The article describes a complaint-based noise problem in a residential building from an unknown mechanical source, which was determined to be a chiller housed located in an interior mechanical room.
Background
I was called in to investigate the source of an audible sound tone in several residential units in a 5-story residential building. The first floor is enclosed parking, with 4-floors of residences above, and a 3rd floor mechanical room on one end of the building. The building was new construction, occupied for about 1-year, and the noise complaint surfaced at the beginning of summer when air conditioning was first needed. Additionally, the building is a concrete structure.
There are several big clues in the description above; one, the noise complaint did not occur during the winter months, so initially it is most likely associated with some equipment which only runs when cooling is required, two, new construction, which means some piece of equipment was installed and had not been run during the summer months, except during the previous summer during building commissioning when no residents lived in the building, and three, a concrete structure, concrete structures readily transmit vibration and hence sound well over large distances.
Measurements
During the initial site visit, I was accompanied by the building maintenance director, and we went to talk to the resident complaining of the tonal noise.
Immediately upon entering the residence, I could hear the tonal sound. I then asked everyone to be quiet and made two sound level measurements, one with the chiller on and the second with the chiller off. The results of these measurements clearly showed a sound level peak, with the chiller on, in the 250 Hz 1/3-octave band. We then went to the mechanical room and made vibration measurements on the chiller.
Analysis
This case was unusual, in that, the culprit noise source was easily identified and all parties, the resident, the building facility manager, and myself could all hear the noise and could all hear it stop when the chiller was turned off. There was no way the building facility people were going to get out of this one.
Sound Level Data
There is a tonal or very narrow band, of sound causing the complaint. The 250 Hz 1/3-octave band sound levels are 23 to 24 dB higher than the 200 and 315 Hz 1/3-octave band sound levels. This is a significant level of sound to mitigate.
As an aside, in order to successfully mitigate tonal noise levels, the tone must be reduced to at least the levels of the sound levels of the two adjacent sidebands. Otherwise, because the character of the tone is so different from other background noise levels, the tone will still be audible.
Vibration Data
Analysis of the vibration data is a bit more complicated than the sound level data.
Next, there is a strong vibration peak at 238 Hz, which is in the 250 Hz 1/3-octave band (for sound data). The chiller motor was operating at ~30Hz, so this 238 Hz peak is six times the operating speed of the motor. The chiller vibration levels are in the ASHRAE Severity range of Fair to Slightly Rough. Additionally, there are two small peaks near 40 and 60 Hz, which are at the 1-times and 2-times of the drive motor speed. The vibration levels are in the ASHRAE Severity Smooth range and not a concern.
During the initial inspection of the chiller, it was disclosed that the chiller had been installed during construction, had not been run for ~1 year, and had not been periodically rotated. Storage of motors and pumps requires them to be rotated a minimum number of revolutions at periodic intervals and stopped at different angles of revolution to minimize damage to rolling elements.
The next analysis shows measured vibration levels measured on the discharge piping. The 238 Hz vibration peak is still present, although more than an order of magnitude lower than on the chiller. This does not rule the piping as the transmission path because small vibration levels can lead to clearly audible sound levels.
The fourth and fifth analyses show vibration levels on each side of the spring vibration isolators. There is an amplitude difference between the chiller side of the spring and the base side of the spring. This shows that the spring isolators are functioning as designed and providing over 95%.
A bit of caution at this point. While, the data shows the spring isolators are providing significant vibration isolation, the isolator may still be part of the transmission path, ONLY because the background sound levels inside the residential unit with the chiller off are so low, in the range of 26 dB in the 250 Hz 1/3-octave band.
Mitigation
The analysis above has identified the source of the tonal noise heard in the residential unit, and there are several ways to mitigate this tonal noise; reduce the vibration level of the driving force (the chiller defect), install vibration isolation in the intake and discharge piping, and last install more effective vibration isolation between the chiller and the building structure:
- Reduce the vibration level of the driving force, the chiller defect. Reducing the driving force will reduce the transmitted force and energy that goes into the building structure. In this case, remove the drive motor, inspect and if needed replace the drive motor bearing, and then inspect the bearings or bearing elements in the compressor. While this is the obvious first step, it is also expensive, and there may be other two options may be sufficient. In this case, the building owner decided to try the other options first.
- Install vibration damping elements in the discharge and intake piping. In this case, the initial design had a typical spherical isolator; which, are typically installed more for alignment than vibration isolation. Piping vibration isolation looks like a U or V-shaped loop, perpendicular to the pipe axis, with two braided steel flexible sections. In this case, the owner has opted to try this solution first.
- Install more effective vibration isolation between chiller and the building structure. There are situations, especially with concrete structures, that spring or neoprene isolation elements are not sufficient, even with very high isolation efficiencies. In these cases, the only solution is to install pneumatic isolators, in place of a typical spring isolator. The cases where pneumatic isolators are needed, is where the background sound levels are very low; as in this case or in cases where spaces require very low noise levels.
Summary
While the initial measurement data pointed directly to the source of the noise complaint, implementing a solution has been hampered by cost concerns. The building ownership has chosen to use the second isolation method, in hopes this will be sufficient. As I write this article, the second option has been installed, over the winter months, and we are waiting for warm enough weather to operate the chiller. Hopefully, no additional bearing damage has occurred over the winter months.
Looking to get NEBB Certified? Request your application today.