3 Gholmohamadi

JRHS 2010; 10(1): 22-30

Copyright © Journal of Research in Health Sciences        

Evaluation of Noise Propagation Characteristics of Compressors in Tehran Oil Refinery Center and Presenting Control Methods

Rostam Golmohammadia, Mohammad Reza Monazzamb*, Maryam Nourollahia, Ali Nezafatb, Samaneh Momen Bellah Fardc,

a Department of Occupational Hygiene, School of Public Health and Research Center for Health Sciences, Hamadan University of Medical Sciences, Hamadan, Iran

b Department  of Occupational Hygiene, School of Public Health, Tehran University of Medical Sciences, Tehran, Iran

c Department of Environmental Science, Graduate School of the Environment and Energy, Science and Research Branch, Islamic Azad University, Tehran, Iran

*Corresponding author: Monazzam MR (PhD), E-mail addresses: mmonazzam@gmail.com

Received: 19 December 2009, Revised: 17 March 2010, Accepted: 26 May 2010, Available online: 21 June 2010        

Abstract        

Background: The adverse effects of noise are well known and noise problems due to industrialization of communities are increasing over the time. Oil industries due to the process and nature of production; contain many noise sources such as compressors, turbines, and pumps, which cause excessive noise exposure. The objective of this study was to evaluate the noise characteristics of compressors in Tehran Oil Refinery and study on visible control measures. 

Methods: To get to the appropriate control method, procedures such as basic theories, measuring sound parameters, frequency analysis, related diagrams and noise propagation schemes due to the measurement results, equivalent noise exposure level (Leq(8h)) and exposure noise dose and  technical specification of compressors are considered in this paper. Considering field and analytical re­sults, module enclosure with particular specifications (like absorbent layer, specific wall, window and door design etc.) is predicted to be the best control method. 

Results: Calculation results of multiple layer density of the enclosure (W = 16.5 kg/m2) and needed density for the dominant frequency of the source (W = 12 kg/m2) demonstrated that the designed enclosure satisfies the goal.

Conclusion: Results of designing sandwich layers module demonstrated that install­ing the designed enclosure causes 20 dB(A) reduction in total sound pres­sure level of the sources dominant frequency.

Keywords: Oil refinery, Noise enclosure, Dominant frequency, Noise exposure        

Introduction

Nowadays harmful noise effects are well known and noise problems due to in­dustrialization of communities are in­creasing over the time. Of harmful noise effects can point to masking noise, relation with sec­ond kind diabetes and some psychological dis­orders [1]. In addition, effects on visual organ (interference in colla­tion control and detecting items and reduce eye reaction to light) and equi­librium system (Nau­sea, confusion and walking interference) are other harmful effects of exces­sive noise expo­sure.

Compressors are one of the main noise sources in industries, which cause noticeable damages to working community annually. Oil industries due to the process and nature of production; contain many noise sources such as compres­sors, tur­bines, and pumps, which cause exces­sive noise. In this case, Nassiri et al. reported that noise exposure in the studied oil fields were far more than Iranian permissible levels [2].

On another study, excessive noise exposure was detected and so necessitated engineering noise controls were outlined [3]. Evaluation of noise pollution on oil field determined that compres­sors are one of the main noise sources in the field. In addition, this study resulted that apply­ing enclosures to some noise sources would cause 14 to 19 dB noise reduction [2].

Esmail Zadeh et al. studied on noise pollu­tion of compressed air conditioning unit on a fac­tory and showed that compressors noise were excess of the allowable limit and control strate­gies on all the sources such as air outlet pipes and air inlet vents are the only way for reduc­ing the noise and installing silencer and muffler sound is useful in the field [4]. From this con­text, Hakimi by applying a module in the air outlet has estimated a 20 dB reduction in the sound level [5].

Other survey done by Speich on controlling compressors' noise showed that for its noise con­trol program noise should be evaluated on the view point of technical and spectral specifi­ca­tions [6].

Knight et al. denoted enclosure compressors with an enclosure composed of soft synthetic layers instead of applying hard layers would reduce the noise of compressors. Applying the acoustic enclosure reduces the compressor noise level by 9 dB [7]. Other evaluation for re­ducing compressors noise to permissible level speci­fied that acoustic enclosure as the best method [8].

Joseph et al. survey entitled control of shear cutting noise effectiveness of enclosures showed that ignoring structural paths, which generated sound leaks from the controlling de­vice, re­duces the efficiency of enclosure. In addition, it was determined that precise recog­nition of the noise source and the field surfaces plays a great role in assessing the acoustic effi­ciency of de­vice [9]. Applying multi purpose enclosure can reduce sound pressure level up to 40 dB [10].

Knight et al. study conduced to soft syn­thetic multi layer and absorbent module design in­stead of applying hard layers for enclosure air compressor [7]. This study showed that install­ing soft synthetic multi layer and absorbent ma­terial could reduce sound pressure level about 9 dB(A). Nathak et al. stated that the best result is achieved by applying soft synthetic multi layer and absorbent materials [8].

The control measures presented in this study are in accordance with similar studies stated be­fore [8]. In this study, hydrogen gas pressure en­hancer compressors were applied in the field (3 devices including A, B, C, 2C-401) to pro­vide needed hydrogen in the unit with increase hy­dro­gen pressure received from hydrogen gener­ated unit or catalyst exchange unit around 2854 psig. Compressors, which were used in Isomax unit of Tehran Oil Refinery center, were com­posed of three parts: main compressor body (first section), re­du­cer gear (second sec­tion), and turbine (third section).

In this paper, firstly the result of evaluated characteristics of noise generated by the com­pres­sors and then a recommended control me­thod will be introduced.

Methods

First cooperating with HSE department of Te­hran Oil Refinery center, basic information such as plans, noise source place, and some other tech­nical information like device longev­ity, size, com­ponent, and task or function of the device was col­lected. In the plan of the unit, most im­portant noise sources were pointed out, circled, and coded. In addition, number and task position of workers ex­posure to excessive noise were de­termined.

Second step was to measure sound parame­ters like SPLrms, SPLmax on specified zone. TES-1385 sound level meter was applied in this study and it was calibrated by B & K 4231 standard calibra­tor. For this purpose, 21 meas­uring points, which are shown in Figure 2, were considered. The overall dimension of the in­vestigated unit was 32 by 15 meter. It is worth noting that the measuring points were chosen by the net method and in fact, they are the centers of squares of 5 by 5 meter in the field. In this step, frequency analysis was done on C-weighted and results were given to excel and SURFER 7 software. Then related diagrams and noise propagation schemes were obtained. Stan­dard method ISO 9612 (1997) is used for meas­urement [11]. Finally, in this step, review­ing rou­tine procedure in the unit and following personal exposure time, equivalent noise expo­sure level (Leq (8h)) and noise exposure dose were calculated.

On third step, to determine the dominant fre­quency of compressors by theoretical method, their technical specifications such as rate per min­ute were evaluated and estimated. Apply­ing re­sults of compressor's noise measuring, sound contours of the unit was drawn (Figure 1).

Finally, a control measure for the dominant noise source using both experimental and theo­retical findings is presented and the overall per­formance of the designed enclosure is pre­dicted.

Figure 1: Noise contours in Isomax unit

Results

Compressor's specification of Isomax unit

There were three reciprocating compressors in Isomax unit. Two compressors work simulta­neously and another one is left stand by. The results LArms and LAmax of 21 measuring points are shown in Table 1. The locations of the above-mentioned measuring points are showing in Figure 2. Using the above results, the crest factor was calculated by subtracting rms sound pres­sure level of maximum sound pressure level. The results are also shown in Table 1.

Figure 2: Noise zone plan of Isomax compressors, Compressor A is off and both compressors B and C aon

Table 1: The A-weighted sound pressure level and maximum sound pressure level along with the crest fac­tor of Isomax compressors (dB)

Measuring point

LArms

LAmax

CF

1

84.4

98.1

13.7

2

84.6

99.6

15

3

86.1

100

13.9

4

88.4

103.5

15.1

5

88.7

103

14.5

6

89.6

104.5

14.9

7

89.8

104.8

15

8

84

99.2

15

9

88.1

102

14

10

88.4

104.1

15.7

11

90

105.4

15.4

12

91.2

107

15.8

13

92

108

16.4

14

90.6

107

16.6

15

84.7

100

15.4

16

85.5

102.2

16.7

17

89.2

104.6

15.4

18

91.7

108

16.3

19

91.5

107.4

15.9

20

91.6

108.9

17.3

21

89.7

105.9

16.2

The noise was analyzed in octave band center frequency band in six measuring point. The re­sults of turbine, gear, and compressor's side in Isomax unit are shown in Table 2 and Table 3. Point numbers is the ta­bles are shown in Figure 2.

Table 2: Octave band center frequencies of turbines' side in Isomax unit

Measuring point Frequency(Hz)

Points

1

4

6

63

80.4

84.7

84.4

125

79.1

83.6

82.7

250

78.4

81.4

82.3

500

75.3

77.6

79.8

1000

76.7

82.5

81.6

2000

72.7

77.8

79.9

4000

69.2

76.1

77.5

8000

62.7

69

71.2

 

Table 3: Octave band center frequencies of gear and compressor's side in Isomax unit

Measuring point Frequency (Hz)

Points

16

18

20

63

79.5

79.4

78

125

83

83.1

82.1

250

81.3

82.2

81

500

78.2

79.8

78.4

1000

82.5

81.7

82

2000

81.8

81.4

81.8

4000

87

86.8

87.3

8000

84

83.3

84

Dominant noise frequency of Isomax compres­sors

The rotation speed of the three compressors component (including compressor, gear and tur­bine) of Isomax unit was measured using a RPM meter. The rotation speed of the turbine was 4000 RPM with 49 blades and the rotation speed of gears and compressor were 1/13 of that of the tur­bines with 8 blades. Therefore using the well-known equation (see equation 3-1 Ref. No. 12) the domi­nant fre­quency for the turbine 3266.6 Hz and that is 41 Hz for the gears and compres­sor. In an oc­tave band analysis the above fre­quencies is lo­cated respec­tively in 4000 and 63 Hz center frequencies. There­fore, very good agree­ment can be seen by com­pari­son between the above predic­tion results and the field measure­ment results as shown in Table 2 and Table 3.  

Equivalent workers' exposure level

To determine the equivalent exposure level, the exposure time of related workers with differ­ent sound pressure levels were specified. Work­ers usually work on three shifts with du­rations of 2.5, 1 and 4 h and the sound exposure levels in each section are respectively 71, 65 and 92 dBA. Therefore the LAeq(8h) was found to be 89.5 dBA and the received dose according to Iran standard levels (85 dB per 8 h work shift and criterion shift parameter of 10) was 282%. There­fore, based on Iranian standard limits, al­lowable exposure limit for workers is just 2.8 h per day.

Control measure

Designing a noise enclosure for the Isomax com­pressors are as follows:

  1. Critical frequency of main insulator of the enclosure (2mm steel)

For designing enclosure, it is firstly impor­tant to determine the critical frequency of the main insulator, which is considered 2mm steel.  By applying the well-known equation of cal­culating the critical frequency (see equation 6-17 Ref. No. 12), this frequency is predicted to be 8978 Hz which is far above the dominant frequency of our main noise source.

  1. Layout and specification of needed module sandwich layers 

A: Absorbent: For the purpose of this design, a layer of slag wood with 2.5 kg/m3 surface den­sity and 25 mm thickness applied as an absorbent for the considered frequency.

B: Frame: A wooden frame with15 mm thickness and surface density of 7 kg/m3 was used

C: Insulator: A 2 mm steel with surface density of 17 kg/m3 was applied in the center­line of the pan­els for insulating the structure born noise.

D: Filler: A 20 mm polyurethane foam was used as a filler within the panels.

E: External surfaces: A 9 mm chipboard with surface density of 7 kg/m3 was used for covering the external sur­faces of the enclosure

F: Door: A common gash door with 43 mm thickness and surface density of 9 kg/m3 was used for the entrance of the enclosure. The dimen­sion of the door was designed to be 1.8 by 0.7 m.

G: Windows: 8 double layer glass windows with 9mm thick­ness and surface density of 7 kg/m3 and di­men­sion of 1.7 by 1.7 m were used for the enclo­sure.

  1. Calculating the least surface density for the dominant frequencies (63, 4000 Hz)

The least surface density for the dominant frequencies was calculated to be 12 kg per cu­bic meter [12].

  1. Size and area of Enclosure

The dimensions of the designed enclosure were 3×5×5.5 m and as it was mentioned 8 windows with size of 1.70×1.70 m and a 1.80×0.70 m door were used in the design. To­tal surface of the enclosure was 118 square meter and using the following equation the overall panel density was found to be 16.5 kg per cubic meter.

         (1)                                                                                        

Where Wi and si  are the surface density and the area of each panel component respectively (Figure 3).

Figure 3: Detail structure of the main enclosure panel

  1. Frequency analysis, and TL and NR

Using the above mentioned field measure­ment results [92 dB(A)] and Iranian noise expo­sure limit [85 dB(A)], it is easily found that the total noise reduction required is 12 dB [92 dB(A) - 85 dB(A) +5 dB(A)] and in that of dominant fre­quency is 16 dB.  The 5 dB is mostly added to achieve the practical results.

Total noise reduction achieved by installing the designed enclosure was calculated about 20.1 dB, which is gained by subtraction of total outdoor and indoor noise levels. In this case, the overall noise level inside the enclosure was measured to be 92 dB while the noise level out­side the enclosure was estimated to be 71.9 dB. Figure 4 provides sound pressure level varia­tions before and after installing the enclosure.

Figure 4: Comparison between field sound pressure level before and after installing the enclosure

The architectural plans and related details were designed by AutoCAD software and are shown in Figure 5 to Figure 7.

Figure 5:  A detailed three dimensional plan of the designed enclosure (the unit of undefined numbers are based on cm)

Figure 6: A cross section of the designed enclosure for Isomax unit


Figure 7: The horizontal plan of the designed enclosure for Isomax unit

Discussion

Results of field measurement of compres­sors noise (Table 1) indicates that in 17 of 21 stations evaluated, sound pressure level was above 85 dB(A) and just in 4 stations, the meas­ured noise was just 1 dB(A) bellow Iranian stan­dard. These four stations were close to the stand by compressors (Figure 2).

Results of sound plan presented in compres­sors zone showed that the same level contours which placed between two running compres­sors contains maximum sound pressure level [91.5 dB(A)]. Also evaluating sound counters plan illustrated that sound pressure level around com­pressors placement was above Iranian standard levels, which is in consistent with pre­vious stud­ies [2, 4] (Figure 1).

Equivalent exposure level of workers indi­cated that workers noise exposure was above allowable level and due to standard limits, al­lowable working shift should be reduced to just 2.8 h, also daily exposure dose results showed that exposure dose was about 282% and there should applied control methods to reduce the effects of harmful noise.

The results of both theoretical prediction and field frequency analysis demonstrated that the dominant frequency at the turbine side was 63 Hz, while that was 4000 Hz for gear and compressor side of the noise sources (Table 2 and Table 3).

To control the compressor noise and reduce its harmful effect with less noise exposure level, it was decided to design and apply an insulating enclosure.

Considering the large scale of the needed enclosure for controlling the noise and also with regard to results achieved by frequency evaluation limit of insulation applied for the enclosure, it was demonstrated that the critical frequency of a 2 mm steel insulator was far above the dominant noise frequency of the source (fc = 8978 Hz). So the steel panel was used as a main insulator of the enclosure. It is worth noting that using steel panel with thick­ness more than 2 mm will reduce the critical fre­quency bellow the dominant frequency of the source. This obviously reduces the per­formance of the control measure.

Conclusion

To avoid the multiple reflections from the hard surfaces of the insulators, utilizing an effi­cient absorbent material is unavoidable.

In addition, results of evaluating absorbent coefficient of the enclosure and minded trans­mission loss in dominant octave band frequen­cies indicated that applying a layer in module designing, would not get to the considerate trans­mission loss. Therefore, a multiple layers (sand­wich layers) were applied.

Sound absorbent evaluation also showed that the slag wool was a suitable absorbent for the control measure. Fiberglass was also an­other suitable choice but using perforated sheet or metal lace is necessary for installing a fiber­glass on a panel, which in this study and due to the large size of the enclosure, using this ab­sorbent increase metal surface area of the en­closure and so increase the reflection of the sound inside the enclosure which leads to in­crease the sound pressure level inside the en­closure.  To reduce the noise reflection of the outside of the enclosure, a 9 mm chipboard is used as a last layer.

Results of multiple layer density of the en­clo­sure (W = 16.5 kg/m3) and needed density for the dominant frequency of the source (W = 12 kg/m3) demon­strated that the designed enclosure satisfies the goal. In fact, the findings listed above are in agreement with the results of the other studies in the literature [7-9].

Results of designing sandwich layers mod­ule demonstrated that installing the designed enclo­sure causes in 20 dB(A) reduction in total sound pressure level of the sources dominant fre­quency.

Finally, it should mention that this article is only presented according to prediction and de­signs and the results should be validating with field research.

Acknowledgements

The authors would like to thank the HES department and managing of Tehran Oil Refinery and Hamedan University of Medical Sciences for financial sup­port of this study. The authors declare that they have no conflicts of interest.

References

  1. Golmohammadi K. Evaluation of impact noise in pneumatic breaker. Sci J Hamadan Uni Med Sci. 1996; 5(3):17-25.
  2. Nassiri P, Ahmadi N. Evaluation of noise pollu­tion in Lavan Oil Field and determination of source enclosure for noise reduction [M Sc.Dis­sertation]. Tehran: Tehran University of Medi­cal Science, School of Public Health; 2004.
  3. Gholshah H. Engineering noise control in oil industry. 1st Congress on noise and its effect; 1997; Mashahd.
  4. Esmail Zadeh E, Hakimi HA, Farrokhi MN. Measurement and evaluation of Noise pollution at grinding and flotation units of concentration plant. 6th Congress on safety, Health and Envi­ronment in Mines and related industries; 2006; Tehran.
  5. Hakimi HA, Ali Mohammadi T, Farrokhi MN. Modeling of noise propagation in outside area of Sarcheshme new concentration for determin­ing best noise control method by sound plan software. 6th Congress on safety, Health and Environment in Mines and related industries; 2006; Tehran.
  6. Speich. CF, Elements of compressor noise con­trol. Int J Refriger. 1981;4 (5):281-7.
  7. Knight SB, Evans JB, Himmel CN. Air Cooled Chillers With Rotary Helical (Screw) Compres­sor at Hospital With Impact on Patient Rooms Residential, Neighborhoods, and Open Park. The 2001 International Congress and Exhibi­tion on noise control Engineering; 2001 August 27-30; Netherlands.
  8. Nathak S, Puranik A, Schut J, Wells L, Rao MD. Study of Noise Transmission from an Air Compressor, Noise_ Con Minneapolis, Minnesota. Acoustical Society of American Journal. 2005;118 (3):1872-3.
  9. Joseph CSL, Speakman C, Williamson HM. Con­trol of shear cutting noise effectiveness of enclosures. Appl Acous.1999; 58:69-84
  10. Chiu MC, Chang YC.  Numerical studies on venting system with multi-chamber perforated mufflers by GA optimization. Appl Acoust. 2007;69(11):1017-37.
  11. ISO, Guidelines for Measurement and assessment of exposure to noise in working environment, International Standard Organization, ISO-9612, 1997.
  12. Lewis HB and Douglas HB. Industrial Noise Control, fundamental and applications. 2nd ed.   New York: Marcel Dekker Inc; 1994.


JRHS Office:

School of Public Health, Hamadan University of Medical Sciences, Shaheed Fahmideh Ave. Hamadan, Islamic Republic of Iran

Postal code: 6517838695, PO box: 65175-4171

Tel: +98 81 38380292, Fax: +98 81 38380509

E-mail: jrhs@umsha.ac.ir