HomeMy WebLinkAbout2011-07-13 PACKET 08.2.B.SEH MEMORANDUM
TO: Jennifer Levitt, PE, City of Cottage Grove
FROM: Thomas Henning, PE
DATE: June 30, 2011
RE: Review of Ambient Monitoring Data and 3M Compliance Data
SEH No. 109553 001
At your request Short Elliott Hendrickson, Inc. (SEH) has reviewed the results of ambient air monitoring
conducted at the Cottage Grove air monitoring station and evaluated trends in compliance data from the
3M Cottage Grove facility. SEH also obtained the incinerator operating records from 3M and conclude
that the incinerator has operated nearly continually from January 1, 2011 through April 1, 2011. It was
operating during each of the sampling dates over this period.
Review of Air Monitoring Data
SEH was asked to assess the results of monitoring for three compounds: dichlorodifluoromethane (Freon
12), hexane, 2- propanol, and methyl ethyl ketone (MEK).
Freon 12 — The issue related to Freon 12 is its frequent detection at the monitor. It has been measured
above the detection limit in 11 of the 14 samples collected. The monitored concentrations have ranged
from 1.6 to 24.0 micrograms per cubic meter (ug/m All the measurements have been below the Freon
12 health benchmark of 200 ug/m Freon 12 is regularly detected because the ambient concentration is
above the typical laboratory detection limit — about 1.8 ug/m The Twin City's average concentration
during 2009 measured by the Minnesota Pollution Control Agency (MPCA) was about 2.9 ug /m Freon
12 is not expected to be emitted from the 3M incinerator.
Hexane — Hexane has been detected in the Cottage Grove monitor in 10 of the 14 samples. The
monitored concentrations have ranged from non - detect to 24.8 ug/m All these concentrations are less
than the health benchmark of 2,000 ug/m During 2009, MPCA recorded concentrations ranging from
non - detect to 94.3 ug /m in monitors located across the state. Hexane is a common laboratory solvent so
it could be a contaminant in a sample. Hexane could also be used at one of the 3M facilities located at the
3M Cottage Grove Complex. We note that the highest concentration, 24.8 ug /m was monitored on
February 14, 2011. Monitored wind direction on that day was from the west so the 3M facility did not
contribute to the 24.8 ug/m reading (the plant is located south east of the monitor).
2- ProRanol Concentrations of 2- propanol have ranged from non - detect to 18.2 ug/m Eleven samples
have had non - detect concentrations of 2- propanol. The MPCA has monitored a wide range of 2- propanol
concentrations, from non - detect to 180 ug /m We note that the highest 2- propanol concentration was
monitored on February 14, 2011 —the same day as the highest hexane concentration and a day when the
wind was blowing toward the 3M plant.
Methyl Ethyl Ketone — MEK has been detected in 10 of the 14 samples collected. The concentrations
have ranged from non - detect to 102 ug /m (the second highest concentration is 10.0 ug /m The
concentrations are less than the health benchmark of 5,000 ug /m Like hexane, MEK is a common
Short Elliott Hendrickson Inc., 3535 Vadnais Center Drive, St. Paul, MN 55110 -5196
SEH is an equal opportunity employer I www,sehinc.com 1 651,490 2000 1 800.325.2055 1 651.490.2150 fax
Review of Ambient Monitoring Data and 3M Compliance Data
June 30, 2011
Page 2
laboratory solvent so it is prone to contaminate ambient samples. It may also be used at the 3M plant.
The 102 ug/m sample was collected on December 28, 2010, a day when the wind was primarily from the
south and east (i.e. from the plant).
Review of Reported Deviations at 3M Cottage Grove incinerator
SEH was asked to review data provided to the City by others. The data was an analysis of the number of
deviations reported at the 3M Cottage Grove Corporate Incinerator from 2006 through 2010. 3M is
required by their air pen to provide semiannual monitoring reports which include descriptions of
deviations. In the evaluation, the number of deviations were counted for each year and divided by the
operating hours of the incinerator for that year. The data includes graphs of the results of the analysis
with linear regressions suggesting a trend in the data. The conclusion of the analysis is that the frequency
of reported deviations per thousand hours of incinerator operation is increasing.
SEH reviewed the data analysis and the 3M monitoring reports for the years 2007, 2008 and 2010 (i.e. the
years that SEH had complete 3M monitoring reports) and provide the following comments:
1) Many the trends reported are a result of the reduced hours of operation of the incinerator
rather than a significant increase in deviations. For example, there is no significant difference
in the number of "CO Deviations > 1,000 ppm" over this five year period (data = 1, 0, 1, 2, 1).
Yet, when divided by the number of operating hours per year an apparent trend develops. Another
example is the "Record Keeping" deviations. The reported number of these deviations (data = 0,
4, 3, 5, 3) do not show a trend. But when divided by the operating hours per year, an apparent
trend develops. Each of these damsels was tested by SEH using the Mann- Kendall statistical test.
The Mann - Kendall testis widely used in environmental science as a non - parametric test for the
detection of trends in a time series of data. The test results indicate that there are no trends in the
data at both the 80% and 90% confidence level and the data presents a stable (not increasing or
decreasing) pattern.
2) The counting of the deviations is relatively suhjeetive. SEH was able to reproduce the number
of deviations identified for each category in the 2010 monitoring report. However, using the
same counting method, we were not able to reproduce the number of deviations in the analysis
from the 2007 and 2008 reports. For example, we counted a different number of "Other
Deviations" and "Record Keeping" deviations in both 2007 and 2008 than reported in the
evaluation. Given this subjectiveness, a trend based on this data is questionable.
3) The analysis does not take into account the duration each deviation. In 2007 most of the
"Other Deviations" reported have very short durations (less than about 4 seconds) yet are
weighted equally with other deviations which may have lasted for a much longer time. In
addition, there is a wide variation of the length time for each "CO Deviation ". To count each
deviation equally may skew the data.
In conclusion we do not agree that this data supports the conclusion that the number of deviations is
significantly increasing with time. We are not providing an opinion of whether the number of deviations
is acceptable or not, rather only commenting on the change of deviations with time.