The Role of Insulation in Noise Control
A significant source of noise at pipeline facilities is piping radiated noise. Although the noise generation is due to the internal geometries and flow conditions of the rotating equipment or due to the orifices or valves, the acoustical radiator is the piping. Piping and compressor casings on numerous gas turbine-driven turbocompressor packages as well as piping at meter stations and regulator stations have been treated using acoustical laggings, as a retrofit as well as during design.
However, very little information exists in the industry regarding the acoustical performance of site-installed acoustical lagging systems. With the advent of innovative testing techniques, installed lagging system insertion losses have been obtained. Insertion loss is a method of testing and comparison developed for piping and equipment. Significant experience has been gained regarding material selection and system performance. As opposed to the massive panels typically used for acoustical enclosures on turbocompressor packages, acoustical laggings utilize lightweight and often removable materials.
Radiated noise from natural gas transmission compressor stations, meter stations and regulator stations is a problem of increasing concern, due to stricter regulatory enforcement as well as due to greater public awareness of noise as an environmental problem. For example, U.S. legislation by the FERC set maximum levels at 55 dBA Ldn; and similarly, Canadian legislation by the AEUB in Alberta and by the MOE in Ontario set maximum Permissible Sound Levels from gas transmission operations between 40 to 50 dBA Leq, based on the ambient noise environment.
One component that is a predominant contributor to environmental noise is the gas piping associated with centrifugal compressor packages at gas turbine-driven compressor stations. Noise is generated by the turbulence of the rotating wheel or impellers of the large turbocompressors but only a minimum amount of this noise is then radiated by the sufficiently thick compressor casing. The primary noise radiator is the relatively thin-walled process piping which is the focus of this article.
The solutions to the piping-radiated noise problem are acoustical silencers and acoustical laggings. The contractual responsibility for silencer design and supply is usually accepted by the silencer O.E.M. or by the equipment packager, attempting to achieve an overall acoustical performance requirement that usually has been specified by the gas transmission company as the purchaser. As in-line silencers typically have unacceptable pressure drops imposed upon the gas stream, acoustical pipe lagging becomes the preferred alternative.
The responsibility for piping-radiated noise is not addressed by any packager, supplier or O.E.M. Furthermore, acoustical lagging performance is not designed by or guaranteed by material manufacturers or by industrial insulation contractors. The responsibility is usually left to the facility owner as part of his design, hereby being the gas transmission company. In similar industries where centrifugal compression is utilized, such as in gas-processing plants and petrochemical complexes, the same approach occurs.
A lot is known about the acoustical performance of silencers, but relatively little is known about the acoustical performance of laggings. However, the acoustical performance of acoustical lagging systems must be known so the gas-transmission companies can specify material requirements to achieve their intended performance requirements.
MEASUREMENT OF RADIATED NOISE
The common descriptor for rating the acoustical performance of a layer of lagging material is laboratory-tested sound transmission loss, yet this is not relevant for the installed performance of multiple-layered systems that are in physical contact with a vibrating source. A lagging systems sound insertion loss more characteristically describes the installed acoustical performance. We believe that the most appropriate means to determine an acoustical lagging systems insertion losses over the audible sound frequency range is by field testing.
In order to determine the noise-reduction requirements for an existing facility, the preferred approach is to first measure and then quantify the noise source radiation characteristics. For gas-piping radiated noise, one should rely on site measurements. This is because turbocompressor and flow measurement equipment manufacturers do not typically have data available during the equipment procurement process, and those responsible for the piping design many times do not have sufficient information to carry out the acoustical calculations necessary.
Compressor-generated noise is a function of the compressor wheel configuration and geometry; the amount of piping-radiated environmental noise is a function of piping characteristics such as wall thickness, the extent of above-ground piping, and the types of support structures. Therefore, when conducting site measurements to choose acoustical lagging systems, one should concentrate on the amount of noise being radiated, without concentrating on its generation. In essence, the main aspect of consideration is determining how much noise should be attenuated by the lagging system.
The most accurate method of quantifying the piping-noise contribution at existing facilities is to isolate the predominant sources one at a time, and determine their radiated sound power levels. With these goals, a standard microphone placed nearby the noise source is not a reliable approach, as one runs a great risk of contamination of the measurement due to other nearby extraneous noise sources. To assist, we have used advanced sound measurement techniques. For example, the sound-intensity measurement technique utilizes a directional microphone probe, which can isolate the amount of acoustical energy originating from a defined noise source surface. Its measurement results exclude the effects of other nearby noise sources.
The results of a sound-intensity analysis on the suction and discharge piping as well as on the gas generator, power turbine and compressor casing of a turbocompressor package have previously been reported using this technique (Frank and Milner, 1995). This analysis showed the relative ranking of the various noise sources, gas piping was number one, with the discharge and suction piping ranking second and third, respectively.
Alternatively, we have had considerable success using an acoustical pipe-box measurement technique. The technique utilizes the result of noise measurements obtained inside a sealed box that straddles the pipe to determine the effect of that individual pipes noise-radiation characteristics, isolated from other nearby noisy pipes. Figure 1 shows a schematic of the acoustical pipe box on a pipe. The result of either the sound-intensity or pipe-box measurements are islolated sound power levels of the piping noise sources, expressed in terms of the acoustic power per lineal meter of pipe.
Based upon past experience, it is possible to estimate the amount of noise that would be produced by gas piping at proposed facilities before start-up. This means that potential noise problems can be detected during the facility design process, and noise control recommendations can then be implemented during the equipment procurement and construction phase. The accuracy of such theoretical predictions may not be as good as having conducted site measurements,. however, they are generally within the range of +/-5 dBA. The advantage of using facility design recommendations is that the problem is solved before it occurs, which usually justifies the initial capital costs spent to ensure that project design criteria are met.
ACOUSTICAL LAGGING SYSTEMS
Acoustical laggings consist of combinations of insulations and jacketings. The insulation must be porous to function acoustically, which means it is usually a porous fibrous material, such as mineral wool, glass-fiber or E-glass. Rigid or closed-cell materials should not be used for acoustical applications, as they provide extremely little noise reduction, and in some cases can actually increase the amount of noise radiated from insulated piping. For thermo-acoustical lagging systems, which is when noise control is required for cold service piping, combinations of first cellular glass and then mineral wool can be used as the insulation layers, to provide the thermal protection directly over the piping, and then provide the acoustical requirements thereafter.
Acoustical jacketings differ from conventional jacketings, in that they have a limp mass layer adhered to the outer metal jacket material. These limp mass layers usually consist of sheet lead or impregnated vinyl, which are thoroughly bonded to the outer jacketing which is usually aluminum. The use of lead is beginning to be questioned in certain regions, due to potential lead contamination during the installation process and due to potential ground water contamination if the lead were to leach out into rain water. Alternatively, some impregnated vinyls are not as suitable in very cold climates due to cracking if the pipes are walked on during unit shut-downs.
The porous fibrous insulation provides absorption of the airborne sound, and it structurally decouples the outer jacketing layer from the radiating pipe wall. The laminated jacketing provides additional mass for reduction of sound transmission. In addition, the composite of two materials laminated together damps out vibratory energy which would otherwise reradiate at critical resonant frequencies.
In some lagging systems, a septum layer is also utilized for increased performance at higher frequencies. This septum layer consists of an intermediate limp mass layer sandwiched between two layers of insulation, which are then covered by the acoustical jacketing. Acoustical pipe lagging visually appears similar to thermally insulated piping once installed. Figure 2 presents a schematic of an acoustical lagging system installed over a pipe.
The advantages of acoustical pipe laggings are as follows:
- they are relatively inexpensive when including labor and materials
- they can be installed relatively easily during facility operations without requiring equipment shut-down, and
- no pressure drop occurs as it does with in-line silencers. Alternatively, the disadvantages of acoustical pipe laggings are the following:
- they must be somewhat resilient to function acoustically, thus they are damage prone,
- they are susceptible to moisture buildup in certain climates, so there is a potential for pipe wall corrosion, and
- they are generally not removable and reusable, thus inspection for corrosion protection is not easy.
Vapor barriers and moisture barriers can be installed to mediate corrosion. The most successful applications have used bituminous materials installed between the insulation layer and the jacketing layer to prevent moisture from penetrating into the insulation and thus reacting with the pipe wall.
As an alternative to acoustical laggings, acoustical blankets are often used. Acoustical blankets contain materials similar to acoustical laggings, except the jacketings are silicone-impregnated fiberglass or Teflon-impregnated fiberglass cloths. These resemble what are known as turbine blankets, which are widely used on steam turbines in the petrochemical industry. Figure 3 presents a schematic of an acoustical blanket installed over a pipe.
An advantage of acoustical blankets is that they are designed for removability and durability. This feature provides easy access for critical locations such as valves which require frequent servicing or for compressor stations where management requires frequent inspections to monitor for corrosion. Alternatively, disadvantages of acoustical blankets are they provide slightly lower acoustical performance compared to that of acoustical laggings, and they also have higher costs than laggings.
As a compromise between acoustical laggings and acoustical blankets, removable covers are often used. They are manufactured out of semi-rigid frames using lightweight lagging materials. Although removable covers feature ease of removability and durability, they cost more than acoustical laggings.
LAGGING SYSTEM PERFORMANCE
As acoustical performance data do not exist with manufacturers or contractors, independent studies have been conducted to determine comparable acoustical performances of various lagging and blanket systems. Laboratory tests performed by a couple of acoustical jacketing manufacturers revealed various properties of insulation materials, such as density and rigidity, that affected acoustical performance (Holton, 1976). We recently conducted a detailed study of the relative performance of ten different acoustical pipe lagging systems and three different acoustical blanket systems (acknowledgment is hereby given to Pacific Gas Transmission for supporting this project and to NOVA Gas Transmission for availing a compressor station).
With a turbocompressor operating at a manual set-point to maintain consistency of the noise source, we first conducted acoustical measurements at specifically determined points at various locations along a bare compressor discharge pipe. The pipe was then treated with specifically chosen acoustical lagging systems, and then retested. The difference between the before and after measurements yielded the pipe lagging systems acoustical insertion losses. This work allowed us to determine the specific acoustical performance of lagging systems typically used, compared to other lagging systems which were believed to have enhanced performance.
Figures 4 and 5 present the results of the insertion losses determined for the 10 pipe lagging systems assessed. These data represent a plot of the noise that is attenuated versus sound frequency. We observed that minimal performance occurs at low frequencies, and the best performance occurs around 2,000 Hz. As most piping-radiated noise from centrifugal compressors occurs between 1,600 to 2,500 Hz, and most piping-radiated noise from flow meters and pressure-reducing regulators occurs between 1,000 to 2,000 Hz, the frequency range of the lagging system performance well covers the frequency range of the noise source. We also observed that a significant variation in performance between different lagging systems occurs around 2,000 Hz, potentially due to a coincidence dip as a result of lagging system stiffness and thickness. Thus, the proper lagging system selection is important.
The results of these tests and other similar studies have determined the following conclusions:
- Typical laggings perform best for mid-and high-frequencies, commencing above 250 Hz;
- Insulation greater than a 2-inch thickness provides little additional improvement;
- Lagging systems with a thicker or heavier jacketing having greater skin rigidity do not perform as well as other systems with thinner or lighter jacketings having less rigidity;
- Lagging systems where the mass layer and the outer jacketing layer are not laminated or which have been delaminated provide significantly poorer performance, as compared to the intended system design having laminated mass barrier and jacketing; and,
- The provision of the septum layer provides worthwhile enhancement for mid and high frequencies, generally above 1,000 Hz.
In a similar manner, Figure 6 presents the results of the insertion losses determined for the three acoustical blanket systems assessed. The relative variations in performance centered around 2,000 Hz. Therefore similar conclusions were drawn as were for the pipe lagging systems.
Further work in these areas is needed, including an investigation of the comparative effects of insulation type and density, and more detailed work regarding the effect of insulation thickness. In conclusion, choosing the acoustical pipe lagging system or the acoustical blanket system that meets the performance requirements, without over-performing, can greatly optimize needed performance versus costs.
Figure 7 presents plant process area or compressor station yard measurements both before and after noise control with acoustical pipe lagging, which shows that noise reductions in the neighborhood of 25 dBA are achievable.
There are five recommended steps for the gas transmission company engineer to follow in order to achieve effective acoustical performance from lagging systems. They are as follows:
- Measure, or quantify from past experience, the amount of noise radiating from the gas piping;
- Assess the performance, from data available from acoustical lagging system insertion loss tests, that will be required to meet the project design goals;
- Issue detailed installation specifications (not performance specifications), including generic material requirements and a listing of approved material suppliers;
- Include specification of specific treatment locations, including pipe supports; and,
- Include provision for inspection and testing, to ensure that all required treatment locations are covered.
Successful acoustical pipe-lagging installations have achieved 25 dBA noise reduction from gas piping, based upon before and after measurements at gas-processing plant and compressor-station facilities. Noise-control engineering on gas piping is a predictable science, where achievable results utilizing lagging systems with known acoustical performance can produce cost-effective noise-control installations.
Frank, L.D., and Milner, G.J., 1995, “Isolation of Major Noise Sources on Natural Gas Turbomachinery Packages,” ASME Paper 95-GT-86.
Holton, K., 1976, “Acoustical Insulation,” Construction Specifier.