# Accuracy limitations for low-velocity measurements and draft assessment in rooms.

INTRODUCTIONAirflows in ventilated spaces are typically turbulent, with velocities varying in magnitude, direction, and fluctuation frequency. Mean air velocities from below 0.05 m/s up to 0.6 m/s, turbulence intensities from less than 10% up to 70%, and frequency of velocity fluctuations as high as 2 Hz that contribute up to 90% of the measured standard deviation of fluctuating velocity (RMS) have been identified in the occupied zone of rooms (Finkelstein et al. 1996).

Accurate measurement of a low air velocity is difficult. Several methods, such as thermal anemometry, laser Doppler anemometry, flow visualization, sonic anemometry, etc., may be used to measure air velocity in rooms. At present, a thermal anemometer with an omnidirectional velocity sensor is most often used in practice due to its low price and easy and convenient operation. These anemometers, in fact, measure speed but not velocity.

Draft is one of the most frequent complaints indoors. Draft is defined as an unwanted local cooling of the body caused by air movement. Draft discomfort increases when the air temperature decreases and the mean speed and turbulence intensity increase (Fanger et al. 1988). The same airflow from the back causes more draft discomfort than that from the front (Mayer and Schwab 1988; Mayer 1992; Toftum et al. 1997). Research also identifies that the frequency of the speed fluctuations is important for the sensation of air movement (Fanger and Pedersen 1977). Room airflow is felt most uncomfortable when the equivalent frequency of the speed fluctuations is around 0.5 Hz (Zhou and Melikov 2002; Zhou et al. 2002).

Present standards acknowledge the importance of draft discomfort and define requirements for either summer and winter allowable maximum mean speed in the occupied zone or maximum percent of dissatisfied occupants due to draft (EN 1995; ASHRAE 2004, 2005). The percentage of occupants in a space that may experience draft discomfort is assessed by a draft rating index, %, calculated by the following equation:

DR = (34-[t.sub.a])([[V.sub.mean]-0.05).sup.0.62](0.37 . [V.sub.mean] . Tu + 3.14) (1)

or

DR = (34-[t.sub.a])([[V.sub.mean]-0.05).sup.0.62](37 . [V.sub.RMS] + 3.14) (2)

In these equations, [t.sub.a] ([degrees]C) is the air temperature, [V.sup.mean] (m/s) is the mean speed, [V.sub.RMS] (m/s) is standard deviation of speed fluctuations, and Tu (%) is the turbulence intensity of the flow. The mean speed is defined by the instantaneous speed averaged over an interval of time, while the turbulence intensity is the standard deviation of the speed divided by the mean speed, Tu = [V.sub.RMS]/[V.sub.mean]. The above equation is valid when [V.sub.mean] is higher than 0.05 m/s; for smaller than 0.05 m/s, [V.sub.mean] = 0.05 m/s should be used; for DR< 100 %, DR = 100% should be used. This equation is based on results from human subject experiments (Fanger et al. 1988). The importance of airflow direction and frequency of speed fluctuations on draft sensation is not addressed in the standards.

Apart for evaluation of occupants' thermal comfort, air velocity measurements in rooms are needed with regard to air quality research, validation of CFD predictions of room air distribution, balancing and commissioning of air-conditioning systems, etc.

The accuracy of measuring mean speed, standard deviation of speed, and turbulence intensity by a low-velocity thermal anemometer (LVTA) can be affected by several error sources defined later in this paper. Comprehensive analyses for the impact of the error sources on the measurement accuracy of the mean speed and the standard deviation of speed have been performed (Loomans and Schijndel 2002; Popiolek et al. 2007; Joergensen et al. 2004). Measurements without considering the impact of these factors on the accuracy of speed measurement and draft assessment may lead to erroneous results (Melikov and Sawachi 1992; Popiolek et al. 1998).

Requirements for the characteristics of instruments measuring low air velocity are given in the present indoor climate standards (EN ISO Standard 7726, Ergonomics of the Thermal Environment--Instruments for Measuring Physical Quantities [EN 1998]; ANSI/ASHRAE Standard 55-2004, Thermal Environmental Conditions for Human Occupancy [ASHRAE 2004]; ANSI/ASHRAE Standard 113-2005, Method of Testing for Room Air Diffusion [ASHRAE 2005]). The standards define range of velocity, required and desirable accuracy of its measurement, and measuring time, as well as requirements regarding dynamic response of the anemometer and directional sensitivity of the velocity transducers. However, the standards do not specify requirements for design of the velocity transducer or for the sampling rate. Several other requirements that will ensure low uncertainty of velocity measurements are not specified in the standards either. ASHRAE (2004) requires DR < 20%, while ISO Standard 7730, Moderate Thermal Environments--Determination of the PMV and PPD Indices and Specification of the Conditions for Thermal Comfort (EN 1995a), requires DR < 15%. The accuracy of determination is not known and is not defined in the standards. CEN Report 1752, "Ventilation for buildings: Design criteria for the indoor environment" (CEN 1998), defines three categories of indoor thermal environment, A, B and C, with draft ratings of 15%, 20%, and 25%, respectively. However, it is questionable whether these requirements are realistic, since the uncertainty in determination of draft rating is not known.

In this paper, the uncertainty of mean speed, standard deviation of speed, and turbulence intensity measurement with LVTAs due to the impact of all important error sources is identified. Also identified are realistic requirements for the characteristics of LVTAs with omnidirectional velocity sensors, which will ensure the highest possible accuracy of mean speed and standard deviation of speed measurements in practice. The minimum uncertainty in draft rating determination with LVTAs is defined.

REQUIREMENTS FOR LVTAS

Tested LVTAs

Four low-velocity thermal anemometers available on the market, A, B, C, and D, were selected to be used during this study. The anemometers had omnidirectional velocity sensors with temperature compensation circuits. The transducers, which consisted of a heated (velocity) sensor and an unheated (temperature) sensor, were of quite different design (Figure 1). The velocity sensor of anemometer A is designed as a spherical body with a diameter of 3 mm. The body of the sensor is made of quartz and coated with a heated nickel layer. The velocity probes also have a separate sensor for measuring air temperature. The overheating temperature of the velocity sensor A is 25[degrees]C (the overheating temperature is the difference in the temperature of the heated velocity sensor and that of the air). Transducer B has a velocity sensor with an ellipsoid body made of a highly insulating thermal foam material, ground to shape and supported by a thin stainless-steel tube. The ellipsoid carries three coils of electrically heated nickel wire coated with a thin layer of white epoxy enamel. The diameter of the ellipsoid at the location of the heated wire is approximately 5 mm. The overheating temperature of the velocity sensor is 13[degrees]C. The velocity transducer of anemometer C is a small thermistor with a diameter of less than 1 mm. The overheating temperature of the velocity sensor is 10[degrees]C. Anemometer D has a spherical velocity mass sensor of 2 mm diameter made of enameled copper wire molded into a sphere. The overheating temperature of this sensor is 25[degrees]C. The four probes have an unheated sensor used to correct the measured speed when the air temperature is different from that of the air temperature from the calibration.

[FIGURE 1 OMITTED]

The design of the four anemometers was previously modified by the manufacturers in order to improve their dynamic response. The aim was to reach an upper frequency of at least 1 Hz, which would be sufficient to capture the significant speed fluctuations occurring in rooms (Finkelstein et al. 1996). The upper frequency is defined as the highest frequency of sinusoidal velocity fluctuations up to which the thermal anemometers should be able to measure the standard deviation of the velocity with an accuracy of [+ or -] 10% (Melikov et al. 1998a). The dynamic response of the four anemometers was tested under laboratory conditions with experimental facilities and a test procedure similar to that described by Melikov et al. (1998b). The anemometers were calibrated in the same wind tunnel prior to testing. Figure 2 shows the standard deviation ratio as a function of the frequency of the periodically fluctuating velocity obtained from the tests with the anemometers. The standard deviation ratio is defined as the standard deviation of speed measured by the tested anemometers divided by the standard deviation of speed measured by a reference anemometer (in this case, a hot-wire thermal anemometer). The results show that the four LVTAs had an upper frequency higher than 1 Hz. The results in Figure 2 also show that the differences in the standard deviation ratio for the anemometers, i.e., the difference in the standard deviation of the speed fluctuation measured by these four anemometers under identical flow conditions, are small.

[FIGURE 2 OMITTED]

Simultaneous measurements with the four anemometers, A, B, C and D, were performed in the occupied zone of full-scale rooms under different air distribution conditions comprising mixing and displacement ventilation. The four velocity sensors were placed as close as possible (approximately 0.03 m) to each other but without disturbing each other. Measurements performed with one of the anemometers revealed that the velocity field was rather uniform at the location where the four anemometers were placed. The mean of the standard deviations measured by the four anemometers was calculated. The difference between the standard deviation measured by each of the anemometers and the mean standard deviation (of the four LVTAs) was then calculated. The results shown in Figure 3 reveal that the difference in the standard deviation of speed measured by the anemometers was rather small.

[FIGURE 3 OMITTED]

Error Sources Affecting Velocity Measurement

All important error sources related to measurement with LVTAs were identified as follows:

1. Natural convection

2. Directional sensitivity

3. Frequency response

4. Calibration reference

5. Conversion and reproducibility

6. Velocity gradient

7. Temperature compensation

8. Temperature gradient

9. Air temperature fluctuations

10. Barometric pressure variations

11. Humidity variations

The expanded uncertainty in the measured speed introduced by the error sources was quantified. The expanded uncertainty (95% confidence level) was defined in accordance with the Guide for the Expression of Uncertainty in Measurement (EN 1995b) accepted by the International Organization for Standardization (ISO). The impact of natural convection, directional sensitivity, and dynamic response of the anemometer was modeled. The models developed, together with a database of instantaneous-velocity records obtained in full-scale rooms by means of a three-dimensional laser Doppler anemometer, were used to estimate the uncertainty of mean velocity and standard deviation of velocity measurements by the LVTA due to the separate and combined impacts of natural convection and directional sensitivity of omnidirectional velocity sensors, as well as the dynamic response of the anemometers. The expanded uncertainty due to the other individual error sources was defined on the basis of existing knowledge of the room airflow in general and the characteristics and the behavior of the tested LVTAs. Then, the combined effect of all error sources on uncertainty of speed measurement was identified. This part of the research is described in detail by Popiolek et al. (2007) and Melikov (2000).

Figures 4 and 5 show the absolute expanded uncertainty for the mean speed and the standard deviation of speed for anemometer A. Similar analyses were performed for anemometers B, C, and D (not included in this paper). The estimated absolute expanded uncertainty due to the combined impact of natural convection, directional sensitivity, and frequency response--i.e., [[^.U].sub.1-3]([V.sub.mean]) and [[^.U].sub.1-3] ([V.sub.RMS])--and the absolute expanded uncertainty due to the remaining error sources--i.e.,[[^.U].sub.4]([V.sub.mean]), [^.U.sub.5]([V.sub.mean]), etc. and [[^.U].sub.4], ([V.sub.RMS]), [[^.U].sub.1-3][(V.sub.RMS)], etc.--is shown in the figures. The absolute expanded uncertainty due to the combined effect of all error sources--i.e.,[[^.U].sub.1-11]([V.sub.mean]) and [[^.U].sub.1-11]([V.sub.RMS])--is shown in Figures 4 and 5 as well. Figures 6 and 7 compare the absolute expanded uncertainty of mean speed and the absolute expanded uncertainty of standard deviation of speed measurement, respectively, for the four anemometers due to the different error sources. The estimation is for a flow with a mean speed of 0.15 m/s and a turbulence intensity of 40%. The analyses of the results identified that the uncertainty due to the impact of the studied error sources is different for the four anemometers since they have different characteristics. The error due to calibration reference remains constant and dominant with regard to the mean speed for all four LVTAs, while the uncertainty due to the remaining error sources increase with the increase of the mean speed (except for the uncertainty due to conversion and reproducibility). The results also show that the uncertainty due to barometric pressure changes may be significant, especially at higher speed (0.4-0.5 m/s). Therefore, the requirements specified and suggested in Table 1 for the barometric pressure should be fulfilled. In regard to the standard deviation of speed, the uncertainty due to directional sensitivity, dynamic response, calibration reference, and air temperature fluctuations is substantial. The results reveal that the uncertainty due to velocity gradient, temperature compensation, and humidity is rather small with regard to the mean speed and the standard deviation of speed.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

Recommendations for Characteristics of LVTAs

Based on the results of this study, as well as of previous studies (Melikov et al. 1998b), new updated requirements for the characteristics of LVTAs were developed and suggested for inclusion in future standards. They will ensure the highest accuracy realistically attainable in practice for measurements with LVTAs. The requirements are listed in Table 1. The requirements specified in the table are strict but possible to achieve in practice.

Table 1. Requirements for the Characteristics of Instruments for Low-Velocity Measurements in the Occupied Zone of Rooms Parameter Requirements Comments Requirements for the Instruments Measuring range 0.05 to 1 m/s Design of Distance between Any shield protecting the transducer velocity sensor sensor from damages shall and sensor for not influence accuracy of temperature measurements. compensation: <50 mm Maximum size of The velocity sensor and the velocity sensor: sensor for temperature <10 mm compensation shall not influence each other. Measuring Required: 3 min. Average value of three or period more measurements is recommended. Sampling rate Two or more samples per second Requirements for the Expanded Uncertainty of Measurement Calibration Required: <0.02 Calibration of the velocity reference m/s, Desirable: transducer with the <0.01 m/s anemometer with which it will be used is recommended. Conversion and Required: <0.01 reproducibility m/s, Desirable: <0.005 m/s Frequency Required: upper The upper frequency can be response (10% frequency of 1 determined by the test damping) Hz, Desirable: method suggested by Melikov upper frequency et al. (1998a). of 2 Hz Directional Required: mean Velocity sensor shall be sensitivity speed: <5% omnidirectional. The turbulence placement of the sensor intensity: <10%. with regard to the mean flow direction shall be as close as possible to its placement in the flowduring calibration. The mean speed directional sensitivity and the turbulence intensity directional sensitivity of low velocity probes can be determined by the test method suggested by Stannov et al. (1998). Natural Required: <0.01 convection m/s at speed >0.1 m/s Temperature Required: Applied for air temperature sensitivity <0.4%/K; in the range of Desirable: 15[degrees]C-35[degrees]C. <0.2%/K Barometric Systematic error pressure caused by changes in barometric pressure has to be corrected.

The requirements in Table 1 constitute a considerable improvement over the requirements in the present standards (EN 1998; ASHRAE 2004; ASHRAE 2005). For example, measurement of standard deviation of speed is important for validation of CFD predictions. Data measured by an LVTA with a fast enough response must be used. The response time of 0.2 s specified in the standards for anemometers for indoor velocity measurements is sufficient for accurate measurement of standard deviation. ISO Standard 7726 (EN 1998) specifies a response time of 0.5 s (desirable 0.2 s), while ASHRAE Standard 113 (ASHRAE 2005) requires a response time of 0.2 s). The response time, [[tau].sub.R], is defined as 2.3 times the time constant, [[tau].sub.C], of the instrument, and the time constant is defined as the time for a measuring sensor to reach 63% of the final value after a step change (ASHRAE 2004). The first problem is that the relationship [[tau].sub.R] = 2.3 [[tau].sub.C] is not valid for some of the LVTA anemometers used today, since they do not behave as a first-order inertial system. Furthermore, the response time determined by a step-change test, as suggested in the present standards, can be different for step-up change and step-down change (Melikov and Popiolek 2004). The upper frequency recommended for use in the present study (Table 1) can be defined for any LVTA. The requirement for an upper frequency of 1 Hz specified in Table 1 will ensure measurement of the standard deviation of speed fluctuations with sufficient accuracy (defined in the following). This requirement is realistically achievable since all four improved low-velocity thermal anemometers tested in this study had an upper frequency of 1 Hz.

ACCURACY LIMITATION FOR SPEED MEASUREMENT AND DRAFT DETERMINATION

The absolute expanded uncertainty (95% confidence interval) of measuring mean speed and standard deviation of speed (RMS) determined for anemometers A, B, C and D, with characteristics that comply with the requirements in Table 1, are shown in Figures 8 and 9. The uncertainty of anemometer B with regard to mean speed and standard deviation of speed is only slightly higher than the uncertainty of the three other anemometers. The analyses revealed that this was mainly due to the much higher error of conversion and reproducibility because, for this anemometer, the velocity transducer was calibrated separately from the instrument it was used with during the measurements. The analyses revealed that this error for anemometer B will be significantly reduced when the velocity sensor is calibrated together with the instrument with which it will be used, as recommended in Table 1. The results in Figure 8 allow for determination of the minimum expanded uncertainty for mean speed measurement with LVTA that is realistically achievable in practice when the characteristics of the instrument comply with the requirements in Table 1. Similarly, the minimum absolute expanded uncertainty achievable in practice for measuring standard deviation of speed can be defined (Figure 9).

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

[^.U]([V.sub.mean]) = 0.025 + 0.025 . [V.sub.mean] (3)

[^.U]([V.sub.RMS]) = 0.007 + 0.15 . [V.sub.RMS] (4)

Equations 3 and 4 define uncertainties in mean speed and standard deviation of speed measurement due to the impact of all significant error sources. The defined uncertainties are slightly higher than the uncertainty achieved with the tested anemometers, which, as already discussed, comply with the strict but achievable in practice requirements for the characteristics of LVTAs specified in Table 1. The requirements in the table relate to the most significant error sources. Therefore, the uncertainty of speed measurement with any anemometer with characteristics that comply with the requirements in Table 1 will be lower than the uncertainty specified by Equations 3 and 4. The equations can be used to define the minimum expanded uncertainty of speed measurement that can be achieved in practice by any LVTA, complying at the same time with the requirements specified in Table 1. The defined minimum expanded uncertainty of the mean speed measured by LVTAs includes the impact of all error sources. It is almost twice as low as the uncertainty of [+ or -]0.5 m/s required in the present standards (EN 1998; ASHRAE 2005). It is important to note that the uncertainty specified in the present standards ([+ or -]0.5 m/s) addresses only the error due to static calibration. The total uncertainty due to the combined impact of all possible error factors is not considered in the standards. The minimum uncertainty of measuring standard deviation of speed as defined by Equation 4 is new. The present indoor climate standards do not specify uncertainty in measuring standard deviation of speed. The equations for calculating the uncertainty of mean speed and standard deviation of speed for each of the four tested anemometers are given in Appendix A.

Table A1. Equations for Calculating Expanded Uncertainty of Measurement of Mean Speed and Standard Deviation of Speed Absolute Expanded Uncertainty of Mean Speed, U([V.sub.mean]) = a.[V.sub.mean.sup.2] + b. [V.sub.mean] + c Anemometer Coefficient A B C D a -0.0041 0.0119 0.0008 -0.0053 b 0.0256 0.0299 0.0253 0.0262 c 0.0214 0.0252 0.0228 0.0214 Absolute Expanded Uncertainty of RMS of Speed, U([V.sub.RMS]) = a.[V.sub.RMS.sup.2] + b. [V.sub.RMS] + c Anemometer Coefficient A B C D a -0.1191 -0.12 -0.1252 -0.1217 b 0.1436 0.18 0.1602 0.1504 c 0.0051 0.006 0.0054 0.0053

The absolute uncertainty of turbulence intensity, Tu = [V.sub.RMS]/[V.sub.mean], can easily be defined as follows:

[^.U](Tu) = [[square root of ([([[partial derivative]Tu]/[[partial derivative][V.sub.RMS]]).sup.2][[[^.U]([V.sub.RMS])].sup.2] + [([[partial derivative]Tu]/[[partial derivative][V.sub.mean]]).sup.2]]][[[^.U]([V.sub.mean])].sup.2]) (5)

Equation 5, together with Equations 3 and 4, was used to determine the absolute expanded uncertainty in determination of turbulence intensity by the tested four anemometers. It is shown in Figure 10. The results identify that the turbulence intensity determined (normally the term measured is used) is rather high since it includes the uncertainty of the mean speed and the standard deviation of speed. This is important information, especially when measurements with LVTAs are used for validation of CFD predictions. Therefore, validation with regard to the mean speed and standard deviation of speed is recommended using the method suggested by Popiolek and Melikov (2005). The relatively high uncertainty of measuring turbulence intensity should be considered during airflow distribution measurements under laboratory conditions and in field surveys.

[FIGURE 10 OMITTED]

The draft rating is a function of the air temperature, mean speed, and standard deviation of speed (Equation 2). Equation 2 can be used and, in accordance with the error calculation theory, the absolute uncertainty in determination of draft rating can be determined by the following equation:

[^.U](DR) = [[square root of ([([[partial derivative]DR]/[[partial derivative][t.sub.a]]).sup.2][[[^.U]([t.sub.a])]].sup.2] + [([[partial derivative]DR]/[[partial derivative][V.sub.mean]]).sup.2][[[^.U]([V.sub.mean])].sup.2] + [([[partial derivative]DR]/[[partial derivative][V.sub.RMS]]).sup.2][[[^.U]([V.sub.RMS])].sup.2]) (6)

Equation 2, together with Equations 3 and 4, was used to determine the absolute expanded uncertainty of draft rating when measurements with anemometers complying with the requirements in Table 1 were performed. The analyses were performed for an air temperature range from 20[degrees]C to 26[degrees]C, mean speed from 0.1 to 0.5 m/s, and turbulence intensity from 10% to 60%. An uncertainty of the air temperature measurement of [+ or -]0.2[degrees]C, as recommended in ISO Standard 7726 (EN 1998) was assumed (ASHRAE Standard 113 [ASHRAE 2005] is even more strict and specifies an uncertainty of [+ or -]0.1[degrees]C). Air temperature has a significant impact on, but the uncertainty of its measurement has a rather small impact on the expanded uncertainty of determination. Figure 11 shows the expanded uncertainty of the draft rating determination. The turbulence intensity is used as a parameter. The maximum allowed in ASHRAE Standard 55 (ASHRAE 2004) is 20% and it is 15% in ISO Standard 7730 (EN 1995a). up to 30% is considered in ISO/CD Standard 7730rev, Ergonomics of the Thermal Environment--Analytical Determination and Interpretation of Thermal Comfort by Using Calculations of PMV and PPD Indices and Local Thermal Comfort Indices, which is currently under revision (EN 2003). Typically, the turbulence intensity in rooms with a reasonably high velocity is below 50%. Therefore, the maximum absolute expanded uncertainty in determination of DR, based on measurements with LVTAs that comply with the requirement specified in Table 1, can be defined as [+ or -]5% (see Figure 11). The uncertainty of determination using the tested anemometers will be even better.

[FIGURE 11 OMITTED]

The identified uncertainty in the determination of makes it possible for the first time to perform a realistic evaluation of draft discomfort. As already discussed, CEN Report 1752 (CEN 1998) defines three categories of indoor thermal environment, A, B and C, with draft ratings of 15%, 20%, and 25%, respectively. The absolute uncertainty identified above in draft rating determination ([+ or -]5%) suggests that the ranges for the classes, as specified in the document, will overlap each other; therefore, they are not realistic. The draft international standard under review, ISO/CD 7730rev (EN 2003), also specifies three categories, A, B and C, of indoor environmental quality. It takes advantage of the uncertainty identified in the present research and specifies realistic requirements for three categories of indoor environment with regard to draft discomfort: category A with <10%, category B with < 20%, and category C with < 30%. The identified uncertainty of [+ or -]5% allows for reliable draft assessment. For example, in order to comply with the requirement for < 20% (ASHRAE 2004), the determined should be less than 15% in order to incorporate the 5% uncertainty. The requirement of < 10% in rooms with the highest category A will require a rather low speed in the occupied zone. Studies show that some people prefer higher velocities. This, however, will not be a serious problem in practice, as it is easy to provide local high velocity, for example, by table fan. However, in order to obtain the highest category of indoor environment with regard to occupants' thermal comfort and inhaled-air quality, as well as to decrease the risk of airborne transmission of infectious agents, personalized ventilation that provides clean air directly to the breathing zone and allows for control of airflow speed and direction is recommended (Melikov 2004).

CONCLUSIONS

The separate and the combined impact of all significant error sources on low-velocity measurements and draft assessment in practice was analyzed. The static calibration of LVTAs is the most important factor for accurate measurement of mean speed indoors, while, with regard to the standard deviation of speed, the errors due to directional sensitivity, dynamic response, calibration reference, and air temperature fluctuations are substantial. The use of LVTAs with an upper frequency minimum of 1 Hz is recommended in order to measure the standard deviation of speed with sufficient accuracy.

Requirements for LVTAs that will decrease the uncertainty of low-velocity measurements and improve the accuracy of indoor environment assessment are suggested for inclusion in future standards.

The minimum absolute uncertainty in mean speed measurements with LVTAs is defined for anemometers that comply with the suggested requirements. This uncertainty is almost twice as low as the uncertainty specified in the present standards. For the first time, uncertainty of measuring standard deviation of speed is defined.

The uncertainty of draft rating determination ([+ or -]5%) that can be realistically achieved in practice is identified. This will make it possible to determine reliable recommendations in future indoor climate standards.

The results of this study will allow for realistic field assessment of the indoor environment and reliable validation of CFD predictions by measurements with LVTAs.

ACKNOWLEDGMENTS

This research was funded by the European Community under the IMT/SMT program, contract number SMT4-CT97-2172.

NOMENCLATURE

Symbols

DR = draft rating

[V.sub.mean] = mean speed, m/s (ft/min)

[V.sub.RMS] = standard deviation of speed fluctuations, m/s (ft/min)

Tu = [V.sub.RMS]/[V.sub.mean] = turbulence intensity, dimensionless (-)

[t.sub.a] = air temperature, [degrees]C ([degrees]F)

[[^.U].sub.j=1-11][(V.sub.mean)] = absolute expanded uncertainty in mean speed due to different error sources as defined in the paper, m/s (ft/min)

[[^.U].sub.j=1-11]([V.sub.RMS]) = absolute expanded uncertainty in standard deviation (RMS) of speed due to different error sources as defined in the paper, m/s (ft/min)

REFERENCES

ASHRAE. 2004. ANSI/ASHRAE Standard 55-2004, Thermal Environmental Conditions for Human Occupancy. Atlanta: American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.

ASHRAE. 2005. ANSI/ASHRAE Standard 113-2005, Method of Testing for Room Air Diffusion. Atlanta: American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.

CEN. 1998. Ventilation for buildings: Design criteria for the indoor environment. Report 1752, European Committee for Standardization TC 156/WG 6.

EN. 1995a. ISO Standard 7730, Moderate Thermal Environments--Determination of the PMV and PPD Indices and Specification of the Conditions for Thermal Comfort. International Standard Organization for Standardization, Geneva, Switzerland.

EN. 1995b. Guide for the Expression of Uncertainty in Measurement. International Standard Organization for Standardization, Geneva, Switzerland.

EN. 1998. ISO/DIS Standard 7726, Ergonomics of the Thermal Environment--Instruments for Measuring Physical Quantities. International Standard Organization for Standardization, Geneva, Switzerland.

EN. 2003. ISO/CD Standard 7730rev, Ergonomics of the Thermal Environment--Analytical Determination and Interpretation of Thermal Comfort by Using Calculations of PMV and PPD Indices and Local Thermal Comfort Indices. International Standard Organization for Standardization, Geneva, Switzerland.

Fanger, P.O., and C.J.K. Pedersen. 1977. Discomfort due to air velocities in spaces. Proceedings of the Meeting of Commission B1, B2, E1 of the International Institute of Refrigeration, Belgrade, Yugoslavia, pp. 289-96.

Fanger, P.O., A.K. Melikov, H. Hanzawa, and J. Ring. 1988. Air turbulence and sensation of draught. Energy and Buildings 12(1):21-39.

Finkelstein, W., A.K. Melikov, T. Sefker, and G. Langkilde. 1996. Laser Doppler measurements of airflow characteristics in rooms with mechanical ventilation. Proceedings of Indoor Air '96, Nagoya, Japan, 1:785-90.

Joergensen, F.E., Z. Popiolek, A.K. Melikov, and M.G. Silva. 2004. Total uncertainty of low velocity thermal anemometers for measurement of indoor air movements. Proceedings of the 9th International Conference on Air Distribution in Rooms (ROOMVENT), Coimbra, Portugal.

Loomans, M.G.L.C., and A.W.M. Schijndel. 2002. Simulation and measurement of the stationary and transient characteristics of the hot sphere anemometer. Building and Environment 37:153-63.

Mayer, E. 1992. New measurements of the convective heat transfer coefficient: Influences of turbulence, mean air velocity and geometry of the human body. Proceedings of ROOMVENT '92, Aalborg, Denmark 3:263-76.

Mayer, E., and R. Schwab. 1988. Direction of low turbulent airflow and thermal comfort. Proceedings of Healthy Buildings '88, Stockholm, Sweden 2:577-81.

Melikov, A.K. 2000. Accuracy of draught measurement. Final Report, EC Research Project, Contract No. SMT4-CT97-2172, Technical University of Denmark, Department of Energy Engineering.

Melikov, A.K. 2004. Personalized ventilation. Indoor Air 14(supplement 7):157-67.

Melikov, A.K., and T. Sawachi. 1992. Low velocity measurements: Comparative study of different anemometers. Proceedings of International Conference ROOMVENT '92, Aalborg, Denmark 3:291-306.

Melikov, A.K., Z. Popiolek, and F.E. Joergensen. 1998a. New method for testing dynamic characteristics of low velocity thermal anemometers. ASHRAE Transactions 104(1b):1491-506.

Melikov, A.K., Z. Popiolek, W. Finkelstein, T. Sefker, and I. Care. 1998b. Requirements and guidelines for low velocity measurements. ASHRAE Transactions 104(1b):1529-539.

Melikov, A.K., and Z. Popiolek. 2004. Comparison of different methods for the determination of dynamic characteristics of low velocity anemometers. Measurements Science and Technology 15:1709-715.

Popiolek, Z., A.K. Melikov, F.E. Joergensen, W. Finkelstein, and T. Sefker. 1998. Impact of natural convection on accuracy of low velocity measurements by thermal anemometers with omnidirectional sensor. ASHRAE Transactions 104(1b):1507-518.

Popiolek, Z., and A.K. Melikov. 2005. A bench mark database and validation method for CFD predictions of airflow in a room. Proceedings of Indoor Air 2005, Beijing, China, 3:2635-639.

Popiolek, Z., F.E. Joergensen, A.K. Melikov, M.C.G. Silva, and W. Kierat. 2007. Assessment of uncertainty in measurement with low velocity thermal anemometers. Journal of Ventilation 6(2):113-26.

Stannov, T., A.K. Melikov, Z. Popiolek, and F.E. Joergensen. 1998. Test method for describing directional sensitivity of anemometers for low velocity measurements indoors. ASHRAE Transactions 104(1b):1481-489.

Toftum, J., G. Zhou, and A. Melikov. 1997. Effect of airflow direction on human perception of draught. Proceedings of CLIMA 2000, Brussels, Belgium, August 30-September 2, paper 366.

Zhou, G., and A. Melikov. 2002. Equivalent frequency--A new parameter for description of frequency characteristics of airflow fluctuations. Proceedings of Roomvent 2002, Copenhagen, Denmark, pp. 3357-360.

Zhou, G., A.K. Melikov, and P.O. Fanger. 2002. Impact of equivalent frequency on the sensation of draught. Proceedings of Roomvent 2002, Copenhagen, Denmark, pp. 297-300.

Arsen K.Melikov, PhD Fellow ASHRAE

Z.Popiolek, PhD, DSc

M.C.G.Silva, PhD

I.Care, PhD

T.Sefker, PhD Member ASHRAE

Arsen K. Melikov is an associate professor at the International Center for Indoor Environment and Energy, Department of Mechanical Engineering, Technical University of Denmark, Lyngby. Z. Popiolek is a professor and head of the Department of Heating, Ventilation and Dust Removal Technology, Silesian University of Technology, Gliwice, Poland. M.C.G. Silva is an associate professor at ADAI, Department of Mechanical Engineering, University of Coimbra-Polo II, Coimbra, Portugal. I. Care is a project manager in the Metrology Division of CETIAT, Domaine Scientifique de la Doua, Villeurbanne Cedex, France. T. Sefker is manager of research and development, TROX GmbH, Neukirchen-Vluyn, Germany.

Printer friendly Cite/link Email Feedback | |

Author: | Melikov, Arsen K.; Z. Popiolek; M.C.G. Silva; I. Care; T. Sefker |
---|---|

Publication: | HVAC & R Research |

Geographic Code: | 1USA |

Date: | Nov 1, 2007 |

Words: | 5625 |

Previous Article: | Whole-field measuring methods in ventilated rooms. |

Next Article: | Analysis and design of room air distribution systems. |

Topics: |