2.5 Monitoring Network	5

Figure 3:	Proposed PM2.5 Speciation Network	7

Figure 4:	Optimal PM2.5 Monitoring Network	8

Figure 5:	Optimal PM10 Monitoring Network	10

List of Tables

Table #	Title									Page

Table 1:	Michigan PM2.5 Network Summary FY 2001	3

Table 2:	Data Completeness and Quarterly Averages of Fine

	Particulate Material in Michigan	11

Table 3:	98th Percentile PM2.5 Values Averaged Over 3 Years	16

Table 4:	Locations of Co-Located PM2.5 Monitors	17

Table 5:	Percent Differences for Co-Located Measurements 

	Greater than 6.0 ug/m3	18

Table 6:	Mass Comparisons Between R & P FRM and Met-One

	Speciation Monitors	19

Table 7:	PEP Audit Results 1999 - 2000	20

1.0  Purpose:

Annual Review Requirements

According to 40 CFR part 58.26 an annual state fine particulate (PM2.5)
air monitoring report must be submitted to the regional administrator by
July 1, that details the “proposed changes to the PM Monitoring
Network Description and be in accordance with the annual network review
requirements in § 58.25.  Specific details about the contents of the
annual report are found in previous versions of this report. Additional
details about network design are contained in the PM Monitoring Network
Description and the annual reviews from previous years.

2.0 Monitoring Objectives:

The objectives of the PM2.5 monitoring program in Michigan are to:

	(	Elucidate levels of fine particulate material at population-oriented
sites in major population centers in Michigan

	(	Provide real time concentrations of fine particulates in major
metropolitan areas

	(	Generate data to determine long term trends

	(	Measure PM2.5 levels being transported into Michigan

	(	Determine transnational transport

	(	Assess background levels of fine particulate material

	(	Collect data to assess compliance with 24-hour and annual NAAQS

	(	Investigate spatial and temporal trends; assess homogeneity

	(	Determine the impact of sources on areas expected to have the highest
levels of fine particulates

	(	Determine the chemical composition of fine particulate material

3.0  Network Design & Deployment Status

The total statewide particulate network consists of multiple components
which together provide a complete picture of the nature of particulate
material within the state.  The concentrations of PM2.5 measured over a
24-hour time period are determined using the federal reference method. 
Shorter-term measurements, of either PM2.5 or PM10 are updated on an
hourly basis by sites in the continuous network.  Chemical composition
of fine particles is determined by the speciation network and by a more
limited chemical analysis of the filters generated by the federal
reference method (FRM) network. This section summarizes the design
components of the fine particulate network and deployment status but
omits a discussion of the PM10 network, which will be considered in the
annual NAAQS network review, scheduled for fall 2001.

All components of the monitoring networks within Region 5 are currently
subject to a regional review.  A “top down” approach is being taken
by EPA staff to assess impacts of removal of redundant monitors.
Simultaneously, a “bottom up” approach is being considered by state
and local staff within the region that will allow the design of an
optimal monitoring network, without restraints imposed by federal
monitoring requirements.  These network reviews are being conducted in
addition to those normally required by federal regulations. Michigan’s
2001 Annual PM2.5 Report merely acknowledges the existence of this
parallel effort.  Many elements of the “bottom up” and “top
down” reviews may not consider the Federal monitoring requirements and
are not included in this report. At this time, it is too premature to
integrate the results from the “bottom up” and “top down”
reviews. Only the preliminary “bottom up results are included in this
report.  Also, other elements such as possible reductions in sampling
frequency are suggested due to support from federal guidelines.

A.  PM2.5 FRM Network Design:

The Rupprecht and Pattashnick (R&P) sequential PM2.5 sampler has been
deployed at the 25 sites that make up Michigan’s PM2.5 FRM network,
shown in Figure 1.  The two tribal sites in Sault Ste. Marie are also
using R & P monitors. 

No boundary changes are proposed for the monitoring planning areas. 
Spatial averaging will not be used in Michigan. 

FRM Deployment Status:

The FRM network is fully deployed, including the most recent additions
of the tribal sites in Sault Ste. Marie.  Monitoring site information is
summarized in Table 1. 

Continuous Network Design:

The MDEQ and WCAQMD are in the process of deploying the continuous PM2.5
network shown in Figure 2.  The Allen Park population-oriented site in
Wayne County meets the requirement for continuous monitoring in MSAs
with population levels greater than 1 million population. A second
continuous PM2.5 monitoring site in Wayne County, which was planned for
Dearborn (a maximum impact site), was eliminated due to insufficient
funding.  The Flint continuous monitor is still pending.

A continuous monitor is planned by the Ontario Ministry of the
Environment in Sault Ste. Marie as part of a tribal monitoring project
at Lake Superior State University (261330901).  The expected deployment
date is around June 15, 2001.  The monitor will be operational through
2002.

All the PM2.5 continuous monitors currently deployed are Rupprecht and
Pattashnick (R&P) Tapered Element Oscillating Microbalance (TEOM)
instruments.  They are being operated at 50 degrees C without a naphion
dryer. If guidance is issued to support operation of the samplers at 30
degrees  with naphion dryers, modifications will be made to the network
contingent upon securing adequate levels of funding.

Continuous Monitor Deployment Status:

Six of the seven sites planned to be operated by state and local
agencies are currently operational. Currently, a continuous PM10 TEOM is
operating at Dearborn. The MDEQ has one additional TEOM which is
awaiting deployment to the Flint site. Another continuous site was
planned for either Holland or Muskegon to provide supplementary
information about levels of fine particulate in disjoint areas of the
Grand Rapids/Muskegon/Holland MSA.  However, given the relatively
homogeneous nature of the PM2.5 levels generated by the FRM network
within the Grand Rapids MSA, the MDEQ is removing this second site from
the continuous network in the Grand Rapids MSA. 

The date when the Flint continuous PM2.5 monitor may become operational
is uncertain. The current goal is to start up the monitor by 1/1/02.

Speciation Network Design:

The primary monitoring objective of Michigan’s speciation network is
to provide data that will ultimately safe guard populations in the
state. Most of the sites shown in Figure 3 are population oriented
stations that will provide information to support the development of
attainment strategies.  Speciation data from the monitors in Kalamazoo,
Ypsilanti and Detroit will allow the creation of attainment strategies
for these areas not likely to meet the annual standard. A speciation
site is also planned for Grand Rapids, a densely populated area very
close to the annual standard. . The figure also reflects changes in
Michigan’s network, that were the result of regional network design
activities.

A speciation monitor is planned to start monitoring on 1/1/02 at
Houghton Lake to provide background information for Regional visibility
modeling.   Sites to address transport and more rural concentrations of
speciated PM2.5 are included at Luna Pier and Holland.

The Ypsilanti site is also located in an area with elevated asthma
levels. Operation of this site is a cooperative venture between MDEQ and
the Washtenaw County Health Department.   The data will assist Washtenaw
County in their investigation of elevated asthma cases in Ypsilanti.

Allen Park is the trend site for the Detroit MSA, due to the
population-oriented nature of the site.  It has been operational since
January 1, 2001. A Met-One single event monitor was deployed at the
trend site.

To supplement the Detroit Pilot Project database, 2 speciation monitors
have temporarily been deployed at E. 7 Mile (261630019) and at the
intersection of 696, the Lodge Freeway and Telegraph Road (261250010). 
Speciated data at these two sites will allow comparisons to be made
between levels of elemental and organic carbon and emissions from motor
vehicles and diesel exhaust, as well as provide comparisons between data
collected in California with other areas.  After the Pilot Project is
over, these two monitors will be moved to Dearborn and to another
population oriented site outside of the Detroit Metro area.  Dearborn is
proposed as a population-oriented maximum impact site due to the
plethora of facilities whose emissions impact on the station.  Both of
these monitors are single event Met-One models.

A tribal speciation monitor will be deployed at Sault Ste. Marie (Lake
Superior University, 260330901) to address concerns about emissions from
Algoma Steel in Ontario, Canada.  Two IMPROVE sites at Seney Wildlife
Refuge (26153xxxx) and Eagle Harbor (the surrogate site for Isle Royale,
26083xxxx)  became operational during December, 2000.

No co-located speciation sites are planned in Michigan.

Speciation Monitor Deployment Status:

TREND Site:

The trend site at Allen Park (261630001) became operational January 1,
2001.

IMPROVE Network

Both IMPROVE sites at Seney and at Eagle Harbor, which is a surrogate
for Isle Royale, became operational during December, 2000.

Tribal Monitors:

The Sault Ste. Marie monitor became operational during June 2001.

May 2002

Upon completion of the Detroit Pilot Study in April 2002, two speciation
monitors will become available for deployment at Luna Pier and Dearborn.

4.0 Proposed Changes

The following minimal changes are proposed to the fine particulate
networks. 

FRM Network Proposed Changes:

Both the bottom up and top down network reviews may impact upon the
design of the FRM PM2.5 network.  Figure 4 shows the results of the
bottom up network review, which is included here as a preliminary
proposal of modifications to the network.

As a result of possible changes in the network, the proposed NAMS sites
are also modified.  Proposed NAMS stations remain in the major urban
centers of the state that include: Grand Rapids, Kalamazoo, Lansing, Oak
Park, Ann Arbor and Saginaw OR Bay City.  The number of NAMS within
Detroit is reduced to Allen Park, to represent the urban area and
Livonia, to represent the more suburban community monitoring zone. 

The following changes in sampling frequency for the FRM network are
proposed.

All sites sampling every day will reduce the sampling frequency to once
every 3 days beginning January 1, 2002.

All sites sampling on a once every three day schedule will reduce their
frequency to once every 6 days beginning January 1, 2002.

All co-located precision monitors will continue to sample once every 12
days.

Continuous Network Proposed Changes:

The TEOM at either Holland or Muskegon will not be deployed because the
FRM PM2.5 measurements are homogeneous throughout the Grand
Rapids/Muskegon/Holland MSA.  There is an existing TEOM at the Grand
Rapids station.

A PM2.5 TEOM may be deployed, eventually,  at Port Huron, contingent
upon additional funding.

Speciation Network Proposed Changes:

Figure 3 illustrates the results form the regional PM2.5 speciation
network design efforts.  Michigan will continue to participate in the
Regional efforts.

Michigan currently has 3 speciation monitors running, all in the Detroit
MSA.  The trend site is located at Allen Park. Two special purpose
monitors are located at E. 7 Mile and 696.  A 4th monitor will become
operational at the rural Houghton Lake site on 1/1/02.  Purchasing and
deploying instrumentation for the remaining (5) sites is delayed pending
the results of continuous speciation monitor study taking place in
Chicago. 

PM10 Network Proposed Changes:

Figure 5 presents the results from the bottom up PM10 review. 
Historically, Dearborn has had the highest PM10 levels in Michigan, and
hence, the site should be retained but at the reduced sampling
frequency. A co-located PM10 monitor will continue to be operated at
Dearborn to supply precision measurements.  We are proposing to convert
the PM10 TEOM into a PM2.5 TEOM, due to relative magnitude of PM2.5. 
The South West High School site is retained as a second NAMS station in
Detroit. 

Fine Particulate Data Summary

The following section summarizes PM2.5 values and the degree of data
capture by for each monitor since its start up date.  Only the first
quarter data for 2001 is available at this time.

Data Completeness

Table 2 illustrates the data capture by monitoring site. The
calculations in the table are limited to each individual monitor; no
data substitution to correct for missing data has been performed. 

All State and Local monitoring stations achieved at least 80% data
capture for the first quarter 2001.  Less than 75% complete data was
captured during 2000 at Traverse City during the fourth quarter when the
filters were destroyed. The incomplete data collected at Oak Park during
the third quarter was due to 

confusion about the correct dates various filters actually sampled. 

The Ann Arbor monitor collected data that was less than 75% complete
because the site was temporarily closed from March 25, 2000 through May
3, 2000 while the roof was reshingled and asphalt was applied to the
parking lot.  These renovations required either removal of the monitor,
which was located on the roof, or caused restricted access to the
building.

The Wyandotte monitor has experienced poor data collection due to power
supply problems.  Originally, the monitor was connected to the ground
fault interrupter by a 25 foot long extension cord, which created
electrical problems. On March 1, 2000, underground wiring and conduit
was added to the site, resolving the problem.

Site renovations occurred at the Ypsilanti site when a fence was
relocated to allow the addition of a trailer during January 2000.

The monitor at Port Huron experienced temperature calibration problems
and was repaired during the first quarter. 

Annual Averages

The data in Table 2 show that most of the sites in Southeast Michigan
had a 3-year annual average greater than 15 ug/m3, using only the first
quarter data to represent 2001. Because the first quarter tends to be
elevated at most sites in Michigan, this creates a positive bias in the
estimation of the three year annual averages for Grand Rapids and
Kalamazoo.  Although the average is greater than 15 ug/m3, after 3 other
quarters of data are collected, the averages will most likely drop below
15, as shown by the trends from 1999 and 2000. The three year average
calculated for Luna Pier is 14.7 ug/m3 and is probably biased low, due
to the limited number of samples (5) collected during 1999. After three
years of data are collected, Luna Pier may exceed the standard, based on
a single complete year of data that has been collected so far.

Three sites, Oak Park, Ypsilanti, E. 7 Mile had annual averages between
14 and 14.9. ug/m3.  The averages at most of the remaining sites were
within 2 ug/m3 of 15 ug/m3. 

During the first quarter of 2001, as in every year PM2.5 has been
measured, all sites experienced elevated fine particulate levels, with
the statewide average reaching 15.47 ug/m3. During the same time
interval in 2000, the average equaled 16.30 ug/m3. Sites in Southeast
Michigan continue to be the highest in the state.  The highest first
quarter average during 2001 was measured in Wyandotte which reached 21.2
ug/m3.  Allen Park located close by, reached 20.41 ug/m3.  Dearborn, 
further upwind from this area experienced elevated levels that reached
20.95 ug/m3 in Allen Park and 19.3 ug/m3 in Wyandotte.  The rural,
upwind site in Luna Pier, which was setup to measure concentrations
coming into Detroit, averaged 16.9 ug/m3 during the first quarter.

Other metropolitan areas in Michigan also experienced first quarter
averages that exceeded 15 ug/m3.  Both Flint and Lansing reached an
average of 15.3 ug/m3.  Grand Rapids was slightly higher with an average
of 16.3 ug/m3.  Nearby, Jenison equaled 17.0 ug/m3.  Kalamazoo was
higher with an average of 19.3 ug/m3.  Yet, sites near the Michigan
lakeshore measured average particulate values less than 14 ug/m3. 

The data show that the annual average is the controlling form of the
standard in Michigan.  None of the annual values shown in Table 2 have
undergone spatial averaging because Michigan is not exercising that
option. 

24-Hour Values

The 98th percentile values of the 24-hour measurements are shown in
Table 3.  As shown by the table, none of the sites have a 98th
percentile value greater than 65 ug/m3, nor are any sites within 80% of
this value.  However, the three year average for five of the sites, and
one co-located monitor in Southeast Michigan, are within 60% of 65 ug/m3
and have a medium likelihood of exceeding a 24-hour standard if it were
set at this level.  The remaining 14 sites are less than  60% of 65
ug/m3, and hence have little likelihood of exceeding a standard set at
this level.  

Quality Assurance/Quality Control

Specific protocols regarding quality assurance, laboratory analysis
sample handling, collection, transport, laboratory sample identification
and archive procedures are covered in the Quality Assurance Project Plan
for the PM2.5 Program in Michigan.  The following section describes the
locations that contain duplicate monitors and the precision results. 

Co-Located Sites

Table 4 lists the sites that have duplicate R&P sequential monitors. The
co-located monitors that were previously located in Saginaw and Muskegon
were moved to Ypsilanti and Port Huron during January 2001.  A better
assessment of monitor precision can be obtained if a greater proportion
of the duplicate values are above 6 ug/m3.  The PM2.5 concentrations at
Saginaw and Muskegon are relatively low so the co-located monitors were
moved to locations with higher levels. This will provide a better
estimate of precision within the network. As shown by the table, the
portion of the network operated by the MDEQ has 27.8% co-located sites
and WCAQMD has 14.3% co-located sites.  In the original network design,
an additional monitor was planned for the Linwood site in Wayne County,
which would have made the percent co-location equal to 28.5%, but the
allocation was cut.  If a co-located monitor in the MDEQ network were
shifted to Wayne County, the 

percent co-location for MDEQ would drop to 22.2% and Wayne County’s
percent would rise to 28.9%. 

Precision

Table 5 shows the percent difference between the co-located pairs of FRM
monitors from 1999 through the first quarter of 2001 for those data
points with a concentration greater than 6.0 ug/m3. During 1999, 25
pairs of values  had a percent difference greater than |10|. There was a
marked improvement in 2000, when the number of occasions were reduced to
16.

As shown in Table 6, comparisons are made between the FRM mass
measurements and the new speciation monitors, using the rather limited
database generated by the Met One speciation instruments. The
measurements are not very comparable, with the Met-One tending to be
higher than the R&P. 

PEP Audit Results

The audit results received from USEPA, as of June 8, 2000, are shown in
Tables 7 and 8 for MDEQ and Wayne County respectively.

  The actual name is Washtenaw County Department of Environment &
Infrastructure Services, Environmental Health Division.

2001 PM2.5 Annual Report for Michigan

 PAGE   vi 

 PAGE   vi 

2001 PM2.5 Annual Report for Michigan

 PAGE   iii 

2001 PM2.5 Annual Report for Michigan

 PAGE   17 

	

 PAGE   20 

