Sound Power Measurement of Domestic Boilers

Transcription

1 Sound Power Measurement of Domestic Boilers Dipl.-Phys. R. Edenhofer, Dr.-Ing. K. Lucka, Prof. Dr.-Ing. H. Köhne, Oel-Wärme-Institut ggmbh, Aachen In order to create a standardised method for the measurement of the noise emissions of domestic heating appliances, which facilitates the determination of comparable and reproducible values for such devices, seven European laboratories originated the Boilernoise project with the aim to develop a reliable measurement procedure for this type of appliances. The project consists of several work packages. First a draft for the test method was created, which outlines the general conditions for the measurement. Then detailed studies of a lot of parameters, which potentially influence the measurement results, have been carried out. Based upon the outcomes of these tests the final proposal for the test method will be written and sent to the CEN and manufacturers associations. Introduction In the past decades noise has been recognised as a form of undesirable emission similar to exhaust gases. Beside the prominent sound sources like traffic or airports also domestic appliances have been identified to emit unwanted sounds. One kind of such devices are heating systems, especially those, which are driven by fossile fuels. With noise becoming a topic of increasing interest it is desirable to introduce a standardised method to measure the noise emitted by a boiler in order to be able to compare it to other products and to get an input value for the calculation of local sound pressure levels, which are becoming more and more important in the planning of domestic installations. Due to this background a team of seven European research laboratories created the Boilernoise-Project within the framework of the Competitive and Sustainable Growth Programme, which was originated by the European Commission in order to generally promote the quality and competitiveness of products on the European market. In detail the project partners are the Centre Technique des Industries Aerauliques et Thermiques (CETIAT, France), the Belgian Building Research Institute (BBRI, Belgium), the Building Research Establishment (BRE, England), Gas de France (GdF, France), the Technical Research Centre of Finland (VTT, Finland), Institutio Giordano (IG, Italy) and the Oel-Waerme- Institut Aachen (OWI, Germany). The project is funded by the European Community (Contract Number G6RD-CT ). Its aim is to create a standardised measurement method of the airborne noise emissions of oil- and gas-driven heating appliances in the power range up to 70 kw. The project is scheduled for the duration of three years. The work began in early Elementary conditions The acoustic quantity that characterises the noise emissions of a device in the best manner, is the sound power level. Therefore this has been chosen to be the measurement value to be determined in contrast to former approaches, which tried to achieve a significant comparison of different boilers by the measurement of sound pressure levels at certain positions around the boiler. The problem with this method was the directivity of the sound emissions, which was quite different for the tested devices, so that it was possible that a boiler with a better sound insulation to the direction of the measurement point was rated more silent than another boiler, although it obviously generated more noise. Due to the integral character of sound power measurements this problem will be avoided by the choice of sound power to be the measured quantity.

2 The grade of accuracy of the sound power levels to be measured has been chosen to class 2, which means a standard deviation of about 1.5 db, depending on the frequency band. This accuracy allows a significant distinction of different boilers and can be achieved by reasonable means regarding the measuring equipment and the properties of the laboratory. An investigation of national legislation and standards throughout the nations of the European Community regarding requirements on noise emitted inside living quarters respective by comparable domestic products like refrigerators, dish washers and washing machines points out that accuracy class 2 is quite commonly used for this kind of characterisations. Scientific goals The international standards of the ISO 3740-series and the three parts of ISO 9614 are describing at great length how to measure the sound power levels of any kind of device. The methods described in these standards are well known and proven to produce reliable values, so the method to be created will be strongly based on these procedures. But when it comes to a special type of facility then often problems arise which are not covered by these standards, especially considering the installation and operational settings of the sound source. In regard to domestic boilers there are a lot of parameters that potentially influence the noise emission such as the type and sound power level / db frequency / Hz Figure 1: Results of the preliminary round robin test geometry of the chimney, the wall or floor the boiler is fixed at, the water connections as well as operational factors such as the nature of the fuel, the heat input and the properties of the water circuit. The main goal of the Boilernoise project is to determine, which of these parameters in fact have a non-negligible influence on the noise emissions and to fix these to appropriate values in order to ensure reproducibility of the measurements within the required accuracy. Measuring techniques The project partners are actually applying three different class 2-methods for the measurement of the sound power levels of the boilers, as there are the measurement in a reverberant room according to ISO 3743, parts 1 and 2 [1, 2], and the intensity method described in ISO [3], which are the most common methods for this kind of testing. Through this the validity of the test method concerning these different procedures can be verified. In order to assure that variations in the measurement results coming from the different laboratories are not conditional upon the utilised acoustic technique, a preliminary round robin test was performed with a small fan as a noise source. The results coming from the project partners are outlined in figure 1 and demonstrate a good conformity. Figure 2 depicts the standard deviations of the seven measurements, which are below 1 db except for the frequencies lower than 200 Hz respective those above 4000 Hz. The method OWI is using is the sound intensity procedure according to ISO For this kind of measurement a special sound intensity probe is needed, which consist of two opposed microphones. This arrangement allows the simultaneous measurement of the sound pressure level

3 and its gradient along the axis of the probe. These two quantities enable a dedicated analyser to indirectly determine the sound particle velocity and hence the sound intensity vector including its direction along the probe s axis. The advantage of this technique is that the requirements on the surroundings of the sound source during measurement are much lower than for other procedures, since the probe can distinguish between sound waves coming from the source to be examined and background noise respective reflections out of the surroundings. This way all sound waves which enter the surface at one point and leave it at another point are compensated by the integration over the measurement surface, since they contribute to the final result with a positive and a negative fraction of equal size as it is depicted in figure 3. In theory the sound power level can be determined in any kind of environment, but it shows up that a dedicated laboratory makes the measurement much easier (see figure 4). Nevertheless the installation of a dedicated laboratory for sound intensity measurement usually is exceedingly cheaper and easier than for example the construction of a reverberant room according to the requirements of ISO The main disadvantage of the sound intensity measurement compared to other methods in addition to the more complex measurement surface sound source standard deviation / db Sounds contributing to the final result Sounds not contributing to the final result Figure 3: Sound intensity method. 4 3,5 3 2,5 2 1,5 1 0, external noise frequency / Hz Figure 2: Standard deviations of the preliminary round robin test equipment is that depending on the sound source the determination of one sound power spectrum with high accuracy potentially needs more time, since the number of measuring points generally is higher than for the reverberant method. In order to reach a class 2-result in all frequency bands of interest (third-octave bands from 50 to 6300 Hz), up to 96 measuring points around the boiler have been realised at OWI (see figure 4). Using sophisticated commercial software it takes about half a minute to measure one point so that the overall time for a complete measurement is nevertheless contained within reasonable limits. Test programme The main part of the project consists of the variations of the installation and working conditions of the boiler and the determination of their influence on the emitted sound power levels. Taking into account the knowledge from former tests of the project partners the following variables have been chosen for the parameter variations: - Chimney draught and geometry - Duct length and configuration - Fixing of the boiler - Type of wall or floor - Fluid connections - Nature of gas (gas boilers only) - Heat input - Water temperature - Settings of the pump - Water static pressure

4 Figure 4: Test room with boiler inside measuring mesh - Setting of the burner (air draught boilers only). The boilers that are used during these tests are chosen to cover the most common products sold on the European market. In detail they are - 2 wall hung atmospheric boilers - 2 wall hung room sealed atmospheric boilers - 2 wall hung premix burners - 2 floor based atmospheric boilers - 4 floor based boilers equipped with a fuel oil forced draught burner. Every parameter variation is performed at least with two different types of boilers. Crucial parameters like the heat input are tested for all kinds of boilers. OWI s task is the examination of two floor based fuel oil burners. The parameters, which are tested on these devices comprise installation conditions concerning the chimney and the use of a vibration isolation as well as operational parameters like the heat input, properties of the water circuit and air ratio. Test results The test programme described in the previous section and the analysis of the measured data is not completely finished, so the final results are not yet available. Nevertheless some trends concerning the influence of certain parameters on the noise emissions can already be recognised. The first result obtained during this work package does not concern the parameter variations, but the intra-laboratory reproducibility, as it could be shown that two consecutive measurements on the same boiler generally produce the same results, even if the boiler has been reinstalled between the two tests. These experiments proof once more that the chosen acoustical methods are reliable. A second result concerns the scope of the acoustic measurement. Taking into account the spectral information gathered during the tests, it became obvious that the main part of the noise emissions of the examined boilers is in the frequency range between 100 and 5000 Hz. This means that the overall level and even more important the A-weighted overall level are not severely influenced by the frequency bands beyond this range. Assuming that the chosen boilers are representative for the available range of products, this means that the analysis of the extreme low as well as the very high frequencies will not be necessary for most devices. Applying this limitation the measurement procedure can be simplified in many cases, as for example utilising the sound intensity method the use of a second spacer suited for the lowest frequency is not necessary. The test results for four floor standing boilers of different types are depicted in figure 5. The actual parameter variations are presented in the following including yet available results: Chimney draught and geometry Since the chimney is directly connected to the boiler and its combustion chamber, the combustion noise can spread to the chimney which therefore becomes a sound source itself. Additionally the combustion chamber and the connected chimney constitute a Helmholtz resonator with an according

5 Figure 5: Sound power spectra of different floor standing boilers [4] natural frequency that depends on the chimney length. Therefore the potential incitement of Helmholtz oscillations can affect the sound emissions of the complete device. The variations of the chimney include changes of the overall length, the number and location of the bends, the material (plastic instead of steel) and the chimney draught, which was adjusted by a flue gas damper. None of the related experiments display a severe impact on the overall sound power levels. The final proposal will recommend minimising the overall chimney length inside the test room. Duct length and configuration For room sealed boilers this parameter is of importance, since the pressure drop of the ducts is compensated by an adjustment of the rotational speed of the fan in order to ensure a constant mass flow of combustion air. The test results indeed show a nonnegligible influence of the length and configuration of the ducts. Regarding the final measurement method this parameter has thus to be fixed to a certain assembly, which has been chosen to be a C1- configuration with short pipes. Fixing of the boiler Through the fixation of the boiler vibrations can be spread to the wall or floor the boiler is mounted to, which then can radiate noise, too. The use of vibration isolation has been examined for wall hung and floor based boilers. The results show no discernible influence on the noise emissions. In the final proposal the use of this kind of application will not be recommended. Type of wall or floor The above mentioned spreading of vibrations from the boiler to the wall can of course also depend on the type of the wall, especially its weight. Walls with different mass per square meter, realised by the use of a plaster wall on the one hand and a wall built of concrete blocks on the other hand, have been investigated with regard to this topic. The available results show an enormous deviation of the measured sound power levels. Further tests to specify these effects have to be done. The final measurement procedure will include a clear definition of the weight of the wall due to its strong influence. Of course this means that the sound power value measured according to this procedure can deviate from the on site value, if the boiler is mounted to a different kind of wall. Fluid connections Also the water connections can potentially be influenced by vibrations from the boiler or the flowing water inside the pipes. Rigid and flexible pipes have been tested for their impact on sound emissions, but so far no effect could be detected. Due to the easier installation the use of flexible pipes will be recommended. Nature of the gas Whereas the properties of fuel oil are more or less constant, a variety of gases with different compositions and calorific values are used all over Europe, depending on the

6 source of the gas. Since the combustion itself is influenced by the relative amount of the different constituents, this is an important parameter to test. Several test gases, as there are G20 (methane), G25 (Groningen gas), G30 (butane) and G31 (propane) have been used for these tests. The results display a strong influence on noise emissions with deviations up to 3 db(a). For the final test procedure the use of G20 will be required, but in the same manner as for the type of the wall or floor the boiler is fixed to, it has to be stated that the obtained sound power levels can be different on location, depending on the locally used type of gas. Heat input A crucial variable to examine is the heat input. All types of boilers have been tested for their dependency on this parameter and not surprisingly most of the measurements show severe differences of the obtained sound power levels. Variations up to 3 db(a) have been observed for only slight changes of about 5 to 10 %. The differences between nominal and reduced heat input can amount up to 8 db(a). The final proposal will recommend a stringent definition of the heat input set up during the tests. Following the requirements for efficiency tests a maximum deviation of 2% from the nominal heat input will be proposed. Water temperature Depending on the water temperature local boiling can appear inside the boiler, which potentially produces noise and therefore can affect the overall sound power level. The investigations concerning this parameter have been executed using water temperatures of 80/60 C and 60/40 C. The differences of the measured sound levels are lower than 1 db(a), so this is not an important factor. According to the efficiency tests a temperature of 80/60 C will be recommended. Settings of the pump A potential source of noise emissions is the water pump and the circulation speed of the water. Changes of the pump speed and, resulting from this, the water flow and pressure drop revealed a slight, but nonnegligible effect on noise emissions, so it will be important to fix these settings in the final proposal. Water static pressure Besides the water temperature the appearance of local boiling can also depend on the water pressure. The tests on this parameter have been done at 1.5, 2.0 and 2.5 bar static pressure, but no vital influence could be detected. The final proposal will recommend setting the water pressure as it is defined in the technical documentation of the boiler or 2.0 bar. Setting of the burner The air excess used for the operation of the burner is another factor which influences the combustion itself and therefore has been examined. For the investigated air draught burners the air ratio can either be changed by the adjustment of a shutter or the head of combustion. Both variations have been executed with deviations of 5 to 10 %. The results show only small changes of the noise emissions. The recommendation for the final measurement procedure will be to set the air ratio as it is required by the technical documentation. The behaviour of the different types of boilers used for the parameter variations displayed big differences regarding the effect of the variations on the emitted sound power levels. The boilers tested at OWI are equipped with fan assisted fuel oil burners. This type of device generally does not seem to be severely affected by any of the changes in the tested installation and operational conditions. The reason for this effect is that a big part of the emitted noise is caused by the fan, the rotational speed of which is constant, since variations of the combustion air mass flow are accomplished by an adjustment of an air shutter. The noise of the combustion mainly contributes to the low frequencies, which are of low im-

7 sound power level / db frequency / Hz nominal heat input only fan Figure 6: Sound power spectra of a floor based boiler with a forced draught burner with and without combustion L A (CEN/TC57/WG6) with the participation of OWI-Aachen. Its scope is much wider than that of the Boilernoise project, as it also concerns solid fuel burners (pellets), does not have any limitations related to heat input and includes a second part regarding the measurement of sound power inside the chimney. portance regarding the calculation of the A- weighted overall level, which is the most important quantity for this kind of test. The measured sound power spectra of one of the tested boilers with nominal heat input and only the fan working without combustion are compared in figure 6. Future work With the parameter variations and the examination of their influence on the emitted noise levels being nearly completed, the draft for the final proposal is now going to be written including the above introduced requirements and recommendations. The reliability of the proposed test procedure will be checked by a final round robin test. Therefore two of the tested boilers will be sent to the project partners sequentially in order to compare the sound power levels measured in the different laboratories. When this test is accomplished with satisfying results, the proposal for the measurement method will be finished and sent to CEN and EHI. Other projects Contemporaneously to the Boilernoise Project the European Heating Industry (EHI), the European organisation of boiler and burner manufacturers, submitted a proposal concerning noise characterisation of boilers and burners to the CEN. This draft is currently discussed by the Work Group 6 of CEN s Technical Committee 57 Conclusion The intermediate results from the Boilernoise project suggest that a standardised measurement of noise emissions generating reproducible and comparable results is possible. The application of the final proposal based on the investigations executed during the project to appropriate standards like the EHI-proposal that is currently discussed, has to be accomplished by the relevant technical committees of the CEN. References [1] ISO , Acoustics; Determination of sound power levels of noise sources; engineering methods for small, movable sources in reverberant fields; part 1: comparison method in hard-walled test rooms, 1994 [2] ISO , Acoustics; Determination of sound power levels of noise sources using sound pressure - Engineering methods for small, movable sources in reverberant fields - Part 2: Methods for special reverberation test rooms, 1994 [3] ISO , Acoustics; Determination of sound power levels of noise sources using sound intensity; part 1: measurement at discrete points, 1993 [4] P. Gonzalez: BONO document 32-V2, Draft of WP2-Task 2.3 synthesis