1、2010 ASHRAE 271ABSTRACTThis is the third paper in a four-part series reporting onthe test and evaluation of typical carbon-dioxide sensorsused in building HVAC applications. Fifteen models of NDIRHVAC-grade CO2sensors were tested and evaluated todetermine the humidity, temperature, and pressure sens
2、itivityof the sensors. This paper reports the performance of thesensors at various relative humidity, temperature, and pres-sure levels common to building HVAC applications andprovides a comparison with manufacturer specifications.Among the 15 models tested, eight models have a single-lamp, single-w
3、avelength configuration, four models have adual-lamp, single-wavelength configuration, and three modelshave a single-lamp, dual-wavelength configuration.The sensors were tested in a chamber specifically fabri-cated for this research. A description of the apparatus and themethod of test are described
4、 in Part 1 (Shrestha and Maxwell2009). The test result showed a wide variation in humidity,temperature, and pressure sensitivity of CO2sensors amongmanufacturers. In some cases, significant variations in sensorperformance exist between sensors of the same model. Eventhe natural variation in relative
5、 humidity could significantlyvary readings of some CO2sensor readings. The effects of tem-perature and pressure variation on NDIR CO2sensors are un-avoidable without an algorithm to compensate for the changes.For the range of temperature and pressure variation in an air-conditioned space, the effect
6、 of pressure variation is more sig-nificant compared to the effect of temperature variation.INTRODUCTIONThis is the third part of a four-part series of papers report-ing on the test and evaluation of typical wall-mounted CO2sensors used in building HVAC systems. In this study, fifteenmodels of NDIR
7、(non-dispersive infrared) HVAC-grade wall-mounted CO2sensors were tested and evaluated. In all, 45 sen-sors (three from each model) were tested in order to determinethe effects of humidity, temperature, and pressure on the CO2sensors readings. The sensitivity of the sensor reading to eachof these th
8、ree parameters was computed and compared to themanufacturers specifications. The experimental procedureused to test and evaluate the sensors is described in Part 1(Shrestha and Maxwell 2009) of this paper. Among the fifteenmodels tested, eight models have a single-lamp, single-wave-length configurat
9、ion, four models have a dual-lamp, single-wavelength configuration, and three models have a single-lamp, dual-wavelength configuration. All single-lamp, single-wavelength sensors and one single-lamp, dual-wavelengthsensor incorporate an “automatic baseline adjustment” algo-rithm in the sensors elect
10、ronics package. The working prin-ciples of NDIR CO2sensors are described in Part 1 (Shresthaand Maxwell 2009) of this paper.PREVIOUS STUDIESIn the past, limited studies have been done to investigatethe effects of humidity, temperature, and pressure variationson HVAC-grade CO2sensors. Fahlen et al. (
11、1992) evaluatedthe performance of two CO2sensors, one photo-acoustic typeand one infrared spectroscopy type, in lab tests and long termfield tests. The lab tests included performance and environ-mental tests. The authors found that the CO2sensors testedwere sensitive to humidity below a threshold va
12、lue, howeverthe threshold humidity level was not determined and the effectof humidity on the sensors readings was not evaluated.During the low-temperature test at 5C (which was within theAn Experimental Evaluation of HVAC-Grade Carbon-Dioxide Sensors Part 3: Humidity, Temperature, and Pressure Sensi
13、tivity Test ResultsSom S. Shrestha, PhD Gregory M. Maxwell, PhDStudent Member ASHRAE Member ASHRAESom S. Shrestha is a R however, the absolutehumidity remained constant.Established procedures, including guidelines for steady-state conditions, described in Part 1 (Shrestha and Maxwell2009) of this pa
14、per were used to perform the testing.HUMIDITY SENSITIVITY TEST RESULTSThe effect of humidity on a CO2sensors output is deter-mined by comparing the sensor readings at 20% and 60% RHto the sensor reading at 40% RH. The humidity sensitivity testresults are presented graphically in terms of the deviati
15、on ofthe sensor readings at 20% and 60% from the readings at40% RH (i.e., deviation from reading at 40% RH = measuredCO2concentration at a particular RH measured CO2 concen-tration at 40% RH). Figures 1, 2, and 3 illustrate the humiditysensitivity test results for the single-lamp single-wavelength,d
16、ual-lamp single-wavelength, and single-lamp dual-wave-length configurations, respectively. 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction,
17、distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission. ASHRAE Transactions 273The humidity sensitivity of each CO2sensor was calcu-lated using a linear regression of the test results. The slope ofthe regression line represents the sen
18、sitivity of the sensor interms of deviation in ppm reading per % change in relativehumidity. The numerical results of the sensor readings at eachof the test conditions along with the calculated sensitivity arealso tabulated. Tables 2, 3, and 4 provide the values forthe three sensor configurations. P
19、ositive values of humiditysensitivity indicate an increase in sensor reading as RH in-creases, and vice versa.The majority of the sensors show little to no sensitivity tohumidity. However, three sensor models were particularlysensitive to humidity. Specifically, all of the sensors frommodels S4 and
20、S7 exhibit positive relative humidity sensi-tivity. Sensitivity values for these sensor range from 1.4 to2.9 ppm/% RH. One sensor from model S14 (sensor A) has asensitivity of 3 ppm/% RH.Table 1. Manufacturer-Specified Temperature Sensitivity, Pressure Sensitivity, and Operating Range for Their Sens
21、orsSensor ModelConfiguration Temperature Sensitivity Pressure SensitivityTemperature and humidity Operating RangeS1 Single-lamp, single-wavelength 5 ppm/1.8F (1C) NA32 to 122F (0 to 50C)0 to 95% RHS2 Single-lamp, single-wavelength NAAdd 6.7% of reading per psi (6.89 kPa) decrease from 14.70 psia (10
22、1.35 kPa)*60 to 90F (15 to 32C)0 to 95% RHS3 Single-lamp, single-wavelength 3 ppm/1.8F (C)*6.7% of reading per psi (6.89 kPa) from 14.70 psia (101.35 kPa)*59 to 90F (15 to 32C)0 to 95% RHS4 Single-lamp, single-wavelength NA NA32 to 122F (0 to 50C)5 to95% RHS5 Single-lamp, single-wavelength 4 ppm/1.8
23、F (C)*6.7% of reading per psi (6.89 kPa)*32 to 122 (0 to 50C)0 to 95% RHS6 Single-lamp, single-wavelength NA9.6% of reading per psi(6.89 kPa) deviation from14.5 psia (100 kPa)*32 to 122F (0 to 50C)0 to 95% RHS7 Single-lamp, single-wavelength NAAdd 7.2% of reading per psi (6.89 kPa) decrease from 14.
24、70 psia (101.35 kPa)*32 to 122F (0 to 50C)0 to 90% RHS8 Single-lamp, single-wavelength NA NA32 to 122F (0 to 50C)0 to 95% RH S9 Dual-lamp, single-wavelength 5 ppm/1.8F (C)9.8% of reading per psi(6.89 kPa)*32 to 122F (0 to 50C)5 to95% RHS10 Dual-lamp, single-wavelength 2 ppm/1.8F (C ) NA23 to 113F (5
25、 to 45C)0 to 85% RHS11 Dual-lamp, single-wavelength NA NA32 to 100F (0 to 40C)0 to 95% RHS12 Dual-lamp, single-wavelength 5 ppm/1.8F (C) NA23 to 131F (5 to 55C)0 to 90% RHS13 Single-lamp, dual-wavelength 2 ppm/1.8F (C)*Add 10.2% of reading per psi (6.89 kPa) decrease from 14.18 psia (97.77 kPa)*23 t
26、o 113F (5 to 45C)0 to 85% RHS14 Single-lamp, dual-wavelength NAAdd 6.7% of reading per psi (6.89 kPa) decrease from 14.70 psia (101.35 kPa)*32 to 122F (0 to 50C)0 to 95% RHS15 Single-lamp, dual-wavelength3 ppm/1.8F (C)*reference 77F (25C)NA23 to 113F (5 to 45C)0 to 85% RHNotes: NA indicates that the
27、 information was not available in the manufacturers product literature.* indicates that the value was calculated from the manufacturers product literature. 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, V
28、ol. 116, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission. 274 ASHRAE TransactionsTEMPERATURE SENSITIVITY TEST RESULTSThe effect of temperature on a CO2sensors output isdete
29、rmined by comparing the sensor readings at 66F (18.9C)and 80F (26.7C) to the sensor reading at 73F (22.8C). Thetemperature sensitivity test results are presented graphically interms of the deviation of the sensor readings at 66F (18.9C)and 80F (26.7C) from the reading at 73F (22.8C) i.e.,deviation f
30、rom readings at 73F (22.8C) = measured CO2concentration at a particular temperature measured CO2concentration at 73F (22.8C). Figures 4, 5, and 6 illustratethe temperature sensitivity test results for the single-lampsingle-wavelength, dual-lamp single-wavelength, and single-lamp dual-wavelength conf
31、igurations, respectively.The temperature sensitivity of each CO2sensor wascalculated using a linear regression of the test results. Theslope of the regression line represents the sensitivity of thesensor in terms of deviation in ppm reading per degree changein temperature. The numerical results of t
32、he sensor readings ateach of the test conditions along with the calculated sensitivityare also tabulated. Tables 5, 6, and 7 provide the values for thethree sensor configurations. Positive values of temperaturesensitivity indicate an increase in sensor reading as tempera-ture increases, and vice ver
33、sa.Temperature sensitivity is not consistent between sensormodels. For many sensors, the temperature sensitivity is neg-ligibly small. Nine sensor models showed temperature sensi-tivity within 5 ppm/1.8F (5 ppm/C). Sensor S12B showed thehighest temperature sensitivity of 10 ppm increase in sensorrea
34、ding per 1.8F (1C) decrease in temperature.PRESSURE SENSITIVITY TEST RESULTSThe effect of pressure on a CO2sensors output is deter-mined by comparing the sensor readings at 13.25 psiaFigure 1 Humidity sensitivity test results of single-lamp, single-wavelength sensors. 2010, American Society of Heati
35、ng, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission. A
36、SHRAE Transactions 275(91.36 kPa) and 11.80 psia (81.36 kPa) to the sensor readingat 14.70 psia (101.35 kPa). The pressure sensitivity test resultsare presented graphically in terms of the deviation of thesensor readings at 13.25 psia (91.36 kPa) and 11.80 psia(81.36 kPa) from the reading at 14.70 p
37、sia (101.35 kPa) (i.e.,deviation from reading at 14.70 psia (101.35 kPa) = measuredCO2concentration at a particular pressure measured CO2concentration at 14.70 psia (101.35 kPa). Figures 7, 8, and 9illustrate the pressure sensitivity test results for the single-lamp single-wavelength, dual-lamp sing
38、le-wavelength, andsingle-lamp dual-wavelength configurations, respectively.The pressure sensitivity of each CO2sensor was calcu-lated using a linear regression of the test results. The slope ofthe regression line represents the sensitivity of the sensor interms of deviation in percent reading per un
39、it change in pres-sure. The numerical results of the sensor readings at each ofthe test conditions along with the calculated sensitivity arealso tabulated. Tables 8, 9, and 10 provide the values for thethree sensor configurations. Positive values of pressure sensi-tivity indicate an increase in sens
40、or reading as pressureincreases, and vice versa. All sensors showed similar responseto pressure change. The maximum and minimum pressuresensitivity were observed as 10.7% and 7.6% reading/psi(6.89 kPa), respectively.Given the sensitivity of a NDIR CO2sensor reading topressure, it is of interest to e
41、stimate the expected change in asensors reading due to the natural variation in barometricpressure for a given location. Using TMY2 weather data, themaximum change in barometric pressure was determined forFigure 2 Humidity sensitivity test results of dual-lamp, single-wavelength sensors.Figure 3 Hum
42、idity sensitivity test results of single-lamp,dual-wavelength sensors. 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction, distribution, or tra
43、nsmission in either print or digital form is not permitted without ASHRAEs prior written permission. 276 ASHRAE Transactionsnine U.S. cities. From the CO2sensor sensitivity test results,the average pressure sensitivity is 8.9% reading/psi (6.89 kPa).Applying this sensitivity to the barometric pressu
44、re variationfor each of the nine cities, the expected variation in CO2sensorreading were calculated. Table 11 summarizes the results. Forthe nine cites considered, Boston, Chicago, and New Yorkhave the largest variation in barometric pressure. For theselocations, the expected variation in a CO2senso
45、r reading is84 ppm for actual CO2concentration at 1100 ppm. The signif-icance of these results is important when considering sensorcalibration. Even for a “perfectly” calibrated sensor, the read-ing could be in error by several ppm depending on the baro-metric pressure at the time the sensor was cal
46、ibrated comparedto the barometric pressure at other times of the year.CONCLUSIONThe results from the tests conducted under accurate andrepeatable conditions show a wide variation in humidity andtemperature sensitivity among the NDIR CO2sensor models.In some cases, significant variations in sensor pe
47、rformanceexist between sensors of the same model while in other cases,all sensors of the same model showed almost identical behav-ior.None of the sensor manufacturers specified humidity de-pendence of their CO2sensors. While majority of the sensorsshow little to no sensitivity to humidity, the test
48、results re-vealed that three sensor models are highly sensitive to humid-ity. The maximum humidity sensitivity was observed as 3ppm/% RH. It is suspected that the sensors with high humiditysensitivity use hygroscopic material as an optical filter.Theoretically, increase in temperature at a fixed gas
49、composition and pressure should decrease the number ofmolecules in optical path of a NDIR CO2sensor and hencedecrease the sensor reading. Some sensors showed an oppositephenomenon. Nine sensor models showed temperatureTable 2. Humidity Sensitivity Test Results of Single-Lamp, Single-Wavelength SensorsSensor Model SensorReading atSensitivity, ppm/% RH20% RH, ppm40% RH, ppm60% RH, ppmS1 A 1192 1201 1212 0.5B 1195 1203 1209 0.3C 1212 1217 1225 0.3S2 A 1222 1229 1236 0.3B 1271 1277 1284 0.3C