NVH-SQ Research Group, University of Windsor, 401 Sunset Avenue, Windsor, Ontario N9B 3P4

Tel.: (519) 253-3000 x 2534 ! Fax: (519) 973-7050 ! www.uwindsor.ca

Final Report


Prepared for:

The Department of Foreign Affairs and International Trade
Foreign Affairs, Trade and Development Canada
125 Sussex Drive
Ottawa, Ontario
K1A 0G2
Canada

Prepared By:

Principal Investigator
Dr. Colin Novak

Graduate Researchers
Jeremy Charbonneau, Ph.D. Candidate
Peter DAngela, M.A.Sc. Candidate



January 17, 2014

INVESTIGATION OF THE WINDSOR HUM


Windsor, Ontario



Final Report
Investigation of the Windsor Hum: Acoustical Assessment i
EXECUTIVE SUMMARY
Since early 2011, residents of the City of Windsor and nearby communities have reported an
intermittent low frequency noise, which has been labeled as the Windsor Hum. A previous
study confirmed the existence of the low frequency excitation and estimated the source to be in
the vicinity of Zug Island; a highly industrialized island located on the US side of the Detroit
River.
A subsequent investigation by the University of Western Ontario (UWO) and the University of
Windsor was supported by the Canadian Department of Foreign Affairs and International Trade
(DFAIT) to try to locate the source of the Hum. The research conducted by the University of
Windsors NVH-SQ Research Group was comprised of two phases. The first phase was to
identify and characterize the acoustic signature of the Hum using noise monitoring equipment
deployed throughout the Windsor-Essex community. Once the Hum was measured and
characterized, the second phase was to determine the exact source of the noise using noise
source identification (NSI) hardware and software using a beamforming algorithm.
This report includes an in-depth discussion of possible sources capable of producing the
reported Hum sound. The results of the literature study compared the potential sources to the
characterised sounds measured in the first phase of the project. For the duration of the project,
the Hum manifested on only a handful of days, which made the identification of the source
difficult. Good data representative of the Hum was measured using the stationed noise
monitors. Conclusive evidence of the source was not achieved using the NSI system since the
Hum was not present on those days that this equipment was deployed on the river.
The conclusion of the research is that the Windsor Hum does exist and has both qualitative and
quantitative characteristics that surmise the likely source of the Hum to be from the blast
furnace operations on Zug Island. It is recommended that community noise monitoring be
continued and that efforts be given to allow NSI measurements to be made on Zug Island lands
during the periods that the monitors identify occurrence of the Hum sound



Final Report
Investigation of the Windsor Hum: Acoustical Assessment ii
TABLE OF CONTENTS

Page
EXECUTIVE SUMMARY I
1. INTRODUCTION 1
1.1 The Hum Worldwide 1
1.2 Impact on Residents 2
1.3 The Windsor Hum 2
1.3.1 NRCan Investigation 3
1.3.2 University Research 3
2. RESEARCH APPROACH AND METHODOLOGY UNIVERSITY OF WINDSOR
5
2.1 Phase 1 Part 1: Long Term Noise Monitoring 5
2.2 Phase 1 Part 2: High Resolution Data Collection 7
2.3 Phase 2: Noise Source Identification using Beamforming 8
3. RESULTS 10
3.1 Phase 1 Part 1: Long Term Noise Monitoring Results 10
3.2 Phase 1 Part 2: Results of High Resolution Data Collection 11
3.3 Phase 2: Results of Noise Source Identification Measurements 13
4. DISCUSSION 16
4.1 Zug Island Description 16
4.2 Blast Furnace Operation Overview 16
4.3 Potential Noise Sources 18
4.3.1 Blast Mains and Ducts, and Tuyeres 19
4.3.2 Air Discharge Valves (Snort Valves) 20
4.3.3 Effluent Stack Burn Off from Hot-Blast Stove 21
4.3.4 Boiler Facility Release 21
4.3.5 Raw Material Screening Process 22
5. CONCLUSIONS AND RECOMMENDATIONS 23
6. BIBLIOGRAPHY 24



Final Report
Investigation of the Windsor Hum: Acoustical Assessment iii

APPENDICIES

Appendix A: NRCan and UWO Data
Appendix B: Product Sheets
Appendix C: Phase 1 Part 1
Appendix D: Phase 1 Part 2
Appendix E: Phase 2
Appendix F: List of Known Hum Reports




Final Report
Investigation of the Windsor Hum: Acoustical Assessment 1
1. INTRODUCTION
The University of Windsor was retained by the Department of Foreign Affairs and International
Trade (DFAIT) to perform a noise source investigation study in Windsor, Ontario. The purpose
of the study was to investigate an unknown noise, first documented when the Ontario Ministry of
the Environment began receiving complaints in early 2011. Residents of the areas of West and
South Windsor, as well as the Town of LaSalle, started hearing an unnatural hum, or drone
sensation in their homes. The media eventually named the sound the Windsor Hum. Often
perceived in certain geographical areas, the hum sensation has been reported to appear
intermittently at all hours of the day, but has been most prominent during the late evening and
early morning hours.
The City of Windsor is located at the southwestern tip of Ontario and is at the international
border between Canada and the USA, separated by the Detroit River. Windsor is a highly
industrialized city with many industries directly related to, or in support of the automotive
industry. The Ambassador Bridge, located on the citys west side, is also the busiest border
crossing between the two countries in terms of trade volume.
1.1 The Hum Worldwide
Since the early 1970s, similar hums have been reported throughout the world. Reports of hum,
howl, and rumble have appeared in the news in communities including, but not limited to the
following, (in chronological order of first known reported occurrence):
- Sydney, Australia (1970);
- Largs, Scotland (1989);
- Taos, New Mexico (1991);
- Hueytown, Alabama (1992);
- Peterborough, England (1994);
- Auckland, New Zealand (2006);
- Liverpool, England (2008);
- Leeds, England (2009);
- Bristol, England (2009);
- Vancouver, British Columbia (2010);
- Woodland, Durham (2011);
- Herbrandston, Wales (2013);

In each of these cases, local newspapers have told stories of residents being kept awake at all
hours of the night, subject to the intermittent hum noise in their own homes and disrupting their



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 2
lives. Some describe the noise as similar to an idling diesel engine or industrial machines in the
distance. While not everyone in these communities has heard the hum, for those that do, the hum
sensation is a constant annoyance, which they do not become accustomed to. With each report
also comes a wide range of potential sources. Predictions vary from mechanical sources
(submarine activity, generators, worn industrial machines), sonar systems, radio towers, and
environmental features (rock faults and coastal shifts), to human physiological traits such as
psychological problems and tinnitus.
1.2 Impact on Residents
While not all residence within an affected community hear the Hum, some of those who do
report hearing it also describe side effects including loss of sleep and insomnia, nausea, muscle
spasms, chronic stress, difficulty concentrating, fatigue, heart palpitations, headaches, increased
tension, eye strain, pins and needles, irritable ear pressure, personality changes and occasionally
thoughts of suicide. Not all of the self-proclaimed hummers, suffer from side effects but for
many of the reported cases it is reported that the quality of life for these individual may be
lessened with negative impacts on relationships and careers.
Similar to those reports mentioned above, the Windsor Hum has been described by some who
have experienced it as an intolerable disturbance. Some people within the Windsor area
communities have also reported adverse health impacts including loss of sleep, headaches,
increased stress and depression. These reports have come from the media and direct
communications by those affected.
1.3 The Windsor Hum
When reports of the Windsor Hum began appearing in 2011, the Ministry of the Environment of
Ontario (MOE) launched an investigation that considered several local industries but could not
identify the source. It was concluded there were too many potential noise sources in the area for
a simple investigation. This was partially due to the number of industry and the geographic
extent of the documented complaints. As a heavy industrialized border city, the ambient noise
levels in the affected area are generally high as a result of high road and rail traffic volumes,
dense industrial operations and its close proximity to the international shipping channel and



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 3
Detroit, Michigan; a city which also has heavy industrialization along the Detroit River, west of
the international bridge crossing.
1.3.1 NRCan Investigation
Following the initial MOE investigation, a second study of the Windsor Hum was conducted by
National Resources Canada (NRCan) during the summer of 2011. NRCan installed several
Seismograph vibration monitors at various locations within the City of Windsor. A study of
ground vibrations was deemed an appropriate approach given that initial reports suggested that
the Hum was the result of ground-borne vibrations. Monitoring locations were selected within
the affected communities to locate the source of the Hum disturbance. Several conclusions were
made by the researchers following several months of data collection: (i) the long duration of the
excitations were not consistent with earthquake or other seismic activities, (ii) the prominent
frequency of the excitation was approximately 35 Hz, and (iii) due to the speed of the
propagating energy, the excitation was an air-borne noise source and not a ground vibration.
Samples of the results from the investigation are provided in Appendix A. By examining the gap
in time in the measured data between each of the various measurements locations, NRCan was
able to triangulate the data to approximate that the source was located in the general area of Zug
Island. Given that the NRCan report concluded that the Hum was the result of an air-borne
excitation, the next step to locate the source should be an acoustical study, (Bent and Woodgold
2011).
1.3.2 University Research
A press conference was held in January 2013 at the University of Windsor to announce a federal
government funded research project to locate the source of the Windsor Hum as a process to
protect the quality of life for citizens of Windsor. To facilitate this research, the Department of
Foreign Affairs and International Trade (DFAIT) contracted the University of Western Ontario
(UWO) and the University of Windsor into a joint research study to investigate the source of the
Windsor Hum phenomenon.
As part of the DFAIT contract, UWO deployed monitoring equipment in the area of LaSalle in
March of 2013. The equipment is capable of collecting low frequency data in the frequency



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 4
range from 0.1 Hz to 200 Hz. Most sounds having frequencies above 20 Hz are considered
audible to an average human listener when presented at sufficient amplitudes. The frequency
capabilities of the equipment assured that the previously observed 35 Hz tone was within the
measurement range of the UWO equipment.
A dedicated email address was established for local residents to report perceived events of the
Hum to aid the researchers in their subsequent analysis of the measured data. Both universities
had access to this information to allow them to refine the search parameters and to assess the
measured data in correlation to the residential feedback.
The preliminary results of the UWO study did not make any definite conclusions. Dominant
frequencies were found to be 15-20 Hz, 30 Hz, 42 Hz, and 60 Hz. It was further determined that
this entire frequency range of data originated from the direction of Detroit, including the area of
Zug Island.



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 5
2. RESEARCH APPROACH AND METHODOLOGY UNIVERSITY OF
WINDSOR
The research conducted by the University of Windsor was separated into two phases, (i) validate
the existence and characterize the noise for the Hum phenomenon, and (ii) upon confirming the
source characteristic, use noise source identification (NSI) technology to pin point the source
emitter; the only methodology available to the research team without having direct access to the
Zug Island lands.
2.1 Phase 1 Part 1: Long Term Noise Monitoring
The objective of Phase 1 was to continuously monitor and log the noise data of the affected
areas. To accomplish this, two outdoor noise monitoring terminals (NMT), Brel & Kjr (B&K)
Type 3639 NMTs, were deployed. These units are designed for permanent, mobile, and portable
monitoring where unattended outdoor noise (and/or vibration) measurements are required. The
NMT hardware meets the IEC 61672 Class 1 standard for noise measurement, as well as several
other standards including DIN 45657 and ANSI S1.4. The B&K product and service information,
product data sheets and general information relative to this project is provided in Appendix B.
The analyzer inside the weather tight NMT enclosure is a Type 2250 Sound Level Meter having
a broadband, linear frequency range from 3 Hz to 20 kHz which is a good range for this projects
requirements. The microphone is a Type 4952 Outdoor Microphone, capable of withstanding
harsh climate conditions, and has a frequency response from 8 Hz to 12.5 kHz.
The first NMT was a permanent installation located approximately 100 meters from the Detroit
River shoreline, directly across from Zug Island. A photo of the permanent NMT installation is
given in Figure 1. The second NMT was a mobile unit that was installed and relocated within
different residential areas. The chosen locations were based on areas where heavy concentrations
of Hum complaints have been noted. Photographs of the portable NMT system are given in
Appendix C. The intent of monitoring at the residential locations was to be able to correlate
reported Hum incidences with collected noise data.
The NMTs are equipped with a 3G cellular modem, which allow remote access to the real time
noise data using a personal computer having Internet access. Day and nighttime noise threshold
trigger levels were used to identify and record the noise when the preset sound pressure level was



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 6
exceeded. The intent of having the recorded sound is to capture, identify and distinguish the
sound of the Hum from other generated exceedances. For each trigger, a 30-second recording is
made and included in a detailed daily noise report, which is received by the researchers each
morning. Other metrics logged by the NMTs include a rolling equivalent energy level (Leq)
averaged over 5-minute intervals for each 24-hour period, statistical exceedance levels as well as
logged maximum and minimum levels and third octave frequency spectra. Results of the logged
data were subsequently automatically summarized into Excel spreadsheets and included in the
daily reports for review by the researchers. Samples of the Excel data sheets are provided in
Appendix C.


Figure 1: Permanent NMT Setup



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 7
2.2 Phase 1 Part 2: High Resolution Data Collection
While the NMTs provided recordings for triggered events with audio playback, as well as other
environmental noise metrics, the data was limited in terms of frequency resolution. For this
reason, a secondary measurement was used to better quantify the hum frequency signature and
consisted of a B&K Module LAN-XI Type 3056 acquisition system which has a frequency range
from 0 Hz to 51.2 kHz. A Type 4189 free-field microphone having a low-frequency cut-off of
6.3 Hz was also used. Using a specialized firmware application (BZ-7848) loaded on the
acquisition module, the LAN-XI system was used as a stand-alone recorder, thus eliminating the
need for a PC to be connected during data collection. The system has the advantage of recording
data at a higher sampling rate and in a format optimized for later post-processing of the acquired
data. This secondary unit was located adjacent to the permanent NMT as shown in Figure 2.


Figure 2: PULSE Acquisition System Setup at NMT
The secondary measurement system was deployed during a period when the Hum was active
from the evening on Friday July 5
th
until late in the morning on Saturday July 6
th
. From these
measurements, detailed frequency spectra, which characterized the low frequency nature of the



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 8
Hum, was acquired. The data, given in Appendix D, along with the sound recordings provided
confirmation of the Windsor Hum.
2.3 Phase 2: Noise Source Identification using Beamforming
With the Hum sound recorded and characterized, the next phase of the research was to locate the
source of the disturbance. During the study period, it was noted that optimum identification of
the Hum was during the late evening and nighttime. These are also times of lowest ambient
environmental background noise levels. Due to access restrictions to the island, a research vessel
owned by the University of Windsors Great Lakes Institute for Environmental Research was
used to get within close proximity of the Zug Island shore to facilitate the acquisition of NSI
data. Figure 3 is an aerial view of the Detroit River and Zug Island. The red circle in Figure 3
also indicates the target measurement zone for the NSI exercise.


Figure 3: Aerial Photo of Zug Island with Distance from Island to be Measured Marked in Red
A pentangular array Type WA-1676-W-02 (comprising of thirty Type 4959 microphones) was
used as the NSI microphone array. Both qualitative and quantitative data was collected of the
Zug Island site to locate potential noise sources for the Hum. The array is capable of outputting
acoustical colour maps, similar to thermal imaging techniques, to identify and quantify sources



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 9
of noise over a large target area using an acoustic beamforming algorithm. Product data and
information sheets describing the system are given in Appendix B. Acoustic beamforming is a
method of mapping noise sources using a microphone array that discerns the direction from
which the sound originates. This is accomplished by employing a time delay approach as the
source passes over the array of microphones. Using this approach, and knowing the target
frequencies identified in Phase 1 of the research, the aim was to identify the Hum signature
within the measurement target zone. A photograph of the pentangular array mounted on the
research vessel is shown in Figure 4. Bruel & Kjaer PULSE Labshop software was used to
acquire the noise data, which allows for real-time data acquisition and performed fundamental
real time and post processed analysis of the data. A 36-channel B&K LAN-XI acquisition system
was used as the data collection front end.


Figure 4: Setup of the Pentangular Array on Research Vessel



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 10
3. RESULTS
The following section details the observations and conclusion resulting from the data collection
and analysis exercises. Discussion of the results from the subjective and objective metrics is also
given.
3.1 Phase 1 Part 1: Long Term Noise Monitoring Results
Continuous noise data was collected by the NMTs over a period of approximately six months for
which the data reports and sound files were examined on a daily basis by the researchers using
both objective and subjective metrics. Regular visits were also made by the researchers to the
permanent NMT site to maintain the equipment and to qualitatively monitor the noise levels and
sounds. It was not until the first week of July that the Windsor Hum was detected using the NMT
recordings. It was then that the secondary measurement system was deployed and an analysis of
the data confirmed both the prominent low frequency spectra of the noise source as well as the
irregular intervals. The researchers also observed a specific nighttime operation, described as a
bright blue flame from several exhaust stacks located on Zug Island, which coincided with
observation of the Hum sound. The blue flames were easily visible from the Canadian shoreline,
nearly one kilometer away. While there are many factors involved, a blue flame is typically an
indication of high levels of oxygen during the combustion process, required in high temperature
industrial operations. While no solid relation can be made by these qualitative observations
between the acquired noise data and the specific nighttime operations at the Island, it is
important to note the correlation. A photograph illustrating the flames is given in Figure 5.



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 11

Figure 5: Photograph of Zug Island and identified Nighttime Operations Exhaust
3.2 Phase 1 Part 2: Results of High Resolution Data Collection
As discussed previously, a secondary equipment deployment and analysis was carried out once
the Hum was identified to measure the sound with equipment capable of higher frequency
resolution analysis. For this, a standalone B&K PULSE acquisition system was installed for a
short duration at the permanent NMT site near the Detroit River.
The data collected during the overnight period on July 5/6 was extensively assessed against
multiple criteria. The Windsor Hum signature was most dominant during the early morning
hours, specifically between 3 AM and 4 AM. Three short term recordings were selected and
post-processed using PULSE Reflex software using; (i) Fast Fourier Transformation (FFT), (ii)
overall time analysis, and (iii) waterfall frequency spectrum (three-dimensional FFT). The FFT
metric is an algorithm to transform the time domain data into the frequency domain. The overall
time analysis computes the sound pressure level (SPL) with respect to the recorded time domain.
The waterfall plot is a three-dimensional graph of the SPL versus the frequency and time
domains. The complete assessment of the collected and processed data is provided in Appendix
D. This includes both linear and A-weighting data analysis for each metric. This data confirmed
the presence of a significant and prominent low frequency component of the measured noise
located at 63 Hz. A sample portion of the data analyzed is provided in Figures 6, 7, and 8
below. These results are similar to those observed by NRCan and UWO.
Blue Flame



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 12

Figure 6: Plot of A-weighted Overall SPL vs. Time for Time Period 1

Figure 7: Plot of A-weighted Overall SPL vs. Frequency for Time Period 1

Figure 8: Waterfall plot of A-weighted SPL vs. Frequency and Time for Time Period 1



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 13
3.3 Phase 2: Results of Noise Source Identification Measurements
Following the first NMT measurements of the Hum noise, deployment of the NSI system was
not possible until the early morning hours on August 22, 2013, due to availability of the research
boat. It became evident to the researchers that the specific nighttime industrial operations,
identified previously and illustrated in Figure 4, were not active at Zug Island during this day. As
a result, the Windsor Hum signature was not captured. The researchers closely monitored for the
Hum sound and were not able to identify it again before the close of the acquisition time line of
the study. However, given that the NSI acquisition equipment was installed, calibrated and ready
on-board, acquisition of NSI data was made to further identify and characterize the other various
noise sources on the island. The collected data was subsequently post-processed using the B&K
array acoustics post-processing software suite. A sample analysis output is provided in the noise
contour plot given in Figure 9. The corresponding analysis of the sound power 1/3 octave
frequency response is shown in Figure 10. For reference, an octave is the interval between two
points where the frequency at the second point is twice the frequency of the first.
The Zug Island region is isolated and shown in red for both the contour and frequency response
figures. Analysis of the data shows that the major contributor to the overall sound power level
from the region of Zug Island is from the low-frequency noise content. This observation is given
even in the absence of the Hum and previously identified nighttime operations. It should be
noted that due to the wave and drifting motion of the boat during the data acquisition, some error
resulting from sound coming from, and reflecting off of the water, was detected. While this does
result in some loss of spatial determination of the noise sources, it does not remove the fact of
the presence of significant measured low frequency noise.



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 14

Figure 9: Contour Plot over Image from Pentangular Array Measurement 1


Figure 10: Sound Power Level Frequency Spectral Analysis Measurement 1



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 15
At the conclusion of the first boat trip, the primary focus was aimed to examine the data to
determine if the methodology undertaken was successful, even though Hum phenomenon did not
occur that night. Once the data was examined, the method and testing procedure was refined to
make it the most accurate and reduce possible sources of error during future trips. It is worth
noting that during the measurement acquisition, a dominant noise source was detected during
measurements 7 through 12. This noise source was identified to be high-pressure steam exhaust
operation (possibly part of a maintenance or equipment testing process) and was not observed
previously by the research group.
To avoid unnecessary measurements, the decision was made to not attempt another nighttime trip
with the NSI equipment until the characteristic Hum was again positively identified. Continued
noise monitoring of the NMT data was the focus of the research from this point forward.
Unfortunately, the Hum did not manifest itself again before the NMT monitors were
decommissioned in early October. In fact, the researchers did not become aware of any further
Hum activity until January 2014, just prior to the completion of this report.



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 16
4. DISCUSSION
Results obtained have shown that the Windsor Hum does exist and is also a quantifiable sound.
While direct access to Zug Island have restricted any conclusions to qualified guesses, the
ground vibration data measured by NRCan and array data obtained by the Western University of
Ontario team support the conclusions of this report that the Windsor Hum phenomenon most
likely originates from the Zug Island area. If anything, there has been no data to date, which
suggests that the source is from anywhere else, especially given the heavy industrialization of the
surrounding area on both sides of the Detroit River. Further, the observed operations at Zug
Island are also consistent and coincide with the time periods of Hum activity. To further the
discussion and strengthen the hypothesis, the following discussions identify potential noise
sources on the island and the most likely to cause the observed characteristic sounds and
frequencies.
4.1 Zug Island Description
Zug Island is an industrial island located on the south-eastern edge of Detroit, Michigan and
fronts onto over one kilometer of waterfront property. The island is a prominent area feature and
can be easily seen from the Canadian side of the border. Transportation on and off the island is
available three ways; (i) a land bridge which services the island for employee traffic and
shipping operations via large trucks; (ii) a rail bridge which is located on either side of the island
to facilitate the export of steel and raw materials; and (iii) large shipping docks which allow
tankers and other large ships to load and unload materials at the waterfront. Operations on the
island include two large blast furnace facilities and a petroleum coke (petcoke) baking process.
While vehicle traffic and railway horns can be heard clearly at the Windsor waterfront, the most
audible sounds are characteristic of the blast furnace operations and associated high temperature
exhaust towers.
4.2 Blast Furnace Operation Overview
The high capacity blast furnace, similar in appearance to those used on Zug Island, is a popular
design used for the production of pig iron as the unique shape allows for a high production rate
while also maintaining a high level of thermal efficiency. While an in-depth detail of the entire



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 17
blast furnace process is not applicable to this research, the brief overview that follows is focused
on a discussion of the characteristics of the airflow (including the temperature) into the furnace,
as well as the general geometry of the stack. These parameters are the principal features related
to the noise emission levels. The described processes associated with the blast furnace operation
are taken directly from Steel Works: The Online Resource for Steel (last accessed in December
2013) and are outlined below.
The blast furnace is essentially a large cylindrical oven, which is used to convert mineral ore into
usable steel for manufacturing. Capable of producing 12,000 tonnes per day, a blast furnace can
run continuously for up to 10 years with only short periods of downtime to perform scheduled
maintenance (American Iron and Steel Institute 2013). Often standing over 20 metres in height,
the stack is the most prominent feature of the furnace. Acting as a funnel, the brick lined oven
channels raw materials through the superheated flame; chemically reducing and physically
converting iron oxides into liquid iron. A conveyor feeds the tower with varying proportions of
iron ore, limestone, and petcoke which steadily pour into the top of the furnace. Typically the
raw materials require 6 to 8 hours to descend to the bottom (this is dependent on several of the
processes parameters, such as temperature, air flow, etc.). At the same time, preheated as air is
force-blown into the bottom of the furnace, which ascends to the top of the stacks in
approximately 6 to 8 seconds and triggers numerous chemical reactions. The superheated air is
forced into the bottom of the tower, providing the necessary oxygen for combustion. It enters at
both high volume and velocity through a series of water-cooled ports known as tuyeres; this
process is further described below. Air enters the blast zone of the furnace igniting the petcoke
and converting the mineral ore and petcoke mix into a molten mixture of slag and pig iron.
The rate of pig-iron production in any given blast furnace is proportional to the rate at which
oxygen, as air or air plus pure O
2
is blown into the furnace. Typical operational blast rates in the
furnace are 40-50 Nm
3
/min of blast air per square metre of hearth area. It is worth noting that a
blast furnace can operate down to about 70% of its normal capacity without any deleterious
effects; below this rate, the charge begins to react unevenly. Blast temperatures in modern
installations are typically from 1000C to 1300C (Peacey and Davenport 1981).
The high burning temperature increases the efficiency of the furnace. To achieve this, hot dirty
exhaust gas is collected at the top of the furnace and fed through a series of gas cleaning



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 18
processes to recirculate the effluent. The exhaust gas is re-burnt as fuel in hot ovens, which
generates steam. This process is used to convert the thermal energy from the effluent gas into
mechanical energy for a turbo blower, compressing the large amount of air required in the
smelting process (American Iron and Steel Institute, 2013). The turbo blower is the first step in
generating the superheated air outside of the furnace with flow-rates between 37 m
3
/s and 108
m
3
/s. The blown air flows through the cold blast main up to the hot stoves exterior of the plant.
This cold blast enters the stove, which has been previously heated by dirty exhaust gas from the
stack. The heat stored in the stoves refractory brick is then transferred to convert the cold air to
a hot blast form. The moving air temperature at this stage can be from 870C to 1260C
depending on the design and conditions of the stove. As the superheated air exits the stove it
enters the hot blast main which runs up to the intake at the bottom of the furnace. There is a
mixer line connecting the cold and hot blast mains equipped with a value to control the blast
temperature to keep it constant. The hot blast main enters into a doughnut shaped pipe that
encircles the furnace, named the bustle pipe. The circular tube distributes the air around the
furnace through equally spaced nozzles, called tuyeres around the circumference of the
furnace. The number of nozzles may vary from fourteen on a small blast furnace to forty on a
large blast furnace. These tuyeres are made of copper and are water cooled since the temperature
directly in front of them may between 1980C and 2300C. Oil, tar, natural gas, powdered coal
and oxygen can also be injected into the furnace at the tuyere level to combine with the coke to
release additional energy. This may be necessary to increase productivity. The molten iron and
slag drip past the tuyeres on the way to the furnace hearth, which starts immediately below
tuyere level.
4.3 Potential Noise Sources
Outlined above is the typical blast furnace process for converting petcoke and iron ore into
commercial grade steel. Unfortunately, as the investigators did not have access to the Zug Island
facility, they can only speculate the specific processes involved at this facility and the associated
potential sources of noise associated with them. From the above described process, noise sources
may include but are not limited to: (i) the high velocity air travelling through air ducts and the
tuyeres which ignites while entering the blast zone; (ii) the discharge of compressed air
fluctuations into the atmosphere; (iii) the effluent gas exiting the waste stack continuously being



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 19
burnt off; and (iv) the intermittent release of large amounts of steam. Secondary noise sources
may come from other processes on the island, most notably from the screening process to sort
the baked coke piles and raw materials after their crushed.
4.3.1 Blast Mains and Ducts, and Tuyeres
As Zug Island is assumed to be a high volume facility, it necessarily follows that the equipment
involved would be of a large industrial scale, capable of emitting a proportional amount of noise.
The airflow requirements alone to run such a plant are massive and can produce a large amount
of associated duct noise. The formula below, taken from (Crocker 2007), is used to aid in the
prediction of air duct noise sound power level (L
w
), based on the flow velocity (v) and cross-
sectional area (A).
L
w
= constant + 10*log(A) + 60*log(v)
The constant is dependent on the type of fitting used in the air duct. As an example, assigning a
value of zero to the constant and approximating the values of flow velocity and cross-sectional
area based on previously provided information, sound power levels exceeding 150 dB are
predicted. While this value is not directly applicable to blast furnace noise, but serves as a
reference for the level of energy generated by a system of this magnitude.
For a larger furnace, a greater number of tuyeres are required to provide constant heat throughout
the blast zone. According to Klyacho and Shigorin (1975), the most intense period of this
operation is at the start of the melting process as the largest amount of minerals are ignited
simultaneously. In order to increase the heat of a flame, it requires a greater airflow. The
Windsor Hum signature is unique in the sense that the noise experienced is likely to be mostly
contained in the low frequency dynamics (LFD), ranging from 0 to 70 Hz. This range is also
subject to both the audible as well as the inaudible, or feel sensation. Furnace operations contain
instabilities caused primarily by temperature fluctuations, with factors such as modifications in
the material volumes, changes in the airflow, etc. can also have an impact (Crocker 2007). These
instabilities result in pressure changes, and it is shown below in Figure 11, the largest spike in
pressure amplitude occurs in the LFD. It is important to note that Figure 11 is only a typical
spectrum depicting the changes in pressure caused by thermo-acoustic instabilities (Crocker



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 20
2007); with this in mind, it is shown that the principle spike in the LFD is estimated to be in the
60 to 80 Hz frequency range, which is within the same range that the Windsor Hum signature
was characterized by the Windsor research team.

Figure 11: Typical Frequency Spectrum of Thermo-acoustically Induced Pressure Oscillations
(Crocker 2007)
4.3.2 Air Discharge Valves (Snort Valves)
As the Windsor Hum is characterized by an intermittent time signature which can last hours at a
time, it is unlikely that air-discharge valves are the primary cause of the disturbance. With that in
mind, it is still important to account for all primary noise sources in a blast furnace environment.
Snort valves allow for the exhaust of excess pressure fluctuations into the atmosphere in case of
emergency. This typically occurs when pockets of air and fuel ignite in the blast furnace and a



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 21
large pressure burst results. When the snort valve is released, it is accompanied by a proportional
blast of noise, which unless properly silenced, can propagate large distances from the facility.
Fortunately, successful attenuating measures have been found for effectively reducing snort
valve noise. Each snort valve may be outfitted with perforated steel tube lined with steel wool as
a sound absorbing-material. With a large enough silencer in place, the installation allows for
effective attenuation with little impact on the capacity of the air-discharge valve, (Klyachko and
Shigorin 1975).
4.3.3 Effluent Stack Burn Off from Hot-Blast Stove
As was shown in Figure 5 of this report, a large blue flame is an indication of a high temperature
combustion process taking place. A flame of this type is capable of producing a large amount of
turbulence and aerodynamic discharge noise which can be easily heard on the Canadian side of
the Detroit River. The flame will typically have medium and high frequency components as
opposed to the low-frequency spectra of the Windsor Hum. Due to the size of the stack and
flame, attenuating this noise source is quite difficult. Smaller exhaust stacks require sound
insulating covers which absorb the acoustic energy from the reaction. However, the high level of
heat and fast flow-rate render this form of attenuation largely ineffective. Rerouting the exhaust
or finding another means of dispersing the effluent may be necessary.
4.3.4 Boiler Facility Release
Once used, the steam from the boiler house is discarded, resulting in a violent gas expansion, see
Figure 12 below. During this release, noise levels can reach dangerous levels for workers
nearby, while the high humidity levels make conventional attenuation methods inappropriate.



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 22

Figure 12: Photograph of Periodic Steam Release from Zug Island
4.3.5 Raw Material Screening Process
Screening operations produce large amounts of low frequency noise and vibrations as a result of
centrifugal loading forces on the filter screens. This force is most often caused by an off-
balanced rotation which if not properly absorbed will transfer into the bearing assembly and can
propagate into the device mounts. (Klyachko and Shigorin 1975). Unwanted vibrations are
commonly known to produce either ground borne noise and vibration, mechanical noise, or a
combination of the two.
Decreasing the operating speed of the unbalanced force or isolating the machine from the
surroundings can often effectively reduce these levels.



Final Report
Investigation of the Windsor Hum: Acoustical Assessment 23
5. CONCLUSIONS AND RECOMMENDATIONS
The conclusion of the research is that the Windsor Hum does exist and has both qualitative and
quantitative characteristics that surmise the likely source of the Hum to be from the blast furnace
operations on Zug Island. It is further concluded that the Hum is audible by a concentrated
portion of Windsors west end community and other nearby residential areas. These conclusions
are reinforced by the periodical nature of the Hum noise, which has been observed and measured
by the researchers and coincides with the observed irregular operations associated with the steel
making operations on Zug Island.
It is recommended that additional noise measurements be made in closer proximity of the target
sources using the same techniques used in this study. Such an assessment should focus on
nighttime operating conditions with acquired data coinciding with reports of the Hum on the
Windsor side of the Detroit River. This however, will require access to the Zug Island if a
meaningful outcome is expected.



Colin Novak, Ph.D., P.Eng.

NVH-SQ Research Group, University of Windsor







Final Report
Investigation of the Windsor Hum: Acoustical Assessment 24
6. BIBLIOGRAPHY
American Iron and Steel Institute. SteelWorks - The online resource for steel. 2013.
http://www.steel.org/ (accessed 12 06, 2013).
Bent, A.L., and C.R.D. Woodgold. Seismic Investigation of the "Rumblings" in Windsor Ontario.
Geological Survey of Canada, Open File 6999, 2011.
Biswas, A.K. Principles of Blast Furnace Ironmaking. Brisbane: Cootha Publishing House,
1981.
Crocker, M. J. Handbook of Noise and Vibration Control. Hoboken: John Wiley & Sons Inc.,
2007.
IHSA. "Construction Health & Safety Manual - Section 34 Steel Mills." Infrastructure Health &
Safety Association. 2013. http://www.ihsa.ca/resources/health_safety_manual.cfm
(accessed 12 06, 2013).
Klyachko, L.N., and P.I. Shigorin. "Reducing the Noise Level in Steel-Melting and Blast-
Furnace Departments." Metallurgy 7 (July 1975): 40-42.
Peacey, J. G., and W. G. Davenport. Principles of Blast Furnace Ironmaking: Theory and
Practice. Cootha, 1981.
Reeve, D. A., J. T. Price, and J. F. Gransden. High Temperature Behaviour of Blast Furnace
Coke A Review. Goverment Document, Canada Centre for Mineral and Energy
Technology (CANMET), 1977.
Silber, E.A., and P.G. Brown. Scientific Research to Characterize and Localize the Windsor
Hum: Preliminary Report. Department of Physics and Astronomy, University of Western
Ontario, 2013.
Steel Works: The Online Resource for Steel. How a Blast Furnace Works. n.d.
http://www.steel.org/Making%20Steel/How%20Its%20Made/Processes/How%20A%20
Blast%20Furnace%20Works%20larry%20says%20to%20delete.aspx (accessed October
2013).
Szekely, J. Blast Furnace Technology: Science and Practice. New York: Marcel Dekker, 1972.







Final Report
Investigation of the Windsor Hum: Acoustical Assessment











APPENDIX A
NRCan & UWO Data






Contents

NRCan Ground-borne Vibration Data
University of Western Ontario (UWO) Data



Final Report
Investigation of the Windsor Hum: Acoustical Assessment A 2
NRCan Ground-borne Vibration Data


A 1: Ariel View of NRCan Installations (Bent and Woodgold 2011)
Station Identifiers:
MEMO Ministry of the Environment
UNIV University of Windsor
POWR Brighton Bean Power Station
LYON Private Residence on Lyoness Avenue




Final Report
Investigation of the Windsor Hum: Acoustical Assessment A 3

A 2: NRCan Source Estimation with 95% Confidence Interval (Bent and Woodgold 2011)
ZUG ISLAND



Final Report
Investigation of the Windsor Hum: Acoustical Assessment A 4
University of Western Ontario (UWO) Data


A 3: UWO Monitoring Equipment Locations (Silber and Brown 2013)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment A 5

A 4: Photograph of UWO Equipment (Centre Element) Installation (Silber and Brown 2013)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment A 6

A 5: UWO Waveform and Spectrogram Plots (Silber and Brown 2013)








Final Report
Investigation of the Windsor Hum: Acoustical Assessment














APPENDIX B
Brel & Kjr Product Data Sheets






Contents

Noise Monitoring Terminal Type 3639-A
Sound Level Meter Type 2250
Outdoor Microphone Type 4952
Input Module LAN-XI Type 3056
1/2 Free Field Microphone Type 4189
Stand-alone Recorder LAN-XI Notar BZ 7848
PULSE Reflex Core Post-processing Software
Pentangular Array Type WA-1676-W-002 (30 x 4959 Microphones)
Array Microphone Type 4959
PULSE LabShop
Array Acoustics Post-processing Software

http://www.bksv.com




Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 2
Noise Monitoring Terminal Type 3639-A


B 1: Noise Monitoring Terminals Types 3639 and 3655 Uses and Features



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 3

B 2: Noise Monitoring Terminals Types 3639 and 3655 Physical Configurations



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 4

B 3: Noise Monitoring Terminals Types 3639 and 3655 Microphone Options



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 5

B 4: Noise Monitoring Terminals Types 3639 and 3655 Noise Monitoring and Analysis



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 6

B 5: Noise Monitoring Terminals Types 3639 and 3655 Power Options



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 7

B 6: Noise Monitoring Terminals Types 3639 and 3655 Compliance with Standards and
Specifications



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 8

B 7: Noise Monitoring Terminals Types 3639 and 3655 Common Specifications



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 9

B 8: Noise Monitoring Terminals Types 3639 and 3655 Common Specifications (Continued)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 10
Sound Level Meter Type 2250


B 9: Hand-held Analyzer Types 2250 and 2270 Uses and Features



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 11

B 10: Hand-held Analyzer Types 2250 and 2270 Applications and Hardware



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 12

B 11: Hand-held Analyzer Types 2250 and 2270 Using the Platform: Key Features



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 13

B 12: Hand-held Analyzer Types 2250 and 2270 Display Capabilities



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 14

B 13: Hand-held Analyzer Types 2250 and 2270 Compliance with Standards and Specifications



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 15

B 14: Hand-held Analyzer Types 2250 and 2270 Specification (Continued)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 16
Outdoor Microphone Type 4952


B 15: Product Data: Outdoor Microphones Type 4952 Uses and Features



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 17

B 16: Product Data: Outdoor Microphones Type 4952 Description



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 18

B 17: Product Data: Outdoor Microphones Type 4952 Description (Continued) and
Compliance with Standards



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 19

B 18: Product Data: Outdoor Microphones Type 4952 Specifications



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 20

B 19: Transducers and Conditioning Microphone History



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 21

B 20: Transducers and Conditioning Outdoor Microphone Comparison



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 22
Input Module LAN-XI Type 3060


B 21: LAN-XI Data Acquisition Hardware Uses and Features



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 23

B 22: LAN-XI Data Acquisition Hardware System Overview



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 24


B 23: LAN-XI Data Acquisition Hardware System Overview (Continued)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 25

B 24: LAN-XI Data Acquisition Hardware Notar BZ-7848-A



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 26

B 25: LAN-XI Data Acquisition Hardware Compliance with Standards



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 27

B 26: LAN-XI Data Acquisition Hardware Specifications



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 28

B 27: LAN-XI Data Acquisition Hardware Specifications (Continued)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 29

B 28: LAN-XI Data Acquisition Hardware Specifications (Continued)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 30
1/2 Free Field Microphone Type 4189


B 29: Transducers and Conditioning Microphone History



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 31

B 30: Transducers and Conditioning Microphone Matrix



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 32

B 31: Transducers and Conditioning Free Field Microphones



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 33
Stand-alone Recorder LAN-XI Notar BZ 7848


B 32: Stand-alone Recorder LAN-XI Notar BZ-7848 Uses and Features



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 34
PULSE Reflex Core Post-processing Software


B 33: PULSE Reflex Core Overview



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 35

B 34: PULSE Reflex Core Overview (Continued)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 36
Pentangular Array Type WA-1676-W-002 (30 x 4959 Microphones)


B 35: PULSE Array-based Noise Source Identification Solutions Hardware and Software



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 37

B 36: PULSE Array-based Noise Source Identification Solutions Arrays



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 38

B 37: PULSE Array-based Noise Source Identification Solutions Measurement Methods



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 39

B 38: PULSE Array-based Noise Source Identification Solutions Measurement Methods
(Continued)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 40

B 39: PULSE Array-based Noise Source Identification Solutions Planar Beamforming



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 41

B 40: PULSE Array-based Noise Source Identification Solutions Planar Beamforming
(Continued)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 42

B 41: PULSE Array-based Noise Source Identification Solutions Pentangular Array System



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 43

B 42: PULSE Array-based Noise Source Identification Solutions Specifications



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 44

B 43: PULSE Array-based Noise Source Identification Solutions Specifications (Continued)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 45

B 44: Beamforming



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 46
Array Microphone Type 4959


B 45: 20 kHz Array Microphone Type 4959 Uses and Features and Description



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 47

B 46: 20 kHz Array Microphone Type 4959 Specifications



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 48
PULSE LabShop


B 47: PULSE LabShop Overview



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 49

B 48: PULSE LabShop Analysis and Performance



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 50

B 49: PULSE LabShop Customization and Programmability and Hardware



Final Report
Investigation of the Windsor Hum: Acoustical Assessment B 51
Array Acoustics Post-processing Software


B 50: PULSE Array-based Noise Source Identification Solutions Array Acoustics Pose-
processing Suite



Final Report
Investigation of the Windsor Hum: Acoustical Assessment














APPENDIX C
Phase 1 Part 1






Contents

Phase 1 Part 1 Installation Photographs
Phase 1 Part 1 Measurement Locations
Phase 1 Part 1 Measurement Data



Final Report
Investigation of the Windsor Hum: Acoustical Assessment C 2
Phase 1 Part 1 Installation Photographs


C 1: Complete Installation of Permanent Noise Sentinel Site Complete View



Final Report
Investigation of the Windsor Hum: Acoustical Assessment C 3

C 2: Installation of Permanent Noise Sentinel Site Higher Portion
Type 4952
Outdoor
Microphone
130 Watt
Solar Panel
Antenna



Final Report
Investigation of the Windsor Hum: Acoustical Assessment C 4

C 3: Installation of Permanent Noise Sentinel Site Lower Portion
Power
Communications
Box
Noise Monitoring
Terminal (NMT)
Type 2250
Sound Level
Meter
3G Modem



Final Report
Investigation of the Windsor Hum: Acoustical Assessment C 5

C 4: Installation of Portable Noise Sentinel at Local Residence NMT
Noise Monitoring
Terminal (NMT)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment C 6

C 5: Installation of Portable Noise Sentinel at Local Residence Microphone
Type 4952
Outdoor
Microphone



Final Report
Investigation of the Windsor Hum: Acoustical Assessment C 7
Phase 1 Part 1 Measurement Locations


C 6: Noise Sentinel Monitoring Locations
Permanent NMT
Location
Portable NMT Location
Cabot Avenue
Portable NMT Location
Baxter Cresent
Portable NMT Location
Gignac Cresent



Final Report
Investigation of the Windsor Hum: Acoustical Assessment C 8
Phase 1 Part 1 Measurement Data


C 7: Noise Sentinel Live Web Monitoring

C 8: Sample Section of Noise Sentinel Daily Report Long Period

Arial View of
NMT Locations
and Noise Levels
Real-time SPL Data Plot
One-hour Leq Plot



Final Report
Investigation of the Windsor Hum: Acoustical Assessment C 9

C 9: Sample Section of Noise Sentinel Daily Report Short Period (Part 1)

C 10: Sample Section of Noise Sentinel Daily Report Short Period (Part 2)



Final Report
Investigation of the Windsor Hum: Acoustical Assessment C 10

C 11: Sample Section of Noise Sentinel Daily Report Logged Data (Part 1)

C 12: Sample Section of Noise Sentinel Daily Report Logged Data (Part 2)




Final Report
Investigation of the Windsor Hum: Acoustical Assessment C 11

C 13: Sample Section of Noise Sentinel Daily Report Logged Data (Part 3)


C 14: Sample Section of Noise Sentinel Daily Report System Performance



Final Report
Investigation of the Windsor Hum: Acoustical Assessment C 12

C 15: Sample Section of Noise Sentinel Daily Report Setup

C 16: Sample Section of Noise Sentinel Daily Report Alert Data




Final Report
Investigation of the Windsor Hum: Acoustical Assessment














APPENDIX D
Phase 1 Part 2






Contents

Phase 1 Part 2 Measurement Photographs
Phase 1 Part 2 Measurement Location
Phase 1 Part 2 Measurement Data



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 2
Phase 1 Part 2 Measurement Photographs


D 1: PULSE Acquisition System Setup at NMT



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 3
Phase 1 Part 2 Measurement Location


D 2: Phase 1 Part 2 Equipment Location and Surrounding Area
Phase 1 Part 2
Equipment Location



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 4

D 3: Phase 1 Part 2 Equipment Location and Surrounding Area Zoomed-in View
Phase 1 Part 2
Equipment Location



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 5
Phase 1 Part 2 Measurement Data


D 4: Plot of Linear Overall SPL vs. Frequency (0 3.2 kHz) for Time Period 1

D 5: Plot of A-Weighted Overall SPL vs. Frequency (0 3.2 kHz) for Time Period 1



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 6

D 6: Plot of Linear Overall SPL vs. Frequency (0 400 Hz) for Time Period 1

D 7: Plot of A-Weighted Overall SPL vs. Frequency (0 400 Hz) for Time Period 1

D 8: Plot of Linear Overall SPL vs. Time for Time Period 1



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 7

D 9: Plot of A-Weighted Overall SPL vs. Time for Time Period 1

D 10: Waterfall Plot of Linear SPL vs. Frequency and Time for Time Period 1

D 11: Waterfall Plot of A-weighted SPL vs. Frequency and Time for Time Period 1



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 8

D 12: Plot of Linear Overall SPL vs. Frequency (0 3.2 kHz) for Time Period 2

D 13: Plot of A-Weighted Overall SPL vs. Frequency (0 3.2 kHz) for Time Period 2

D 14: Plot of Linear Overall SPL vs. Frequency (0 400 Hz) for Time Period 2



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 9

D 15: Plot of A-Weighted Overall SPL vs. Frequency (0 400 Hz) for Time Period 2

D 16: Plot of Linear Overall SPL vs. Time for Time Period 2

D 17: Plot of A-Weighted Overall SPL vs. Time for Time Period 2



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 10

D 18: Waterfall Plot of Linear SPL vs. Frequency and Time for Time Period 2

D 19: Waterfall Plot of A-weighted SPL vs. Frequency and Time for Time Period 2

D 20: Plot of Linear Overall SPL vs. Frequency (0 3.2 kHz) for Time Period 3



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 11

D 21: Plot of A-Weighted Overall SPL vs. Frequency (0 3.2 kHz) for Time Period 3

D 22: Plot of Linear Overall SPL vs. Frequency (0 400 Hz) for Time Period 3

D 23: Plot of A-Weighted Overall SPL vs. Frequency (0 400 Hz) for Time Period 3



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 12

D 24: Plot of Linear Overall SPL vs. Time for Time Period 3

D 25: Plot of A-Weighted Overall SPL vs. Time for Time Period 3

D 26: Waterfall Plot of Linear SPL vs. Frequency and Time for Time Period 3



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 13

D 27: Waterfall Plot of A-weighted SPL vs. Frequency and Time for Time Period 3

D 28: Plot of Linear Overall SPL vs. Frequency (0 3.2 kHz) for All 3 Time Periods

D 29: Plot of A-weighted Overall SPL vs. Frequency (0 3.2 kHz) for All 3 Time Periods



Final Report
Investigation of the Windsor Hum: Acoustical Assessment D 14

D 30: Plot of Linear Overall SPL vs. Frequency (0 400 Hz) for All 3 Time Periods

D 31: Plot of A-weighted Overall SPL vs. Frequency (0 400 Hz) for All 3 Time Periods




Final Report
Investigation of the Windsor Hum: Acoustical Assessment














APPENDIX E
Phase 2






Contents

Phase 2 Measurement Photographs
Phase 2 Measurement Location
Phase 2 Measurement Data



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 2
Phase 2 Measurement Photographs


E 1: Pentangular Array Setup



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 3

E 2: Pentangular Array and Data Acquisition Equipment Setup



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 4

E 3: Zug Island during Phase 2 Measurement Period Photograph 1



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 5

E 4: Zug Island during Phase 2 Measurement Period Photograph 2



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 6
Phase 2 Measurement Location


E 5: Aerial Photo of Zug Island with Distance from Island to be Measured Marked in Red



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 7
Phase 2 Measurement Data


E 6: Contour Plot over Image from Pentangular Array Measurement 1

E 7: Sound Power Level Frequency Spectral Analysis Measurement 1



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 8

E 8: Contour Plot over Image from Pentangular Array Measurement 2

E 9: Sound Power Level Frequency Spectral Analysis Measurement 2




Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 9

E 10: Contour Plot over Image from Pentangular Array Measurement 3

E 11: Sound Power Level Frequency Spectral Analysis Measurement 3



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 10

E 12: Contour Plot over Image from Pentangular Array Measurement 4

E 13: Sound Power Level Frequency Spectral Analysis Measurement 4



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 11

E 14: Contour Plot over Image from Pentangular Array Measurement 5

E 15: Modified Contour Plot over Image from Pentangular Array Measurement 5




Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 12

E 16: Sound Power Level Frequency Spectral Analysis Measurement 5

E 17: Contour Plot over Image from Pentangular Array Measurement 6



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 13

E 18: Modified Contour Plot over Image from Pentangular Array Measurement 6

E 19: Sound Power Level Frequency Spectral Analysis Measurement 6



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 14

E 20: Contour Plot over Image from Pentangular Array Measurement 7

E 21: Sound Power Level Frequency Spectral Analysis Measurement 7



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 15

E 22: Contour Plot over Image from Pentangular Array Measurement 8

E 23: Sound Power Level Frequency Spectral Analysis Measurement 8



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 16

E 24: Contour Plot over Image from Pentangular Array Measurement 9

E 25: Sound Power Level Frequency Spectral Analysis Measurement 9



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 17

E 26: Contour Plot over Image from Pentangular Array Measurement 10

E 27: Sound Power Level Frequency Spectral Analysis Measurement 10



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 18

E 28: Contour Plot over Image from Pentangular Array Measurement 11

E 29: Sound Power Level Frequency Spectral Analysis Measurement 11



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 19

E 30: Contour Plot over Image from Pentangular Array Measurement 12

E 31: Sound Power Level Frequency Spectral Analysis Measurement 12



Final Report
Investigation of the Windsor Hum: Acoustical Assessment E 20
E 32: Pentangular Array Measurement Information and Notes
Measurement # Measurement Time (s) Notes/Comments
1 2 Steady noise levels
during measurement
acquisition
2 2
3 2
4 5
5 5
6 10
7 2 Significant background
noise observed from Zug
Island, determined by
researchers to be some
sort of steam-related
process
8 2
9 2
10 2
11 1
12 2




Final Report
Investigation of the Windsor Hum: Acoustical Assessment














APPENDIX F
List of Known Hum Reports










Final Report
Investigation of the Windsor Hum: Acoustical Assessment F 2
Auckland, New Zealand
2006-10-22 NPR Radio Show
The buzz behind Aucklands hum
Robert Siegel, Host
Tom Moir Engineering Professor at Massey University was brought in to study it.

2011-06-14 ONE News
Auckland Hum
Pippa Wetzell (with Tom Moir)

2013-08-20 North Shore Times
40-year mystery
Liz Willis

Bristol, United Kingdom
2009-05-18 BBC Online
The Bristol Hum: Your Viewpoints

2013-07-29 Bristol Post Online
Have you ever heard the Bristol Hum? Scientists baffled by noise

Herbrandston, Wales, United Kingdom
2013-05-18 Western Telegraph
Herbrandston residents tortured by mystery low frequency noise






Hueytown, Alabama



Final Report
Investigation of the Windsor Hum: Acoustical Assessment F 3
Started January 1992
Coal mining town of 15000 people

1992-04-14 The New York Times
Hueytown Journal; Humming Along and Howling Mad
Ronald Smothers

Largs, Scotland, United Kingdom
Started in 1989 - Ongoing
Detected a frequency between 50Hz and 60Hz.

1990-01-06 The Herald Scotland
Hum adds unwelcome buzz to Largs Nightlife
Ian Sutherland

2001-10-18 The Guardian
Whats that noise? Its a constant, irritating hum that makes life miserable for all who
hear it but nobody knows what it is. Laura Barton Investigates.
Laura Barton
Accessed 2013-07-06

Leeds, United Kingdom
2009-05-19 BBC News
Have you heard the Hum?
James Alexander




Liverpool, England
Started in 2006



Final Report
Investigation of the Windsor Hum: Acoustical Assessment F 4

2008-10-04 Daily Post
Mystery hum makes us glum; Residents suffer years of baffling noise
Eryl Crump

Peterborough, England, United Kingdom
1994-06-22 The Independent
Whats that terrible noise?: All over the country people are plagued by a strange hum.
Are their ears playing up or is it something sinister? Emma Brooker Investigates
Emma Brooker

Sydney, Australia
Started in 1970

2007-10-04 Stuff.co.nz
Mysterious North Shore hum remains elusive

Taos, New Mexico
Started in 1991

Video Uploaded 2006-09-06
Thomaston
Started in 1998

2006-05-26 Foothills Media Group
Residents irritated by hum
Samantha M. Friedrich

Vancouver, British Columbia
2010-06-02 CTV Environment Reporter
Mysterious Hum Harasses Vancouver Residents



Final Report
Investigation of the Windsor Hum: Acoustical Assessment F 5
Mark Stevenson

Windsor, Ontario
2013-08-01 CBC News
Windsors Mysterious Hum is Real, Says Researcher

Wellington, New Zealand
2012-10-11 3News
Singapores frigate Stalwart source of Wellington hum?
Lloyd Burr

2012-10-11 3News
Wellington hum becomes nationwide obsession
Lloyd Burr
http://www.3news.co.nz/Wellington-hum-becomes-nationwide-
obsession/tabid/423/articleID/272369/Default.aspx

2012-10-14 3News
Mysterious Low Hum Noise Baffles Wellington, New Zealand Residents: low hum noise
baffles Wellington residents

2012-10-16 3News
Wellington hum disappears
Lloyd Burr



Woodland, Durham
2011-06-09 Telegraph
Tiny village is latest victim of the The Hum
Richard Alleyne



Final Report
Investigation of the Windsor Hum: Acoustical Assessment F 6

Worldwide News
2013-07-27 Huffington Post
Hum Heard Around World Impacts 2 Percent Of People In Hum-Prone Areas, Study
Suggests
Marc Lallanilla