**<)-

Techniques for Measuring and Controlling Outside Air Intake Rates in Variable Air Volume Systems Final Report, JCEM TR/99/3 Moncef Krarti, Ph.D., P.E. Michael J . Brandemuehl, Ph.D., P.E. Chris Schroeder Erik Jeannette

COPYRIGHT American Society o? Heating, Refrigerating and Air-Conditionirig Engineers, Inc. 1791 Tullie Circle, NE, Atlanta, GA 30329

Executive Summary Control of minimum outside air intake rates is critical to meet standards set by the "Ventilation Rate Procedure" as outlined in ASHRAE Standard 621989 for maintaining adequate indoor air quality within conditioned spaces. VAV systems present additional complications compared to CAV systems. Pressure in the mixed air plenum can fluctuate with changing supply air volumes, making commonly used control strategies for CAV systems inadequate for use in VAV systems. Analysis of theoretical and experimental results indicate that the use of both a fixed minimum outside air damper position and volumetric fan tracking control are inadequate at maintaining minimum outside air intake rates in VAV systems. The ideal control strategies are those based upon direct measurement of outside airflow rates. However, proper installation of direct airflow measurement devices requires long and unobstructed lengths of duct, which may not be possible to achieve for some buildings. Additionally, in HVAC systems with ductwork sized for use with an economizer cycle, measurement of minimum outside airflow rates by an averaging Pitot-tube array is usually inaccurate unless a high-accuracy, auto-zeroing differential pressure transmitter is used due to low airflow velocities. A possible solution is to use electronic thermal anemometry which is more accurate at lower flow rates. Alternatively, a section of duct dedicated to bring in the minimum outside airflow could be added. The size of this duct is reduced to increase the velocity of the minimum outside airflow and make measurement with an averaging Pitot-tube array possible. Control stability problems due to wind and stack effects were not considered in this report. If direct airflow measurement devices cannot be installed, two alternatives to maintain minimum outside air intake rates exist. The first is to use plenum pressure control, where the pressure drop across the outside air damper and louver is held constant, a technique based on the premise that the pressure drop across a fixed orifice can be correlated to the desired minimum air flow rate. Secondly, an indirect measurement of the outside airflow rate can be achieved using a CO2 concentration balance and an accurate measurement of supply air volume. The value of outside air intake rate measured with the CO2 concentration balance can then be used for direct control. However, the C0 2 concentration balance technique is not adequate during times of low occupancy, or for outside air intake rates that represent is a small percentage of the total supply airflow rate.

ii

Contents
1 INTROOUCnrON -................................................ 1
1.1 PROJECT DESCRIPTION 1

1.1.1 1.1.2 Z

Outside Air Intake Rate Measurements VAV Control Techniques

1 2
2

1.2 REPORT ORGANIZAVON

L I 1 ILKA 1 UKJl/ KJiiVllliW *........*........................................................ 3

2.1 AIRFLOW MEASUREMENT TECHNIQUES 3

2.1.1 2.7.2 2.1.3 2.1.4 2.2.1 2.2.2 2.2.3 2.2.4 3

Direct Airilow Measurement Techniques. Indirect Airilow Measurement Techniques Other Airilow Measurement Techniques Accuracy of Airilow Measurement Techniques Fixed Minimum Outdoor Air Damper Position VAV Control Techniques for Economizer Systems. VAV Control Techniques for Systems with a Dedicated Outside Air Duct Other VAV Control Techniques.

3 6 8 10
11

2.2 VAV OUTDOOR AIR CONTROLTECHNIQUES

.-.

11 11 15 17

ERROR ANALYSIS ..-....B*.H.....aM..H.v....a.a.....a..........H.a.,aa.,..........,H.,a....H.,...,...a................ 19

3.1 UNCERTAINTIES..... 19

3.1.1 3.1.2 3.2.1 3.2.2 3.2.3 3.3.1 3.3.2 3.3.3 3.3.4 3.3.5 3.3.6 3.3.7

Data Acquisition Errors Installation and Calibration Errors

... .......-.

20 20
22

3.2'AIRFLOW MEASUREMENT TECHNIQUES ERROR ANALYSES......:.....;.

Direct Airilow Measurement Techniques. Indirect Airilow Measurement Techniques Theoretical Recommendations for Airilow Measurement Techniques Fixed Minimum Outside Air Damper Position Volumetric Fan Tracking. Plenum Pressure Control. Measurement and Control of Outside Airflow Rate with Economizer. Measurement and Control of Dedicated Minimum Outside Duct Airilow Rate Outside Air Duct Injection Fan Theoretical Recommendations for VAV Control Techniques.

23 27 31
32

3.3 VAV CONTROL TECHNIQUES ERROR ANALYSES

33 33 35 36 37 37 37

LABORATORY DESCRIPTION
4.1 LABORATORY AIR SYSTEM.

39
39

4.1.1 4.1.2 4.1.3

System Function System Description Laboratory Modifications.

39 39 40
41 43

4.2 SUPPLYDUCTAVERAGING PITOT-TUBE ARRAYCAUBRAVON. 4.3 LABORATORY TEST CONDITIONS

4.3.1 4.3.2 4.3.3 4.3.4

Laboratory Control. VAV Load Profiles Outside Air Temperature Profile. Summary of Systems to be Tested.

43 44 44 46

ANALYSIS OF LABORATORY TEST RESULTS

5.1 LABORATORY TEST RESULTS

..

48
48

5.1.1 5.1.2 5.2.1 5.2.2 5.2.3

Measurement Technique Analysis VAV System Control Results Fixed Minimum Position Outside Air Damper. : Direct Control with Economizer Duct Direct Control with Dedicated Duct with and without an Injection Fan

49 51
54

5.2 DATA ANALYSIS

54 54 55

iii

5.2.4 5.2.5 5.2.6 6

Volumetric Fan Tracking. Plenum Pressure Control. Direct Control using Concentration Balance Measurement Technique. .

55 :.... 56 56 61
61 62

SUMMARY AND CONCLUSIONS

6.1 MINIMUM OUTSIDE AIR INTAKE CONTROL AND MEASUREMENT RECOMMENDATIONS 6.2 FUTURE WORK

REFERENCES

64 67
67 67

APPENDIX A - JCEM LABORATORY AIR SYSTEM DESCRIPTION

A.1 SYSTEMCOMPONENTS. A.2 COMPONENT DESCRIPTIONS

A.2.1 A.2.2 A.2.3 A.2.4 A.2.5 A.2.6

Main Air Handling Unit Outside Air Conditioning Station Return Fan Fan-Powered Mixing Boxes Dampers Zone Simulators 1 & 2

67 68 71 71 72 73 74

APPENDIX B - ANALYSIS OF THE EFFECTS OF HUMIDITY AND PRESSURE ON THE CALCULATED AIRFLOW RATE

APPENDIX C - CONCENTRATION BALANCE AIRFLOW MEASUREMENT TECHNIQUE TEST RESULTS ............................-.... 77

C.ICO2 SENSOR REPEATABILITY ERRORS C.2MEASUREMENT TECHNIQUE VERIFICATION. 77 78

APPENDIX D - LABORATORY TEST RESULTS

D.1 FIXED DAMPER Pos., 20% OA, CASE 1-A. D.2FIXED DAMPER POS., 30% OA, CASE 1-B. D.3FIXED DAMPER POS., 40% OA, CASE 1-C.. D.4PLENUM PRESSURE CONTROL, 20% OA, CASE2-A D.SPLENUM PRESSURECONTROL, 40% OA, CASE 2-C...,,..... D.6D1RECTCONTROL WITHECON. DUCT, 20% OA AVERAGINGPrrbT-fuBEARRAY, CASE3-/4..... D.7 DIRECT CONTROL WITHECON. DUCT, 40% OA, AVERAGING PITOT-TUBE ARRAY, CASE3-B D.8 DIRECT CONTROL W/ECON. DUCT, 20% OA ELECTRONIC THERMAL ANEMOMETRY, CASE4-A D.9DIRECT CONTROL W/ECON. DUCT, 40% OA, ELECTRONIC THERMAL ANEMOMETRY, CASE4-B D. 10 VOLUME TRACKING, 20% OA, ELECTRONIC THERMAL ANEMOMETRY, CASE 5-A D.11 DIRECT CONTROL W/DEDICATED DUCT, 20% OA, AVERAGING PITOT-TUBE ARRAY, CASE 6-A D.12 DIRECT CONTROL W/DEDICATED DUCT, 30% OA, AVERAGING PITOT-TUBE ARRAY, CASE6-B D.13 DIRECTCNTRL W/DEDICATED DUCT, 20% OA, ELECTRONIC THERMAL ANEMOMETRY, CASE 7-A D.14 DIRECTCNTRL W/DEDICATED DUCT, 30% OA, ELECTRONIC THERMAL ANEMOMETRY, CASE 7-B D.15 INJECTION FAN W/DEDICATED DUCT, 20% OA, AVERAGING PITOT-TUBE ARRAY, CASE8-A D. 16 INJECTION FAN W/DEDICATED DUCT, 30% OA, AVERAGING PITOT-TUBE ARRAY, CASE 8-B. D.17 INJECTION FAN W/DED. DUCT, 20% OA, ELECTRONIC THERMAL ANEMOMETRY, CASE9-A D.18 INJECVONFANW/DED. DUCT, 30% OA, ELECTRONIC THERMAL ANEMOMETRY, CASE9-B

80
81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98

APPENDIX E - PRESSURE AND TEMPERATURE DATA

E.1 DUCT STATIC PRESSURES E.2 TEMPERATURE PROFILES

99
99 99

iv

List of Tables
Table 1: Accuracy and Range of Airflow Measurement Techniques. Table 2: JCEM Laboratory conditions used for error analysis. Table 3: Averaging Pitot-tube array uncertainties for JCEM Laboratory. Table 4: Electronic thermal anemometer uncertainties for JCEM Laboratory. Table 5: Temperature measurement uncertainties for JCEM Laboratory. Table 6: Minimum C02 differences (ppm) between the return and outside airflow for 15% accuracy using the concentration balance method (data acquisition error of 1 ppm, sensor repeatability of 3 ppm) 30 Table 7: Minimum airflow rates for 15% accuracy using direct measurement techniques in the JCEM Laboratory. Table 8: Uncertainty values used in the calculation of predicted errors in Figure 18. Table 9: Predicted errors of direct control techniques in the JCEM Laboratory. Table 10: Key physical and control characteristics of the JCEM Laboratory. Table 11: Systems tested in the JCEM Laboratory. Table 12: Comparison of Outside Airflow Measuremets. Table 13: Absolute difference between averaging Pitot-tube array and electronic thermal anemometer airflow measurements. Table 14: Summary of results for the systems tested. Table 15: Errors calculating air density while neglecting the effects of humidity. Table 16: Errors calculating air density using the annual average atmospheric pressure Table 17: Errors calculating the air density neglecting the effects of humidity and using the annual average atmospheric pressure Table 18: Duct static pressure data for systems tested. Table 19: Supply and outside temperature data for systems tested. 75 99 100 . 51 52 75 75 32 35 37 43 47 50 10 23 25 26 29

List of Figures
Figure 1: Typical Pitot-static tube Figure 2: Outside airflow rate control schematic for fixed position outside air damper. Figure 3: Outside airflow rate control schematic for system using volumetric fan tracking control strategy. Figure 4: Outside airflow rate control schematic for system with economizer damper. Figure 5: Plenum-pressure control schematic. Figure 6: Outside airflow rate control schematic for system with dedicated minimum outside airflow duct. Figure 7: Outside airflow rate control schematic for system with dedicated minimum outside airflow injection fan Figure 8: Outside airflow profile at 707 CFM. Figure 9: Outside airflow profile at 1,130 CFM. Figure 10: Outside airflow profile at 3,276 CFM. Figure 11: Predicted errors of averaging Pitot-tube arrays in the JCEM Laboratory. Figure 12: Predicted errors of electronic thermal anemometry in the JCEM Laboratory. Figure 13: Errors introduced by replacing enthalpy balance with a temperature balance at 20% OA fraction Figure 14: Predicted errors of temperature balance in the JCEM Laboratory. Figure 15: Predicted measurement errors as a function ofOA% for JCEM Laboratory (5% error in SA measurement, 1 ppm acquisition error, and repeatability errors of 3 Figure 16: Predicted errors for the fixed damper position control technique Figure 17: Predicted errors of volumetric tracking control technique in the JCEM Laboratory. Figure 18: Predicted errors of plenum pressure control in the JCEM Laboratory. Figure 19: JCEM Laboratory Isometric. Figure 20: Schematic diagram of new outside air ductwork. Figure 21: Calibration of supply duct Pitot-tube averaging array with 95% confidence intervals. Figure 22: VAVprofiles used for testing. Figure 23: Outside air temperature profile used for testing Figure 24: Comparison of electronic thermal anemometer and averaging Pitot-tube array airflow measurements as a function of air temperature Figure 25: Representative temperature values during laboratory testing. Figure 26: Comparison of various airflow measurement techniques. Figure 27: Outside airflow % ofdeisgn for fixed minimum outside air damper position test Figure 28: Supply and outside airflows for the concentration balance control test. Figure 29: Outside airflow % of design with predicted error bars Figure 30: Comparison of airflow measurement techniques for concentration balance control test Figure 31: Repeatability tests for a commercially available C02 sensor. ; Figure 32: Predicted measurement errors as a function of 0A% for JCEM Laboratory (5% error in SA measurement, 1 ppm acquisition error, repeatability errors of 3 ppm) Figure 33: Measured errors in the calculated outside airflow rate using theconcentration balance method. Figure 34: Outside air velocity versus OA temperature and supply air velocity. Figure 35: Outside airflow % of design with predicted measurement errors. 4 11 12 13 14 15 16 24 24 24 26 27 28 29

33 34 36 40 41 42 44 45 46 48 49 54 57 58 59 77 78 79 81 81

vi

Figure 36: Comparison of direct airflow measurement techniques. Figure 37: Outside air velocity versus OA temperature and supply air velocity. Figure 38: Outside airflow % of design with predicted measurement errors Figure 39: Comparison of direct airflow measurement techniques. Figure 40: Supply and outside airflows with predicted measurement errors Figure 41: Outside airflow % of design with predicted measurement errors Figure 42: Comparison of direct airflow measurement techniques. Figure 43: Supply and outside airflows with predicted measurement errors. Figure 44: Outside airflow % of design with predicted measurement errors Figure 45: Comparison of direct airflow measurement techniques Figure 46: Supply and outside airflows with predicted measurement errors Figure 47: Outside airflow % of design with predicted measurement errors. Figure 48: Comparison of direct airflow measurement techniques. Figure 49: Supply and outside airflows with predicted measurement errors. Figure 50: Outside airflow % of design with predicted measurement errors. Figure 51: Comparison of direct airflow measurement techniques. Figure 52: Supply and outside airflows with predicted measurement errors. Figure 53: Outside airflow % of design with predicted measurement errors Figure 54: Comparison of direct airflow measurement techniques Figure 55: Supply and outside airflows with predicted measurement errors. Figure 56: Outside airflow % of design with predicted measurement errors Figure 57: Comparison of direct airflow measurement techniques. Figure 58: Supply and outside airflows with predicted measurement errors Figure 59: Outside airflow % of design with predicted measurement errors. Figure 60: Comparison of direct airflow measurement techniques. Figure 61: Supply and outside airflows with predicted measurement enrors. Figure 62: Outside airflow % of design with predicted measurement errors. Figure 63: Comparison of direct airflow measurement techniques. Figure 64: Supply and outside airflows with predicted measurement errors. Figure 65: Outside airflow % of design with predicted measurement errors...-. Figure 66: Comparison of direct airflow measurement techniques. Figure 67: Supply and outside airflows with predicted measurement errors. Figure 68: Outside airflow % of design witfi predicted measurement en~ors. Figure 69: Comparison of direct airflow measurement techniques. Figure 70: Supply and outside airflows with predicted measurement errors. Figure 71: Outside airflow % of design with predicted measurement errors Figure 72: Comparison of direct airflow measurement techniques. Figure 73: Supply and outside airflows with predicted measurement errors Figure 74: Outside airflow % of design with predicted measurement errors. Figure 75: Comparison of direct airflow measurement techniques. Figure 76: Supply and outside airflows with predicted measurement errors. Figure 77: Outside airflow % of design with predicted measurement errors. Figure 78: Comparison of direct airflow measurement techniques. Figure 79: Supply and outside airflows with predicted measurement errors. Figure 80: Outside airflow % of design with predicted measurement errors. Figure 81: Comparison of direct airflow measurement techniques Figure 82: Supply and outside airflows with predicted measurement errors. Figure 83: Outside airflow % of design with predicted measurement errors. Figure 84: Comparison of direct airflow measurement techniques. Figure 85: Supply and outside airflows with predicted measurement errors. Figure 86: Outside airflow % of design with predicted measurement errors. Figure 87: Comparison of direct airflow measurement techniques

81 82 82 82 83 83 83 84 84 84 85 85 85 86 86 86 87 87 87 88 88 88 89 89 89 90 90 90 91 91 91 92 92 92 93 93 93 94 94 94 95 95 95 96 96 96 97 97 97 98 98 98

vii

INTRODUCTION

Due to an increased concern about maintaining acceptable indoor air quality and meeting ventilation codes and standards, the accurate control and measurement of outside air intake rates has come to the forefront of attention of several HVAC engineers and designers. Unfortunately, the necessary monitoring equipment and control logic to maintain minimum outdoor intake rates are often nonexistent or are used improperly if they are installed. Consequently, several commercial buildings, and in particular those with Variable Air Volume (VAV) systems, have been found to have inadequate ventilation (Sterling et al. 1992). Unlike Constant Air Volume (CAV) systems, VAV systems present the additional challenge caused by varying pressure in the mixed air plenum that make techniques used to control outdoor air intake in CAV systems ineffective. ASHRAE Standard 62-1989 establishes minimum outdoor air ventilation rates for acceptable Indoor Air Quality (IAQ) standards within buildings. Section 5.4 of the standard requires that "when the supply air is reduced during times the space is occupied (e.g., in variable-air-volume systems), provision shall be made to maintain acceptable indoor air quality throughout the occupied zone" (ASHRAE 1989). Measurement and control of the outside air intake rate is one method to address this requirement. 1.1 Project Description Several techniques exist to measure and control minimum outdoor air intake rates for VAV systems in accordance with the ventilation requirements of ASHRAE Standard 62-1989. However, guidelines are lacking to assess the accuracy and ease of implementation of each of these techniques. In this project, theoretical and experimental analyses are performed to evaluate several techniques for measuring and controlling minimum outdoor air intake rates in VAV systems. The findings of this research will be used to develop a set of guidelines intended for use by HVAC system designers and building operators to ensure that minimum outdoor air intake rates are provided by VAV systems. 1.1.1 Outside Air Intake Rate Measurements Measurements of outside air intake rates are difficult for several reasons. To minimize the possibility of entrainment of rain, snow and other debris, the outside air intake rate face velocity is typically limited to 400 fpm when standard louvers are used. However, low flow rates are difficult to measure with standard pressure sensing devices due to their inherent inaccuracies at low pressures.

Additionally, in most HVAC systems, fans are located near exterior walls which prohibits reducing the duct size to increase the outside air velocity (Solberg et al. 1990). Several other factors hamper the ability to accurately measure outside air intake rates. The wide range of temperature and humidity that is present in outside air makes the calibration of some instruments difficult. In addition, complex ductwork often limits acceptable measurement sites and contributes to non-uniform velocity profiles in the air stream (Janu et al. 1995). Moreover, all the techniques described and analyzed in this report are evaluated with respect to their ability to control minimum outdoor airflow rates in terms of volume per unit of time, e.g. cfm or L/s, since almost all codes and standards establish minimum ventilation rates using these units. Some would argue that the true operative ventilation criterion should be mass flow rate, not volumetric flow rate, since the purpose of the ventilation is to dilute pollutants whose source strengths are measured as mass emissions per unit time. However, it is beyond the scope of this project to discuss this issue, but is should be kept in mind when inaccuracies caused by outdoor air density variations are considered. 1.1.2 VAV Control Techniques In addition to the difficulties in maintaining minimum outdoor air intake rates described above (which apply to almost all HVAC system types), in VAV systems, changing supply air volumes can also cause the pressure in the mixed air plenum to fluctuate. This varying pressure in the mixed air plenum leads to a reduction in the outdoor air intake rate for several of the common control strategies (e.g. fixed minimum damper positions). 1.2 Report Organization Section 2 of this report presents a literature review of existing airflow measurement methods and VAV control techniques. A complete error analysis is outlined in Section 3 for the measurement and control techniques that have been implemented in existing VAV systems. Section 4 provides a description of the laboratory set up and the preparatory work that was performed for each set of tests. Laboratory test results and analyses are summarized in Section 5. Finally, conclusions and recommendations are presented in Section 6.

LITERATURE REVIEW

This section presents a comprehensive literature review of airflow measurement techniques and VAV control methods. A list of references is included at the end of this report in Section 7. In this chapter, only the measurement and control techniques that were tested in the laboratory are described in detail. Brief descriptions of several other measurement and control techniques are discussed for completeness. 2.1 Airflow Measurement Techniques Several techniques exist for measuring airflow in HVAC systems. These techniques can be divided into two categories: direct and indirect measurement techniques. The first method measures the airflow directly, using for instance an anemometer. The second method measures other parameters that are dependent upon the airflow, such as the energy balance method. It would be difficult and impractical to attempt to test and compare all of the airflow measurement techniques that are available. Therefore, only the techniques that were tested in the laboratory for this project are described in detail in Section 2.1.1 and 2.1.2. A brief listing of other possible measurement techniques is included in Section 2.1.3. 2.1.1 Direct Airflow Measurement Techniques Direct airflow measurement techniques require well-developed airflow profiles for accurate readings (Drees et al. 1992). Generally, existing systems do not have enough space within tfie outside air ducts to meet manufacturers minimum recommendations for installing these flow measurement stations. This fact coupled with the low air velocities found in outside air ducts are the primary obstacles for measuring accurately ventilation air directly. Table 4 in Section 14 of ASHRAE Fundamentals lists several airflow measurement techniques, their applications, and expected accuracy (ASHRAE 1997a). ASHRAE Standard 11188 also provides a comprehensive summary of measurement techniques, their expected accuracy, and limitations (ASHRAE 1988). 2.1.1.1 Averaging Pitot-tube array Averaging Pitot-tube arrays are based upon the fundamental airflow measurement device, the Pitot-static tube. Figure 1 illustrates a typical laboratory pitot-static tube. 3

Total Pressure " Inlet

I Static Pressure Inlets

Air Duct

a
Outlet for Static Pressure Measurements

^ t
Outlet for Total Pressure Measurements

Figure 1: Typical Pitot-static tube. Ower and Pankhurst (1977) found that air could be treated as an incompressible fluid with negligible error for flow rates typically encountered in building HVAC applications. Therefore, the airflow velocity can be calculated by: y = where: V gc = Vp = p =
Sc p

Equation 1

Airflow velocity [fpm] 32.17Ubmft/(s2lbM Velocity pressure [inW. G.J Density [ibtr/ft3]

Inserting the value of the constant, g c , and the unit conversions into Equation 1 gives: V = 1096.7 } - ? Equation 2

From the ideal gas law, the density of the air stream can be found from Equation 3: p = R>T where: p = Atmospheric Pressure Ipsia] R = Gas constant for air, 53.347 [ft lbf/(lbmR)] T = Absolute temperature fR] Equation 3 neglects any effects of the air humidity on the air density. This assumption introduces minimal error and will be neglected (see Appendix B). Inserting the value of the gas constant, R, and the unit conversions into Equation 3 gives: 4 Equation 3

p = 2.6993-
Substituting Equation 4 into Equation 2 gives:

Equation 4

\V-T V = 667.517 M

Equation 5

An averaging Pitot-tube array averages both the total and static pressures throughout the duct and then calculates the airflow velocity from the difference between the two pressures, or the velocity pressure. Some inaccuracies are inherent with this method due to the nonlinear relationship between the velocity pressure and the airflow velocity. Specifically, averaging the velocity pressure readings across the duct before calculating the velocity from Equation 5 introduces some errors in the measurement. This is especially true when a nonuniform velocity profile exists in the duct. To reduce this error, manufacturers installation guidelines typically require 7.5 duct diameters upstream of the Pitottube and 3 duct diameters downstream from a disturbance in the flow, such as an elbow in the ductwork (ASHRAE 1997a). Typically, averaging Pitot-tube arrays are not accurate for flow rates below 600-800 fpm unless amplifying sensors and/or auto-zeroing and temperature compensated differential pressure transmitters are used (Drees et al. 1992, ASHRAE 1988). Obviously, such instrumentation is expensive and may not be practical for typical building installations. Additionally, small errors in the differential pressure transmitters can result in large errors in the calculated flow rate. Typical errors in averaging Pitot-tube arrays range from 1 to 5% of full scale, with larger errors at lower flow rates (Janu et al. 1995). 2.1.1.2 Thermal Anemometry Thermal devices such as thermistors, hot-wires and hot-films have long been used as a method to measure flow rates. They have the capability to accurately measure flow rates as low as 1 to 10 fpm (ASHRAE 1997a, Haines 1994). Other benefits include a much closer approximation to constant error as a percentage of airflow than for Pitot-tubes (Solberg et al. 1990). Additionally, the measured flow signals are linear and electronic which allows for simple use with DDC controllers (Solberg et al. 1990). The theory of thermal anemometry is relatively simple; a heated device loses heat at a rate determined by the temperature and velocity of the fluid being measured. Ower and Pankhurst (1977) provide a good description of the physical relationships underlying the theory. Traditionally, the flow rate is calculated from King's Law (Beckwith et al. 1993):

e02 = C + Dsjp-V
5

Equation 6

where: e0 C,D P V

Output voltage [v] Constants Density of fluid Db/f?] Fluid velocity [ft/sj

Flow rates determined from King's Law are typically subject to non-linear output signals and are dependent on the ambient temperature of the fluid. For these reasons, the span and temperature ranges of calibrated thermal anemometers are somewhat limited. There are other drawbacks for the use of thermal anemometer flow measuring stations. These drawbacks include their high sensitivity to turbulence in the air stream and the complicated field calibration procedures that must be performed frequently (Drees et al. 1992, ASHRAE 1997a, Maki et al. 1997, ASHRAE 1988). Electronic thermal anemometry attempts to account for some of these limitations by converting the analog signal from the sensor to a digital signal. If the ambient temperature of the fluid stream is also measured, corrections can be made and a linear output signal can be generated. Experimental results by Dress et al. (1992) indicate that calculated outside air intake rate errors as large as 50% were possible at lower outside air temperatures when using thermal anemometry. No negative effects due to low outside air temperatures were found in the JCEM Laboratory. See Section 4.3.3 for details of these findings. 2.1.2 Indirect Airflow Measurement Techniques Indirect methods provide an alternative for the measurement of outside airflow intake rates, which are typically difficult to measure directly. Two indirect measurement techniques are described in this section: the energy balance and the concentration balance. 2.1.2.1 Energy Balance Assuming adiabatic mixing of the return airflow and outside airflow in an HVAC system, the outside airflow rate can be determined by performing a mass and energy balance on the air streams:

CFMOA = CFMSA
where:
CFMOA CFMSA
PMA POA hRA

PMA

hDA-h *RA
\ "RA

MA
"OA )

Equation 7

PoA

hm hoA

Outside airflow rate [CFM] Supply airflow rate [CFM] Density of mixed air pbjff] Density of outside airpbn/ft*] Enthalpy of return air[BTU/lbda] Enthalpy of mixed air [BTU/lbdal Enthalpy of outside air [BTU/lbda}

At low outside air intake rates, the quantities on the right side of Equation 7 may be more easily measured than the outside airflow rate directly (i.e. CFMOA). This approach has the advantage that it is easily incorporated into today's microprocessor-based control systems. Moreover, temperatures are easily measured with standard instrumentation, with the exception of the mixed air temperature. Indeed, the temperature of the mixed air must be measured before passing through any coils or fans. Hence, stratification of the air can make accurate temperature reading difficult to obtain (Drees et al. 1992). This issue will be addressed further in the error analysis for this technique presented in Section 3.2.2. Equation 7 indicates large errors in the calculated outside air intake rate are possible when the value of TRA-TOA becomes small. Drees et al. (1992) found experimentally that the calculated outside air intake rates were within 10% of full-scale reading using an averaging Pitot-tube array reading. These results were somewhat limited since they are based on only two days of measurements. The supply airflow rate in Equation 7 can be measured directly by an averaging Pitot-tube array or other measurement technique. The air density can be calculated from Equation 4 while the enthalpy can be calculated from Equation 8 (Kreider and Rabl 1993):

h = 0.24 (Td - 459.67)+W (1061.2 + 0.444 (Td - 459.67))

where: Ta = Dry-bulb temperature pR] W = Humidity Ratio [Iby/lbda], given by:

Equation 8

W = 0.62198

0'pVd/ P-Q-PsatVd)

Equation 9

where: $ = Relative humidity f%] p = Atmospheric pressure [psia] PsafW = Saturation pressure at dry-bulb temperature Td [psia], given by:
A: (l_1165.67/

p 5fl/ (r rf ) = 3,226-10 ^
where:

/d)

Equation 10

# = 4.39553-3.469

( Td \ A + 3.072 - ^ - | - 0 . 8 8 3 3 - [ - ^ - | m 1000J uoooj UoooJ

Equation 11

Substituting Equation 4 and Equations 8 through 11 into Equation 7, the can be related to quantities that can be measured directly. The equation is rather large and cumbersome and will not be included here. by neglecting the changes in the humidity ratio, W, in Equation 8, the balance can be simplified to: 7

enthalpy resulting Instead, enthalpy

CFMQ4=CFMS4

PsA PoA

TRA

^SA

Equation 12

TRA ~TQA J

The energy balance technique then becomes a temperature balance. associated with this simplification will be addressed in Section 3.2.2.1. 2.1.2.2 Concentration Balance

Errors

Another indirect method for outside air intake rate measurement is using a concentration balance based on C0 2 concentration levels. Numerous papers have been published dealing with this topic including Drees et al. (1992), Elovitz (1995), Janu et al. (1995), Ke and Mumma (1997a), Ke et al. (1997b), Meckler (1994), and Persily (1993). In the C 0 2 concentration balance model, the outside air intake rate is based on a volume balance of the air streams and is given by Equation 13:
^^2RA
C0

CFM^^CFM^

^2SA

Equation 13
J

COlRA ~ 20A

This model assumes that the densities of all air streams are equal which may not be an accurate assumption (Ke and Mumma 1997a). Similar to the energy balance method, when the value of CO2RA-CO20A becomes small, errors in the calculated outside air intake rate become very large (Janu et al. 1995). Additional information regarding the errors associated with this measurement technique is included in Section 3.2.2.2. 2.1.3 Other Airflow Measurement Techniques Many other methods exist to measure airflow. Brief descriptions of several of these various techniques are presented below. The reader is referred to the references for further information- regarding these measurement techniques. 2.1.3.1 Rotating vane and propeller anemometers These anemometers contain wind-driven wheels with mechanical, electrical, or magnetic pick-ups for measuring flow rates. Typical vane anemometers have an accuracy of 2 to 5% for velocities from 100 to 3000 fpm (ASHRAE 1997a, Dols and Persily 1995, Ower and Pankhurst 1977). However, they are very susceptible to changes in flow rates and the instrumentation is very delicate which limits its usefulness for a typical air distribution system. Additionally, these devices require frequent calibration. Use of these anemometers is not recommended for measurements within ducts due to their large area (ASHRAE 1988). 8

2.1.3.2 Swinging vane anemometers These devices contain a pivoted vane attached to a resistive hairspring. Deflections of the spring correlate to a reading displaced on an indicating scale. Total pressure readings are possible for a velocity range of 50 to 10,000 fpm with a typical error of 10% (ASHRAE 1988). Readings tend to be high on the suction side and low on the discharge side of a fan. Swinging vane anemometers are not recommended for critical measurements (ASHRAE 1988). 2.1.3.3 Vortex shedding meters These meters use the eddy-shedding principle to measure the velocity of air streams. To create the eddies, or "Karmen Vortices", an object is inserted in the flow stream and piezoelectric pressure transmitters measure the periodic pressure fluctuations which are then converted to velocities using calibration constants (Ower and Pankhurst 1977). While high accuracy can be achieved for flows with large Reynolds numbers (e.g. 0.5% at Re=104), at lower flow rates this accuracy drops off to the point where these meters are no longer useful. The accuracy of the pressure transmitters used to measure the fluctuations further reduces the accuracy of the meter. As a rule of thumb, this method is not recommended for airflow rates of less than 500 fpm. 2.1.3.4 Integrated damper and measuring devices Typically, these units contain a pressure-based measurement device similar to an averaging Pitot-tube array built into a damper. The measured flow can then be used to control the damper to maintain a constant flow rate. Errors and limitations on the use of these devices are similar to that of the averaging Pitot-tube array. 2.1.3.5 Laser doppler anemometry (LDA) LDA and a similar operating fiber optic system are very accurate measurement techniques, but prohibitively expensive for both the required equipment and operation. Mease et al. (1992) provide a detailed description of the measurement technique. Errors less than 1% at velocities as low as 15 fpm are possible with this measurement technique. Typically, these systems are used for calibrating other calibration systems (ASHRAE 1997a).

2.1.3.6 Orifice Meters Air flow rates can be measured using orifice meters. ASME Standard MFC-3M (1989) describes measurement of flow through pipes, ducts, and plenums for all fluids using the orifice, nozzle, and venturi. ASME Standard PTC 19.5 specifies their construction. The measurement of flow rates by orifice meters are determined by measuring the pressure difference across the orifice. The accuracy of orifice meters is 1 % as long as the Reynolds number is above 500. The main limitation of the orifice meters is that the determination of the discharge coefficient and thus the accuracy of the measurement depends on the installation conditions. 2.1.4 Accuracy of Airflow Measurement Techniques Based on the literature review and the manufacturers' declared specifications, Table 1 summarizes the accuracy and the range of the direct measurement techniques for airflow rates discussed above. As will be discussed in Chapter 3, the accuracy of the indirect measurement techniques (e.g., enthalpy balance and concentration balance) depends on various factors including the accuracy of both a direct measurement technique (to measure the supply airflow rates) and a measurement technique of the temperature (for the enthalpy balance) or of the C0 2 (for the concentration balance). Table 1: Accuracy and Range of Airflow Measurement Techniques.
Technique Pitot-Tube Thermal Anemometer Rotating Vanes Anemometer Swinging Vanes Anemometer Vortex Shedding Meter range 200 fpm - 9,000 fpm >1fpm 100 fpm-3,000 fpm 50 fpm-10,000 fpm > 500 fpm Accuracy 1%-5% 2%-5% 2% - 5% 10% 1%-5% 1%-5% 1%-3% 1%-5% Comments For low flows (200-600 fpm), high accuracy DP is needed. Sensitive to turbulence. Need frequent calibration Susceptible to changes in flow rates. Need periodic calibration. Not sufficiently accurate for OA measurements. Not accurate for low flow rates. Same errors and limitation that Pitot-Tube. Accurate at low rates. Too costly for field applications. Accuracy is affected by installation conditions.

Integrated 200 fpm - 9,000 fpm damper/measuring device 1 fpm - 5,000 fpm Laser Doppler Anemometer Orifice Meter > 20 fpm

10

2.2

VAV Outdoor Air Control Techniques

Several techniques for controlling minimum outdoor intake rates in VAV systems have been used in the field or suggested in the literature. In this section, detailed descriptions of the proposed control strategies to be tested in the laboratory are presented. For completeness, brief descriptions of several other control techniques are included at the end of this section. A complete error analysis for each proposed control strategy is included in Section 3.3. 2.2.1 Fixed Minimum Outdoor Air Damper Position This system is shown schematically in Figure 2. A fixed outside air damper position is probably the most common method used to try to meet minimum outside airflow intake rates in VAV systems. For systems without outdoor air economizers, the analogous design is simply a fixed minimum outdoor air intake louver and/or damper. This control technique will be used as the base case for this project. Under design flow conditions, the outside air damper is positioned to meet the minimum outside air requirements. This predetermined damper position is then used when only minimum outside airflow is required, even as the supply fan speed is reduced in the VAV system.
,EA

i
////
VFD VFD

^ A

DA \

AHU

SA^

Figure 2: Outside airflow rate control schematic for fixed position outside air damper. In VAV systems, this control method will not deliver the minimum outside air intake due the variation in the static pressure of the mixing plenum (Drees et al. 1992, Mumma and Wong 1990). Outside air intake rates will be much closer to a constant percentage of supply air than a constant volume flow rate (Jahu et al. 1995). Another problem with this method is stack and wind effects on the outside air intake rate (Solberg et al. 1990). Because of the limitations inherent to the technique of using a fixed minimum outside damper position, Ke et al. (1997b) found through simulation that this method was the least effective control strategy to maintain a minimum outside airflow rate among those commonly used. 2.2.2 VAV Control Techniques for Economizer Systems The next three control strategies are for systems where the outside air duct is sized for use during an economizer cycle. The size of the outside air duct

11

must be large enough to safely provide 100% of the design flow. This large size, however, results in very low airflow velocities during minimum outside air intake rate mode. 2.2.2.1 Volumetric Fan Tracking Volumetric fan tracking is a common control strategy used to provide minimum outside airflow rates for VAV systems with economizers and return air fans. Figure 3 shows a schematic of the volumetric fan tracking system. The flow measurement stations, AFS-1 and AFS-2, measure the supply and return airflow rates, respectively. The return fan speed is controlled (AFC-1) to maintain a fixed differential in the return airflow rate compared to the supply airflow rate. The preset fixed differential in the return and supply airflow rates must then be made up by outside air. Damper positions for the return, exhaust, and outside air are generally set to fixed positions during minimum outside air intake mode. The fixed flow differential is typically based upon the initial system air balancing. The outside air provided to the space maintains a slight positive static pressure within the building to reduce unwanted infiltration.
EA

I
////
I

i^VFD|- - {AFt>T}- HAFS-gj

VFD

L--|AFS-1|

DA \

AHU

SA.

Figure 3: Outside airflow rate control schematic for system using volumetric fan tracking control strategy. With this control strategy, the differential air flow between supply air and return air is technically what is required to achieve the desired level of minimum building pressurization to minimize infiltration; it is not however necessarily equal to the minimum outdoor air rate. In fact, this control strategy only works conceptually when the minimum outdoor air rate is equal to the amount of air desired for pressurization. The control strategy does not work if the exhaust air rate is not equal to zero during minimum outdoor air operation since the exhaust air rate is not measured. For instance, in an application requiring a large fraction of outdoor air, such as an assembly space, the minimum outdoor air rate may exceed that which will mildly pressurize the building. Some exhaust air is required to prevent over-pressurization. In this case, the differential between supply air and return air will be less than the required outdoor air rate. Since the outdoor air and exhaust air rates are not directly measured, it is not possible to control the outdoor air at or above the minimum required level using the fan tracking control strategy;

12

Volumetric tracking is one of the more common control methods used in VAV systems today (Kettler 1995, Avery 1992). The benefit of this method is that velocities in the supply and return ducts are generally large enough that standard flow measuring techniques can be sufficiently accurate, provided adequate sections of straight ductwork is available to avoid excessive turbulence. However, several authors have alluded to weaknesses in this control method. Elovitz (1995) states that even small measurement errors in large flow rates can translate to large errors in the calculated outside air intake rates and that a fixed differential flow is not versatile enough to account for exhaust and leakage flow rate changes. Janu et al. (1995) recommend that online measurement of outside air intake rates be provided and that the differential flow vary to compensate for operation of variable exhaust flows. Finally, Kettler (1995) makes the argument that when typical measurement errors are accounted for, the outside air intake rate can vary by as much as 35%. 2.2.2.2 Measurement and Control of Outside Airflow Rate with Economizer A typical arrangement for this type of system is shown in Figure 4. The outside air duct is sized to allow for economizer control, i.e. capable of supplying 100% outdoor air. During minimum outside airflow intake mode, a flow measurement station (AFS-1) records the flow of outside air and controls the return and outside dampers (M-1 and M-2, respectively) to maintain the required minimum outside airflow intake rate.
,EA

. RA '!
/ / / /
VFD

|M-a|

[AFS-I]

VFD

DA

AHU

SA,

Figure 4: Outside airflow rate control schematic for system with economizer damper. Due to the relatively large size of the outside air duct in this system, the measurement of the outside air intake rate at the flow measurement station (AFS-1) can be difficult. The accuracy of this control technique depends directly upon the accuracy with which the outside air intake rate can be measured.

13

2.2.2.3 Plenum-Pressure Control This method relies upon instrumentation such as a manometer or differential pressure transmitter to measure the pressure drop across a fixed orifice. By maintaining a constant pressure drop, the minimum outside airflow requirements can be met (Janu et al. 1995, Haines 1994, Elovitz 1995). It can be implemented either in a dedicated ductwork or in an existing economizer duct. The fixed orifice in this case is the outdoor air damper or the combination of the outside air louver (L-1) and the outdoor air damper as suggested by Mumma and Wong (1990). This system is shown schematically in Figure 5. The pressure drop must be large enough so it can be accurately measured but not so large to create an excessive energy penalty (Ower and Pankhurst 1977, Kettle 1998). The differential pressure transmitter (DP-1) measures the pressure drop and the return air damper is controlled to maintain a constant value determined during an air balance test that results in the desired minimum outdoor air flow. If an actuator is not located on the return air damper, one must be added.

^ '1
r -Wl\ |DP-l| L-1 . ////
VFD VFD

.RA

DA

AHU

SA.

Figure 5: Plenum-pressure control schematic.

The outside airflow intake rate is related to the pressure drop across the damper by Equation 14 (ASHRAE 1997b): V = 1096.7. where: V = Velocity ppm]
Apj = Total pressure loss [inW. G.J P = Density pb^] C local loss coefficient [-]

\p-C

Equation 14

14

For a fixed damper position, the value of the loss coefficient, C, for the damper is constant (ASHRAE 1997b). To maintain a constant airflow rate, the ratio of the pressure loss to the density of the air stream must remain constant. In practice, the effects of changing densities are usually neglected and only a constant pressure drop is maintained. Moreover, it should be noted that the accuracy of the air flow measurement using the fixed damper position depends on whether a modulating damper or a separate minimum outdoor air damper is used (see section 2.2.3.3). The pressure differential measurement is more accurate when using a separate minimum outdoor damper with only two positions (fully closed or fully open). The fully open position provides a reliable fixed orifice. However, an accurate fixed position achieved with a modulating damper can be difficult to obtain due to hysteresis and linkage slip. Further discussion of the characterization of flow through a modulated outside air damper is provided in section 2.2.4.3. 2.2.3 VAV Control Techniques for Systems with a Dedicated Outside Air Duct The next two control strategies attempt to remedy the main disadvantage of an HVAC system equipped with only one outside air duct. By adding another duct through which only the minimum outside air must flow, the size can be made much smaller, thereby increasing the airflow velocities and thus making them easier to measure. Typically, the larger duct is used only during economizer control mode and is closed when minimum outside air intake rates are required. 2.2.3.1 Measurement and Control of Dedicated Minimum Outside Duct Airflow Rate This system is shown schematically in Figure 6. In economizer mode, the damper on the larger outside air duct is controlled to regulate the outside air intake rate. During minimum outside air intake mode, the dedicated outside airflow intake duct is opened while the damper in the larger outside air duct is closed. A flow measurement station (AFS-1) records the outside airflow rate and controls the return (M-1) and the dedicated outside air dampers (M-2) to maintain the minimum outside airflow intake rate.
RA

[M^T}-

////

VFD VFD

|M-2| DA '/

1 AFS-11 AHU

SA.

Figure 6: Outside airflow rate control schematic for system with dedicated minimum outside airflow duct. 15

2.2.3.2 Outside Air Injection Fan This system is illustrated in Figure 7. A dedicated minimum outside airflow intake duct contains a fan used to control outside airflow whenever it is required. A flow measurement station (AFS-1) installed in the dedicated outside airflow duct measures the outside airflow rate. To maintain minimum outside airflow intake rates, the airflow in the dedicated duct can is controlled by either: a variable frequency drive (VFD) on the injection fan as shown in Figure 7. a dedicated outside air dampers (M-2) as illustrated in Figure 6. It may be argued that the VFD fan is not sufficient for two reasons: (i) the air will be pulled through the fan even when the fan is not operating since the pressure in the mixing air plenum is negative , and (ii) a damper is required by Standard 90.1 to prevent infiltration through the outdoor air intake when the air handling unit is not operating. With the damper (M-2), the VFD is not required to control air flow (thus this control strategy is the same as the one illustrated in Figure 6).

*t
////
[VFDJ fvFDl AHU

.RA

|VFDHAFC-I

\-

iArs-i|

(A\ DA

\\\

SA.

Figure 7: Outside airflow rate control schematic for system with dedicated minimum outside airflow injection fan. The fan is chosen such that it has a very flat fan curve and operates almost as a constant volume fan over the expected range of pressures (Elovitz 1995, Avery 1989). In cases where the fan curve is very flat or the expected variation in the plenum pressure is small relative to the overall fan operating pressure, the volume rate of the fan may be sufficiently constant that an active outdoor air volume measurement and control system may not be necessary. 2.2.3.3 Plenum-Pressure Control This system is identical to that described in section 2.2.2.3 except that the fixed orifice is a dedicated two-position minimum outdoor air damper (possibly in combination with the outdoor air louver) rather than a fixed damper position on

16

the large economizer outdoor air damper. In cases where there is no outdoor air economizer and a return air damper does not exist, one must be added. This design is generally more accurate than that described in section 2.2.2.3 as explained in that section. This design is similar to that illustrated in Figure 6 (but without the AFS and with a controller modulating the return air damper). This particular design was used in the error analysis and laboratory testing described in later parts of this report. 2.2.4 Other VAV Control Techniques Control strategies that were not tested in the laboratory are briefly described in this section. For more information on these techniques, the reader is directed to the sited references. 2.2.4.1 Minimum outside air damper position reset This control strategy attempts to compensate for the main limitation of the fixed outside air damper control strategy by allowing the damper position to be reset based upon the supply fan speed. The position of the outside air damper position can be found from either be a linear relationship with the supply fan speed, or a higher order polynomial equation. By varying the position of the damper, the pressure drop across it can be kept constant and therefore the outside airflow intake rate will also remain constant. As With the fixed minimum outside damper position control method, online measurement of the outside air intake rates are not required. However, the minimum ventilation rate may not be met if the supply airflow rate falls too low (Ke and Mumma 1997a). Additionally, since the damper and duct are often the same size, small changes in damper position translate to large changes in flow rates and normal hysteresis can significantly affect the outside air intake rates (Drees et al. 1992). Finally, this control strategy can not account for wind and stack effects on the system. See Solberg et al. (1990) for additional details regarding these errors. 2.2.4.2 Supply/Return fan speed or vane position matching The supply and return fan speeds are controlled, often off the same control signal, to match each other with a fixed differential to maintain a slight positive pressurization. The outside airflow rate is then equal to the difference between the supply and return airflow rates. However, similar to the volumetric tracking control strategy, this is only true when there is no exhaust airflow. Whenever the exhaust flow is greater than zero when in the minimum outdoor air mode, the minimum outside air intake rate cannot be maintained with any reasonable degree of accuracy. While this control method is inexpensive and easy to implement on existing systems, it generally has been unacceptable due to mismatched fan flow characteristics over the typical range of operation (Janu etal. 1995). 17

2.2.4.3 Characterization of flow through a modulated outside air damper By characterizing the outside air intake rate as a function of both the position and pressure drop across the damper, accurate control of ventilation air can be obtained over a wide range of operating conditions. However, this process requires significant amounts of time to properly characterize the airflow rates. In addition, this method is subject to calibration drifts in transmitters and positioners, as well as looseness and hysteresis, any of which can cause substantial errors (Janu et al. 1995).

18

ERROR ANALYSIS

Performing detailed error analyses for airflow measurement and outside air intake rate control techniques in variable air volume systems allows for a theoretical comparison of their practicality. Techniques that prove in theory to be invalid may not need to be tested further in a laboratory or real building environment. A search of existing literature on the topic of airflow measurement and VAV control error analysis provided relatively little on the subject. Solberg et al. (1990) have performed detailed analyses regarding the effects of wind loading and stack effects on various control techniques commonly used today. Kettler (1995) also provides some rough error estimates for the volumetric fan-tracking control method. Drees et al. (1992) provide graphs of predicted error for the indirect outside airflow measurement techniques using temperature and concentration balances based on return, mixed and supply flows. Several simplifying assumptions are made in Drees's analysis that will be addressed in this report. 3.1 Uncertainties

Propagation of uncertainty is the primary method used throughout this report to perform the error analysis. A brief description of this method is presented here for the reader. Uncertainties in calculated and predicted results depend upon the uncertainty in the values used to find the result. If the value y is a function of independent variables (xi, X& ..., x), then the uncertainty in y can be determined by analyzing the propagation of uncertainty. Stated mathematically, given:

y - / ( * i , x2 ,K , xn )

Equation 15

the uncertainty in y, uy, can be approximated from the uncertainties in the x values, (oi, Uz, .... un), by the following equation (Taylor 1982):
2

^H,

+ \ -9u2 +K + {dx2 )

(dy

19

This method assumes that all the uncertainties are independent and occur with equal probability. According to the ANSI/ASME Standard on measurement uncertainty, there are three sources of uncertainty in airflow measurements: calibration, data acquisition, and data reduction (ANSI/ASME 1983). Uncertainties associated with data reduction (curve fits and rounding) are neglected in the following analyses as these errors are often negligible. Further information regarding the errors associated with the data acquisition system, installation problems, and calibration processes in the JCEM Laboratory is presented below. 3.1.1 Data Acquisition Errors The data acquisition equipment used in the JCEM laboratory consists of a DDC control network. The errors associated with the resolution of this equipment were found by determining the minimum resolution of the database storage system. All sensors used in the lab have either 0-5 V or 4-20 mA output signals. 3.1.2 Installation and Calibration Errors Some of the expected calibration errors for different airflow measurement and control techniques are listed below. Where in situ calibrations could not be performed in the JCEM Laboratory, the manufacturer's declared accuracy was used instead. 3.1.2.1 Averaging Pitot-tube arrays Possible errors related to the installation and calibration of an averaging Pitottube array are listed here. Additional information regarding the calibration of the averaging Pitot-tube array in the supply duct in the JCEM Laboratory is included in Section 4.2. Non-uniform flow velocity - A non-uniform flow velocity increases the amount of error in the averaging Pitot-tube array because pressures are averaged across the duct before the velocity is calculated instead of calculating the velocity at each point in the duct and then averaging them. Non-uniform velocities can result from arrays being installed without the proper lengths of unobstructed duct diameters both upstream and downstream. More information regarding the flow profiles of the JCEM Laboratory is included in Section 3.2.1.1. Pitot-static tube alignment - The tip of the Pitot-static tube should be parallel with the direction of airflow movement. A misalignment of the tip in the airflow results in lower velocity pressure measurements. This error is known as the yaw sensitivity of the Pitot-static tube. Loading - When a Pitot-static tube traverse is being performed for calibration purposes, the introduction of the Pitot-static tube into the airflow introduces some amount of error. 20

Compressible vs. incompressible flow - Air is a compressible fluid, but for flow velocities common in HVAC ducts, the assumption is made that air can be treated as an incompressible flow without introducing any significant error.

3.1.2.2 Electronic thermal anemometry Due to the proprietary calibration procedures of electronic thermal anemometer manufacturers, the errors associated with calibrating the instruments are difficult to predict. Some possible installation and calibration errors may include the following: Non-uniform flow velocity - A non-uniform flow velocity increases the amount of error in the averaging readings taken by the sensor. Each duct contains six measurement locations that calculate the flow velocity. The duct airflow rate is then averaged from these calculated values. More information regarding the flow profiles of the JCEM Laboratory is included in Section 3.2.1.1. Turbulence - Thermal anemometers are highly sensitive to turbulence in the air stream. Turbulence effects the heat transfer coefficients and generates false readings using the calibrated algorithms. Air temperature - Since they are based on temperature measurement, the thermal anemometers are sensitive to ambient air temperature. The measured air velocity should be corrected based on the ambient temperature. Typically, these devices include an ambient temperaturesensing RTD or thermistor to provide a "temperature compensated" measurement of the air velocity.

3.1.2.3 Enthalpy balance The calculation of the outside airflow intake rate from .an enthalpy balance is dependent upon several sources of error. Supply airflow rate measurement - Any error associated with the measurement of the supply airflow rate must be taken into account. See Section 3.2.1 for more information regarding direct airflow measurements. Temperature sensor calibration - The calibration of the temperature sensors must also be taken into account. Temperature stratification of airflow - This error is probably most prominent when trying to measure the mixed air temperature. Stratification of the mixed air streams makes the accurate measurement of the temperature difficult.

21

Temperature balance assumption - A minimal error is introduced when the temperature balance replaces the enthalpy balance. Errors associated with this assumption are examined further in Section 3.2.2.1.

3.1.2.4 Concentration balance The calculation of the outside airflow intake rate from a CO2 concentration balance is subject to three main sources of error: Supply airflow rate measurement - Any error associated with the measurement of the supply airflow rate must be taken into account. See Section 3.2.1 for more information regarding direct airflow measurements. CO2 sensor repeatability - CO2 sensors manufacturers typically list a sensor repeatability over a period of one year. For use with the concentration balance technique when using a single sensor all CO2 measurements, the repeatability error of the sensor only needs to be over a period of one day or so. This repeatability error is typically much less than that reported by the manufacturers. See Section 3.2.2.2 and Appendix C for more information regarding the repeatability errors of CO2 sensors. Repeatability can be a significant source of error when multiple CO2 sensors are used. Stratification of CO2 concentration - Similar to the enthalpy balance method, if air streams are not well mixed when a sample is taken for measurement of the C 0 2 concentration, an inaccurate reading may result. However, the problem is typically less severe with this technique than with the enthalpy balance method since the supply air C 0 2 sensor may be located downstream of the supply air fan where mixing is more thorough than the typical mixed air temperature sensor position.

3.2 Airflow Measurement Techniques Error Analyses Included in this section are error analyses for the proposed airflow measurement techniques to be tested in the JCEM Laboratory. The method for calculating the expected error for each measurement strategy is described. Then, the predicted errors are shown specifically for the laboratory setup at the JCEM, using existing instrumentation and duct sizes. Table 2 lists some of the essential characteristics of the JCEM Laboratory used in these error analyses. Section 4.1 and Appendix A provide a complete description of the laboratory.

22

Table 2: JCEM Laboratory conditions used for error analysis.

Design bupr-ly Air [-low Rale Supply Duct Area Return Duct Arc a Econotni7cr Duct Area Dedicated Outsde Duct Area Atmo->fihenc Pressure Suppl, Air Temperature Return Air Temperature Outsidf Air Tmnpurature Supply Air Relative Humidity Outside A r Relative Humidity Outsider C0 2 Level

3.2.1 Direct Airflow MeasurementlTechniques This section includes the expected errors for two different direct airflow measurement techniques; the averaging Pitot-tube array and the electronic thermal anemometer. In an attembt to quantify the effects of a non-uniform flow profile within the ducts on direct airflow measurement techniques, flow profiles from the JCEM Laboratory will be Analyzed first. 3.2.1.1 Velocity profiles Non-uniform flow profiles downstream of disturbances such as elbows in the ductwork can severely affectj the accuracy of direct airflow measurement techniques. To account for this dffect, measurement devices must be installed according to manufacturers' recommendations, or enough measurements must be made across a traverse to get Jhe true flow velocity. Figure 8, Figure 9, and Figure 10 show the velocity prbfile in the economizer duct in the JCEM Laboratory at several flow rates. These profiles where measured with a Pitotstatic tube and micromanometer with an accuracy of +/- 0.00025 inWG. More information regarding these traverses is included in Section 4.2. The uniformity of the velocity profiles in the JCEfi/l Laboratory is better than would typically be found in a building installation.

23

&to^-^^ 4O0l^-^'^

^ ^

^ ^
B 400-500 B 300-400 D 200-300 Q 100-200 Q 0-100

1 ^^^
1

^
^<SS8?
y)- Z4 109 ^"^ / 19-*^-^ySZia 27.9

5 2 0 o i - ^

-<SiS5&i

Vertical Axis tin]

19.4 Horizontal Axis [In]

Figure 8: Outside airflow profile at 707 CFM.

10.9 Vertical Axis [In]

1ft4

194

27.9
27.9

Horizontal Axis 0n]

Figure 9: Outside airflow profile at 1,130 CFM.

,400-500 .300-400

2.4 10.9
194

27.9 Vertical Axis [In] 27.9

Horizontal Axis [Inl

Figure 10: Outside airflow profile at 3,276 CFM. 24

It should be stressed that these airflow profiles are unique for the JCEM Laboratory. The effects of non-uniform flow profiles will typically be much more significant in other installations. Factors such as the number of unobstructed duct lengths before and after the measurement device will need to be considered when the accuracy of other installations is calculated. 3.2.1.2 Averaging Pitot-tube arrays From Section 2.1.1.1, the airflow velocity using an averaging Pitot-tube array is calculated from Equation 17:

V = 667.517-

Wp-T

Equation 17

In buildings where the atmospheric pressure is not measured, minimal errors are introduced if the annual average atmospheric pressure is used in Equation 17. See Appendix B for additional information regarding this simplification. The error in the calculated airflow velocity, u v , is found by substituting Equation 17 into Equation 16 and accounting for the manufacturer's declared accuracy of the Pitot-tube array, upitot:
(

uv

dV

\2 K, Vp

fdV \ (v.u J [?LuT + pl ydT

Equation 18

The uncertainties, up, Uvp, and u-r, for the independent variables of pressure, the velocity pressure, and the temperature in Equation 18 were calculated from Equation 19:
U

x = iU^calibration ' "acquisition calib

2 ,

Equation 19

where: Ucaiibrabon = Uncertainties in instrument calibration "acou/sffion = Uncertainty introduced due to the resolution of the data acquisition system used

Table 3 lists the values for the uncertainties used to predict the error in the averaging Pitot-tube array airflow measurements in the JCEM Laboratory. Table 3: Averaging Pitot-tube array uncertainties for JCEM Laboratory.
"T
^calibration U|csokiLDn

"P

uVp (Outside) 3.6E-10 inWG 0 05 inWG

0 60 .. readirg

t 0 b0% reading 0 25% full scale i 0 03 psi

u "vp (Supply) Pito! 1 00% full scrile + 2'm leading

o?r> r

0 00031 inWG 0 50 inWG

full scale

The predicted errors for the supply and outside duct averaging Pitot-tube arrays in the JCEM Laboratory are shown in Figure 11. The predicted errors are 25

less for the outside air duct due to the higher resolution of the differential pressure transmitter used on this duct. This result illustrates how the proper selection of instrumentation can increase the accuracy of the measured airflow.
15.0% ,

\
13.5% .

\
12.0% . 10.5% . 9.0% . 8 jj X 6.0% . 4.5% . 3.0% . 1.5% . 0 0% . 0 1

\
\

O O S i n W G D P Transmitter

7.5%

I
\
*_

Vx^_
200 400 600 800 1,000

"""--..
1,200 1,400 1,600 Airflow Velocity [fpm]

1,800

Figure 11: Predicted errors of averaging Pitot-tube arrays in the JCEM Laboratory. 3.2.1.3 Electronic Thermal Anemometry Uncertainties associated with measurement of the airflow velocity using electronic thermal anemometry, uv, are calculated from Equation 19, using the manufacturers declared accuracy for the calibration error, Ucaiibration- More accurate values for this error were unavailable due to the proprietary factory calibration procedure of these devices. Table 4 lists the values for these uncertainties and the predicted error of the electronic thermal anemometers is shown in Figure 12. For airflow rates between 200 fpm and 400 fpm, the electronic thermal anemometers provide significantly more accurate measurements than commonly used averaging Pitot-tube arrays. However, when high-accuracy, auto-zeroing differential pressure transmitters are used, averaging Pitot-tube arrays can provide very accurate measurements as illustrated in Figure 11. Table 4: Electronic thermal anemometer uncertainties for JCEM Laboratory.
J

calibration

resolution

10 fpm forflov?s-<500 fpm, 2.0% of reading for flows >500 fpm 1 fpm ~

26

15.0% 13.5% 1Z0% 10.5% 9.0% t 7.5% 6.0% 4.5% 3.0% 1.5% 0.0% 200 400 600 800 1,000 1,200 1,400 1,600 1.800 Air Flow Velocity [fpm]

Figure 12: Predicted errors of electronic thermal anemometry in the JCEM Laboratory. 3.2.2 Indirect Airflow Measurement Techniques This section includes the expected errors for two different indirect airflow measurement techniques; an enthalpy balance and a C0 2 concentration balance. 3.2.2.1 Enthalpy Balance As described in Section 2.1.2.1, the outside airflow intake rate can be calculated from the following equation:

CFMOA = CFMst

PMA

hRA -h tt, 'MA

l

Equation 20

PoA

V hju ~hoA J

The ratio of enthalpy differences in Equation 20 can be simplified to the ratio of the respective dry-bulb temperature differences if changes in the humidity ratio, W, can be neglected. The resulting equation is then:

CFMo^CFM^

PSA

(T

T
RA

^
SA

Equation 21

POA \JRA

~T< OAJ

The errors associated with this assumption are linear with respect the humidity ratio of the supply air. The magnitudes of the errors decline as the outside air fraction increases. Figure 13 shows the error introduced using this assumption for a 20% outside air fraction. Errors up to 4% can be obtained when 27

the effects of humidity are neglected, but errors for typical building operating conditions will usually be less than 1.5%.
4.5% 4.0% 3.5% -a -o Return Air RH = 20% Return Air RH = 60%

A - _ Return Air RH = 40% * Return Air RH= 80%

0.000

0.005

0.010

0.015

0.020

0.025

0.030

Humidify Ratio of Supply Air [lb w flb d .]

Figure 13: Errors introduced by replacing enthalpy balance with a temperature balance at 20% OA fraction. One further simplification can be made to Equation 21 by assuming that the atmospheric pressure of the supply and outside airflow rates are the same. Assuming density is a function of only pressure and temperature (Equation 4), the ratio of the supply and outside airflow densities can be rewritten as the inverse ratio of their absolute temperatures: CFMOA = CFM SA
l

OA

RA

SA

Equation 22
)

SA

TRA ~TQA

This assumption is justified by the fact that the atmospheric pressure in the outside and supply will not likely differ by much, and the error associated with this assumption will be considered negligible. See Appendix B for additional information related to the calculation of the air density. As stated earlier, the measurement of the outside and return air temperatures can typically be accomplished with sufficient accuracy using sensors commonly available today. The measurement of the mixed air temperature, however, can be quite difficult. Due to inadequate mixing of the airflows, stratification can occur in the mixed airflow making temperature measurement difficult even when averaging temperature sensors are used. Quantifying the error associated with the measurement of the mixed air temperature is difficult and is the focus of an upcoming research project 28

(ASHRAE RP-1054). For purposes of this analysis, the uncertainty in the mixed air temperature measurement will be assumed to be 3% of the measured value. However, this error may be less when the outside and return air temperatures are close to one another, or greater when the temperature difference is large. Figure 14 shows the predicted errors for the temperature balance measurement technique in the JCEM Laboratory assuming no error in the measured supply airflow rate and those conditions outlined in Table 2 and Table 5. Table 5: Temperature measurement uncertainties for JCEM Laboratory.
TM
Ucaibration
U

T
0.25 T

T TO 0.25 F

0 50% reading 0 50% reading 3 00% reading 0.25 T

rasolufion

70% . 20% Outside Air Fraction -40% Outside Air Fraction

60%
50%

40% o

41
^30%

20%

10%

0%

10

15

20

25

30

35

40

45

50

Figure 14: Predicted errors of temperature balance in the JCEM Laboratory. From the predicted errors shown in Figure 14, it is clear that using the temperature balance method to measure outside airflow rates will not yield accurate results under most typical building operating conditions. Errors shown in Figure 14 assume no errors in the measured supply airflow rate. This is a simplifying assumption, and actual errors will be larger than shown; especially during times of reduced supply airflow when velocities are lower and more difficult to measure with the common averaging Pitot-tube array.

29

3.2.2.2 Concentration balance Similar to the enthalpy balance applied to airflow rates in an HVAC system, a CO2 concentration balance can be used to calculate the outside airflow rate as shown in Equation 23: CFMOA=CFMI SA
CO2RA CO2SA
^^2RA~^^20A

Equation 23

This measurement is performed using one sensor to measure all three CO2 concentration values. Using a multiple CO2 sensors to determine the outside airflow rate is not sufficiently accurate due to the relatively large error associated with the absolute accuracy of commonly available sensors. When only one sensor is used, however, the absolute errors cancel out of Equation 23. The only source of error associated with the sensor then becomes its repeatability. The use of only one sensor, however, increases the time required to calculate the outside airflow rate. Each airflow must be sampled by the sensor before the outside airflow rate can be calculated, and each airflow typically requires 2 to 3 minutes to be measured with reasonable accuracy.1 Available sensors typically have repeatability errors on the order of 20 to 40 ppm, but this repeatability value is typically based on a one year period. The repeatability within the time frame required for calculation the outside airflow rate from Equation 23 is closer to 2 to 5 ppm. This significantly reduces the error in the calculated outside airflow rate. Errors associated with the supply airflow measurement and data acquisition must still be considered. Table 6 shows the predicted minimum CO2 concentration difference between the return and outside airflow so that errors of less than 15% can be achieved using the concentration balance method. See Appendix C for test results verifying the concentration balance airflow measurement technique. Table 6: Minimum C0 2 differences (ppm) between the return and outside airflow for 15% accuracy using the concentration balance method (data acquisition error of 1 ppm, sensor repeatability of 3 ppm).
5% Uncertainty in SA Measurement s OA Fraction COaA-CO^fppm) 10%"-' 302 20% ~ 145 30% 94 40% 69 10% Uncertainty in SA Measurement '10% 382" 20% 183" ' 30% 119 40%" 87~

Figure 15 shows the predicted errors in the calculation of the outside airflow intake rate for the JCEM Laboratory, using the parameters listed in Table 2.

Testing in the JCEM Laboratory showed sensor repeatability to be +/- 3 ppm when 3 minutes were allowed for the C 0 2 concentration measurement of each airflow. See Appendix C for test results.

30

50% X 45% 40% 35% 30%

A A

*
X
X X X X

x 10% OA Fraction
A

20% OA Fraction

30% OA Fraction o 40% OA Fraction

25%

20% 15% 10% 5% 0% 50

X *x
%

a a 0

"o

x_

: o , /

A 4

T X 4 4

***XXXX

100

150

200250300350400450500550600650 COaRA- C O A

Figure 15: Predicted measurement errors as a function of OA% for JCEM Laboratory (5% error in SA measurement, 1 ppm acquisition error, and repeatability errors of 3 ppm). As illustrated in Figure 15, the accuracy of the concentration balance method is reduced as the difference in the return and outside airflow CO2 concentration levels becomes small. The outside airflow fraction also has a significant effect on the accuracy of this measurement technique, with smaller outside air fractions resulting in higher errors. 3.2.3 Theoretical Recommendations for Airflow Measurement Techniques Direct airflow measurement techniques can provide accurate measurements of airflow rates when installed in accordance with manufacturer's recommendations. Non-uniform flow profiles and turbulence present the largest hurdles in obtaining accurate and reliable measurements. Averaging Pitot-tube arrays and electronic thermal anemometers typically have the same installation guidelines in terms of required unobstructed duct lengths to reduce non-uniform flow profiles and turbulence. These effects may be minimized somewhat by the use of flow straightening devices in the ductwork. When using averaging Pitot-tube arrays, the correct selection of the accompanying differential pressure transmitter is extremely important. The range of this instrumentation should closely match the range of velocity pressures expected within a duct. Generally, the accuracy of electronic thermal anemometry at low flow rates is much greater than that of averaging Pitot-tube array unless high-accuracy, auto-zeroing differential pressure transmitters are used. It should be recalled that such a high-accuracy, auto-zeroing differential pressure transmitter is installed in the outside airflow duct of the JCEM 31

Laboratory, while a more common, 0.50" differential pressure transmitter is used in the supply duct. Specific recommendations for the use of the averaging pitotTube with more commonly used differential pressure transmitters (0.1 inWg and 0.25 InWg ranges) for the direct airflow measurement devices within the JCEM Laboratory are also included in Table 7. Table 7: Minimum airflow rates for 15% accuracy using direct measurement techniques for the JCEM Laboratory only.
Auto-zero, Pitot-Tube fpm 572 93 93 Averaging Array cfm 3 114 667 248 Non Auto-Zero, Averaging Pitot-Tube Array (0 to .1 range) fpm cfm 256 256 256 NA* 1819 682 Non Auto-Zero, Averaging Pitot-Tube Array (0 to .25" range) fpm cfm 390 2 202 390 390 2 876 1 079 Electronic Thermal Anemometry fpm 67 67 67 cfm 365 477 1/9

Supply Economizer Dedicated OA Duct

* The range of 0.1 DP transmitter is too small to handle the design supply air volume of 8,000 cfm for the JCEM lab

Indirect airflow measurement techniques offer an alternative to the direct measurement of the outside airflow rates. However, the measurement of the mixed air temperature is a severe drawback for the use of the temperature balance method. This problem does not appear to have a simple solution in the near future. Additionally, the large errors that result when the return and outside air temperatures become close are unavoidable during many times of the year. The concentration balance technique, however, appears to have potential as an accurate measurement technique when used under the proper conditions. However, It should not be relied upon as the only measurement technique due to large errors when the C 0 2 levels in the return and outside flows become small. In addition, the use of this method with small outside air fractions should be done with caution. 3.3 VAV Control Techniques Error Analyses

Presented in this section are error analyses for the proposed VAV control techniques to be tested for this project. General recommendations for the appropriate use of these techniques based upon the error analyses are included at the end of this section. It should be noted that all control systems require the direct or indirect measurement of the airflow rate and the positioning of dampers to maintain the minimum outside airflow rate. Actuators used to position the damper have an error affiliated with them, and the dampers also have errors associated with their linkages and hysteresis effects. With the use of PID control, however, these errors associated with dampers become precision errors and hence the average of them is zero. The only remaining source of error is that of the outside airflow measurement.

32

3.3.1 Fixed Minimum Outside Air Damper Position In a system where all damper positions are fixed, the outside airflow rate is proportional to the square root of the pressure difference across the outside air damper. In a VAV system, as the supply fan speed is reduced, the pressure in the mixing plenum is also reduced. Due to this reduction in pressure difference across the outside air damper, the outside airflow intake rate is reduced. In the best case scenario, the outside airflow intake rate in a fixed minimum outside air damper system will be a constant percentage of the supply airflow. Several other authors have identified additional drawbacks to this type of control system (Janu et al. 1995, Solberg et al. 1990, Mumma and Wong 1990). Figure 16 shows the outside airflow intake rate vs. the supply airflow rate under ideal conditions. For this analysis, no calibration or acquisition errors were considered. Additionally, stack and wind effects on the system were neglected.
80%

70%

60%

50%

40%

30%

20%

10%

0%., 30%

. 40%

. 50%

,_ 80%

^ 90% 100%

^50% 70% VAV % of Design

Figure 16: Predicted errors for the fixed damper position control technique. From Figure 16, it is apparent that a fixed minimum outside air damper position is not a valid control strategy for a VAV system under any conditions. 3.3.2 Volumetric Fan Tracking The outside airflow intake rate in a volumetric fan tracking system is the difference in the supply and return airflow rates (assuming there is no exhaust flow):

CFMOA = CFMSA - CFMM

Equation 24

33

Whenever the exhaust flow is greater than zero, the outside air intake rate will be increased and an energy penalty may result. This error, however, will always lead to an increase in the outside airflow intake rate and hence will not affect the method's ability to provide minimum outside airflow rates. The error associated with the calculated outside airflow rate is given by Equation 25:
U

CFMoa

= 4uCFMSa + UcFMra

Equation 25

The error analysis for the volumetric fan tracking control strategy is performed using the system parameters for the laboratory at the JCEM shown in Table . Predicted errors for the volumetric fan tracking control method implemented in the JCEM Laboratory are shown in Figure 17.
70%

60%

-B40% OA, Pitot-tube Array - * 20% OA, Pitot-tube Array

50%
40%

-A 20% OA, Electronic Thermal Anemometry -x40% OA, Electronic Thermal Anemometry

I UJ
N 30%.

20%.

10%.
0% 30% 40% 50% 60% 70% 80% 90% 100%

VAV % of Design

Figure 17: Predicted errors of volumetric tracking control technique in the JCEM Laboratory. As illustrated in Figure 17, the use of averaging Pitot-tube arrays in the volumetric fan tracking control method results in exceptionally large errors, especially as the VAV percentage is reduced. Predicted errors are less using electronic thermal anemometry and may be acceptable for large outside air fractions. It should be noted that the predicted errors shown in Figure 17 do not account for any air leakage which may occur between the measurement stations in the return and supply air ducts. Leakage may lead to a false calculation of the outside airflow intake rate and increase the predicted error of this control technique.

34

3.3.3 Plenum Pressure Control The outside air intake rate is related to the pressure drop across a fixed orifice by Equation 14, rewritten here:

V = 1096.7 J - ^ -

ip-C

Equation 14

When the outside air damper is used as the fixed orifice and is left in a fixed position, the value of the loss coefficient, C, remains constant. In Section 32 of the 1997 ASHRAE Handbook of Fundamentals (AHSRAE 1997b), the value for the loss coefficient for any opposed blade damper is equal to 0.52 when the damper is fully open, regardless of the dimensions of the damper. This value for C will be used in the error analysis for this control strategy. The density in Equation 14 can also be replaced by Equation 4, giving Equation 26:

V = 925.7,

AprT

Equation 26

Values for the parameters listed in Equation 26 are shown in Table 2 and the values of the uncertainties used are listed in Table 8. Predicted errors for this control technique are shown in Figure 18 when different ranges of differential pressure transmitters are used. Table 8: Uncertainty values used in the calculation of predicted errors in Figure 18.
"T
"calibration " d a t a acquisition

Up

"Dp

0 50% reading 0.50% reading 1 00% full scale 0.25 F reading 0.25% reading 0.10% full scale

35

15.0% 13.5% 120% 10.5% 9.0%

t 7.5%

ao%
4.5%

ao%
1.5% 0.0%
200 400 600 800 1,000 Velocity [fpm] 1,200 1,400 1,600 1.800

Figure 18: Predicted errors of plenum pressure control in the JCEM Laboratory. As shown in Figure 18, the accuracy of this control method depends significantly upon the selection of the proper differential pressure transmitter to measure the pressure drop across the outside air damper. Since this technique is based on maintaining a constant differential pressure across the damper, the transmitter may be specifically selected for the application. Any desired degree of accuracy may be obtained by sizing the dedicated air damper to obtain a pressure drop close to the maximum range of the transmitter, or vice versa. Values of the predicted error in Figure 18 assume a constant density in Equation 14, and errors associated with this assumption are not included. Under typical building operating conditions, this introduces at most 2% error in the airflow rate (see Appendix B for further details regarding this assumption). 3.3.4 Measurement and Control of Outside Airflow Rate with Economizer Recommendations for the measurement and control of outside airflow in a system with an economizer are the same as those for the averaging Pitot-tube array and the electronic thermal anemometry measurement techniques. Predicted errors for the JCEM Laboratory using a high-accuracy, auto-zeroing differential pressure transmitter in the outside air duct are shown in Table 9. Also, Table 9 provides the errors predicted for a Pitot-Tube array using a more commonly used differential pressure transmitter with a range of 0.10 inWg (without auto-zeroing capability).

36

Table 9: Predicted errors of direct control techniques in the JCEM Laboratory.

System Description 20% OA w/ Economizer . 40% OA w/Economizer 20% OA w/Dedicated Duct 30% OA w/Dedicated Duct Auto-zero, Averaging Pitottube Array 3.1% 2 1% 2.0% 2.0% Pftot-tube Array (0-0.1" range) 19 3% 5 2% 3.4% 2.4% Electronic Thermal Anemometry 4.4% 2 2% 2,0% 2.0%

3.3.5 Measurement and Control of Dedicated Minimum Outside Duct Airflow Rate Similar to the previous system with only an economizer duct for outside airflow, the only source of error is the measurement of the outside airflow intake rate. Any desired degree of accuracy may be obtained by sizing the dedicated air duct to obtain the necessary flow rate. The only requirement is that the face velocity of the airflow at the outside louver should be less than 400 fpm to minimize the possibility of water penetration (ASHRAE 1997b). If space constraints are not an issue, then the dedicated duct can be sized for any desired accuracy. Duct sizes will need to be smaller for systems using averaging Pitot-tube arrays (with the proper selection of the differential pressure transmitter) compared to electronic thermal anemometers in order to accurately measure the minimum outside airflow intake rate. In order to achieve a higher velocity in some systems, a damper may need to be installed in the recirculated duct in order to increase the differential pressure between the outside and the mixed air plenum. Predicted errors for the JCEM Laboratory are shown in Table 9. 3.3.6 Outside Air Duct Injection Fan The only source of error for this control technique is that associated with the measurement of the outside airflow intake rate. Predicted errors for the use an injection fan in a dedicated outside air duct are shown in Table 9. It should be noted that when the fan curve is selected to be very flat, the measurement of outdoor airflow rate may not be required since the fan operates at constant volume over the expected range of pressure drops. 3.3.7 Theoretical Recommendations for VAV Control Techniques The results of the theoretical error analysis conducted in this chapter are briefly summarized in this section. Based on these theoretical results, only the VAV control techniques proven to be sufficiently accurate are further evaluated and tested in the laboratory as described in the following chapters. The control strategy commonly used for CAV systems consisting of a fixed minimum outside air damper position does not work in VAV systems. This strategy will only provide the minimum outside airflow intake rates when the system is operating at design conditions. Volumetric fan tracking also appears to 37

have limited applications. The accuracy of the return and supply airflow rates in this technique must be extremely high to ensure that minimum outside airflow rates are maintained during all times. It also requires that when the system is operating at minimum outdoor air flow rates, the exhaust air damper is fully closed. Maintaining a constant pressure drop across a wide open damper in a dedicated outside air duct may produce acceptable control of minimum outside airflow rates. The reliability of the dedicated outside air duct technique depends therefore on the accuracy with which the small pressure drop across an open damper can be measured. The proper selection of the differential pressure transmitter is critical to the success of this control strategy. The major benefit of this control strategy is the absence of a direct outside air intake rate measurement, making this strategy favorable in existing buildings where proper installation of direct airflow measurement devices is not possible. The use of direct measurement and control of a fixed minimum outside airflow intake rate can provide acceptable levels of reliability. The accuracy of these methods is dependent only upon the accuracy with which the outside airflow rate can be measured. In economizer systems, this may prove to be the limiting factor when extremely low flow rates result from the large duct areas. The use of a dedicated outside air duct both with and without an injection fan appear to account for this limitation by increasing the outside airflow intakes rates to levels that can be measured with greater accuracy. However, the use of these methods may be expensive and difficult to implement in existing buildings without dedicated outside air ducts.

38

LABORATORY DESCRIPTION

This section provides information specific to the JCEM Laboratory and to the preparatory work that was conducted prior to testing. Also included are descriptions of the laboratory conditions simulated during the tests. 4.1 Laboratory Air System This section briefly describes the design and operation of the air systems of the JCEM Laboratory. A complete description of the laboratory is presented in Appendix A. 4.1.1 System Function The air system provides air conditioning and ventilation to the following zones: Full Size Zone East Full Size Zone West Zone Simulator 1 Zone Simulator 2 4.1.2 System Description The system consists of two air handlers, four VAV boxes and a return fan as illustrated in Figure 19. The central air system component is a single zone, draw-thru, built-up air handling unit. This air handling unit is comprised of, in order, an outside air economizer, a filter bank, a chilled water coil, a hot water coil, and a variable speed drive supply fan. The main air handling unit supplies medium pressure conditioned air to the parallel Fan-Powered Mixing Boxes (FPMB) serving the zones. A second air-handling unit located up stream of the main air handler provides control of ventilation air conditions supplied to the main air handling unit. This second unit is referred to as the Outside Air Conditioning Station (OACS). The system also includes a variable speed drive return fan.

39

Figure 19: JCEM Laboratory Isometric. 4.1.3 Laboratory Modifications A complete description of the JCEM Laboratory is included in Appendix A. Some changes were made to the laboratory specifically for this project. These changes include: (i) In the outside air ductwork, a new ultra-low-leakage damper was installed just upstream of the main air handling unit. The existing outside air damper on the air handling unit was de-coupled from the recirculated damper and locked in the full open position. This new damper is approximately 70% smaller than the original damper, allowing for much better control of the outside airflow intake rates. (ii) For systems requiring a dedicated outside air duct, an existing section of the outside air duct upstream of the new outside air damper and main air handling unit was removed and replaced with a new section of ductwork. The cross-sectional size of the existing ductwork is 7.11 ft2, and the new ductwork was only 2.67 ft2, a reduction in size of over 60%. A schematic diagram of the new ductwork is shown in Figure 20. 40

32/32

16124

Air Flow

Figure 20: Schematic diagram of new outside air ductwork. 4.2 Supply Duct Averaging Pitot-tube Array Calibration To calibrate the averaging Pitot-tube array measurement stations in the supply air ducts, Pitot-static tube traverses were performed for several flow rates. Calibrations were also performed for the dedicated outside air duct when a conventional (rather than auto-zeroing) differential pressure transmitter was used. In the supply duct, each traverse was performed with a Pitot-static tube manufactured in accordance with ASHRAE Standard 111-1988 (ASHRAE 1988) and a NIST traceable mieromanometer with an accuracy of 0.00025 inW.G. Measurements were made at 25 separate points for each traverse following the Log-Tchebycheff rule for rectangular ducts (ASHRAE 1988). Each individual point was measured three times and then averaged. The atmospheric pressure in the laboratory and air temperatures in the ducts were also recorded to account for air density in the calibrations. ASHRAE Standard 111-1988 provides recommendations for suitable velocity profiles when performing a Pitot-tube traverse of a duct (ASHRAE 1988). All traverses performed in the JCEM Laboratory had "ideal distributions" where more than 90% of the measurements were greater than 10% of the maximum reading. Figure 21 shows the results of the calibration of the supply duct averaging Pitot-tube array in the JCEM Laboratory.

41

0.50

0.75

1.00

1.25

1.50

1.75

(Vp/denslty)'"

Figure 21: Calibration of supply duct Pitot-tube averaging array with 95% confidence intervals. Confidence intervals were calculated from Equations 4.2.1 through 4.2.3, which provide the uncertainty in the response variable (y) when a particular value of the predictor variable (x) for a simple Ijnear regression is considered (Montgomery and Runger 1994). These 95% confidence intervals were used as the measurement uncertainty, Upnot, in place of the manufacturer's declared accuracy of 2% of reading in the reported data and analyses in Sections 5 and 6.

1
u y --ta/l^-lV0") \ J
\

, (*0-*y
~

\l\

Equation 27

;2_f(y,-^)2
> -2-,
Sxx
(

Equation 28

=n

E*,

2>
J

(' n

Equation 29

M=l

42

where: ta/2_2
=

value of the t-distribution with n-2 degrees of freedom at a confidence level of (1-a)

oy S^

= standard error regression = sum of squares for x

4.3

Laboratory Test Conditions

To provide a fair basis for a qualitative analysis of the various airflow measurement and VAV control techniques, several aspects of the testing are kept the same for all tests. Descriptions of these characteristics are included in this section. 4.3.1 Laboratory Control Control of the laboratory is accomplished through a DDC control and data acquisition system. All relevant measurements needed in the laboratory are recorded every 10 seconds. For the purposes of this test, only the two zone simulators in the laboratory will be used. Refer to Figure 19 and Appendix A for further information regarding the zone simulators in the JCEM Laboratory. Table 10 lists the key characteristics that were used during the testing procedure including static pressure set points, temperature set points, and duct areas. To investigate the effects of various outside air fractions, tests will be performed using 20%, 30%, and 40% outside airflow fractions. The 20% and 40% outside airflow fractions will be used for systems with the economizer duct. Due to the range on the differential pressure transmitter being used as the reference airflow measurement, tests using 40% outside airflow fraction were not possible in systems with the dedicated outside air ductwork. Instead, 20% and 30% outside airflow fractions were used. Each test was four hours long, and simulated a full day of air system operation. A reduced time scale was used due to the absence of mass in the zone simulators and the accompanying timedelayed effects of mass within a building. Table 10: Key physical and control characteristics of the JCEM Laboratory.
Design Supply Air Flow Rate 8,000 cfm Supply Air Temperature 55 F Supply Duct Static Pressure 1 85 inW.G Return Duct Static Pressure -LOOinW.G. Outside Duct Static Pressure 0 00 inW G Supply Duct Area 5.44 sq. ft. 6 00sq ft. Return Duct Area Economizer Duct Area 7 11 sq.ft. 2 6 / sq. ft. Dedicated Outside Duct Area

43

4.3.2 VAV Load Profiles To investigate the effectiveness of the various airflow measurement and VAV control methods, VAV flows were varied throughout the tests. Figure 22 shows the VAV flow profiles used for the tests. For all tests using a 20% outside air fraction, the VAV flow varied from 30% to 100% of the design flow (VAV Profile 1). For the tests with either a 30% or 40% outside airflow fraction, the VAV flow varied from 40% to 100% of the design flow (VAV Profile 2). Control of the flow was achieved by imposing loads within the zone simulators using electric resistance coils. The VAV boxes in the zone simulators were programmed to maintain a constant zone outlet temperature of 70 F.
110% 100%. 90%. 80%. C 70%. ra

^***"
s?S

*'

s^s''

Jj 60%.
o X 50%.

^^ .' ^*~ ."

VAV Profile 2 . . . V A V Profile 1

>
$ 40% 30% 20%. 10%

a 0% 00

0:30

1:00

1:30

200 Time [hr]

230

3:00

3:30

4:00

Figure 22: VAV profiles used for testing. 4.3.3 Outside Air Temperature Profile To investigate the effect of varying outside air temperatures on the measurement and control of outside air intake rates, a typical cooling day outside temperature profile was used. Figure 23 shows the temperature profile used during the testing.

44

90 1

85 EL

^ ^ _ _ _ _ _

^
^ S ^
jS^

.? 80
Q.

J 75

%TA
< o
65.^1

s y^
.

60 J
0:00

,
0:30

,
1:00

,
1:30

,
2:00 Ttme [hr]

,
2:30

,
3:00

,
3:30 4:00

Figure 23: Outside air temperature profile used for testing. The temperature profile shown in Figure 23 was sufficient to test the effects of varying temperatures on the measurement and control of outside air intake rates. Due to some concerns regarding the accuracy of electronic thermal aoemometry at low temperatures, a test was run to compare these readings with those from an averaging Pitot-tube array with a high-accuracy, auto-zeroing differential pressure transmitter. The outside airflow was controlled to a flow rate of 750 fpm as measured by the averaging Pitot-tube array. The outside air temperature was then varied from approximately 35F to 65F. Figure 24 shows the percentage difference between the electronic thermal anemometer and the averaging Pitot-tube array. As illustrated in Figure 24, the maximum difference is within approximately 5% throughout the temperature range investigated with higher readings from the electronic thermal anemometer at lower temperatures; well within the predicted error ranges of the two measurement techniques. Difference between these findings and those of Drees et al. (1992) are due to the use of newer electronic thermal anemometry sensors.

45

Figure 24: Comparison of electronic thermal anemometer and averaging Pitottube array airflow measurements as a function of air temperature. 4.3.4 Summary of Systems to be Tested Table 11 summarizes the systems to be tested for this project in the laboratory. Systems marked with an "X" in the table were not tested in the laboratory. Each system can be represented by a case number, i.e. 3-A for the direct control in a system with an economizer duct for a 20% outside air fraction, using the airflow measurements from the averaging Pitot-tube array for control. These case numbers are used to simplify the analysis of systems in Sections 5 and 6 of this report. For all the tests, the outdoor airflow rates are measured using averaging Pitot-tube array with auto-zeroing differential pressure transmitter having a range of 0 to 0.05 inWg. The measurement values from this technique are used as "reference" against which measurements from other airflow rate measurement techniques are compared. For systems with dedicated duct or injection fans, additional tests using outdoor airflow measurements from averaging Pitot-tube array with a conventional (in lieu of auto-zeroing) differential pressure transmitter -with a range of 0 to 0.1 inWg- are also performed.

46

Table 11: Systems tested in the JCEM Laboratory.

System Description Fixed Minimum Damper Position Plenum Pressure Control Direct Control with Economizer Duct Volume Tracking Direct Control with Dedicated Duct Injection Fan Measurement Technique used for Control -NA-NAAveraging Pitot-tube Array Electronic Thermal Anemometry Electronic Thermal Anemometry Averaging Pitot-tube Array Electronic Thermal Anemometry Averaging Pitot-tube Array Electronic Thermal Anemometry Outside Air Fraction 20% 30% 40%

A B

X X X X

X X X X X C

47

ANALYSIS OF LABORATORY TEST RESULTS

This section presents the results of the various systems tested as described in Section 4.3.4. Also included is an analysis of the tested measurement and control techniques. 5.1 Laboratory Test Results Experimental results for the each tested control and airflow measurement technique are shown in this section. During testing in the laboratory, direct outside airflow measurements were recorded for both the averaging Pitot-tube array with auto-zeroing differential pressure transmitter and the electronic thermal anemometer for all tests. In tests where a system was controlled by a specific airflow measurement technique, the other method was used only to monitor the outside airflow rate for the same test. Section 4.3 provides a complete description of the test methods used in the laboratory. Illustrated in Figure 25 are temperature values during a representative laboratory test. Appendix E contains a complete analysis of the temperature values for each system tested.
90 85 80 _ Outside Air Temp. Set Point Simulatued Outside Air Temp. Zone 1 Outlet Temp. Zone 2 Outlet Temp. Supply Air Temperature

55

teo*<>tlo*o*Mpo">>"!K4&jiwo*>>4>*<^^

50 0:00

0:30

1:00

1:30

200 Time [hr]

230

3:00

3:30

4:00

Figure 25: Representative temperature values during laboratory testing. Included in Appendix D are graphical representations of the supply and outside airflow rates, the outside air percentage of design, and comparisons of the direct airflow measurement techniques used for each test. 48

5.1.1 Measurement Technique Analysis Figure 26 shows a representative comparison of various airflow measurement techniques; averaging Pitot-tube array with auto-zeroing differential pressure transmitter, electronic thermal anemometry, temperature (e.g. enthalpy) balance and C 0 2 concentration balance. Each time step corresponds to a specific set of outdoor air temperature and supply airflow rate as described in 4.3. As illustrated in Figure 26, agreement between the two direct airflow measurement techniques was very good. More analysis of the averaging Pitot-tube array with auto-zeroing differential pressure transmitter is included later in this section.

i
i

0:00

0:18

0:36

0:54

1:12

1:30

1:48

206

224

242

3:00

Time [hr]

Figure 26: Comparison of various airflow measurement techniques. Also illustrated in Figure 26 are the two indirect airflow measurement techniques; the concentration balance and temperature balance methods. Values measured by the C 0 2 concentration balance method are close to the two direct measurement techniques. More information regarding the concentration balance technique is included in section 5.2.6. Finally, values calculated using the temperature balance method are not in agreement with those measured with any other technique. No further experimental tests were therefore performed using the temperature balance airflow measurement technique. The absolute measurements of outdoor airflow rates for the tested systems are summarized in Table 12. Note that the results for the averaging Pitot-tube array with auto-zeroing differential pressure transmitter are provided for all the systems. As mentioned earlier, these results are considered as "reference" values. The measurements obtained with the thermal anemometry device are provided for all tests except for the direct control that uses C0 2 concentration balance technique. Finally, the measurements using the averaging 49

Pitot-tube with a conventional (e.g. with no auto-zeroing capability) differential pressure transmitter are performed and listed for the dedicated duct and injection fan systems (e.g. tests 6-A, 6-B, 8-A, 8-B, 9-A, and 9-B) as the values obtained for non auto-zeroing averaging Pitot-tube array. From the results summarized in Table 12, it is clear that the outdoor airflow rates measured using an averaging Pitot-tube with a conventional differential pressure transmitter (0-0.1" range) agree very well with those obtained using the averaging Pitot-tube array with high accuracy auto-zeroing differential pressure transmitter. Table 1.2: Comparison of Outside Airflow Measurements
Non Auto-Zero, Averaging Auto-zero, (orCOj Averaging Pitot- Pitot-tube Array tube Array balance) Measurement Control 1 NA NA NA -NANS P P E E E P P E E P P E E C
C - Clo^4mrt(#> T I M

Electronic Thermal Anemometry

System Description

Case

Set-point mean (dm) 16it> 2 400 3 2C0 1600 1?00 1600 3 ..30 1600 3 200 1600 16C3 7 1G0 1bC0 2400 1600 2 400 1600 2 400 1600 ' r ^* r*r^ 656

(cfm)

mean (cfm) NA NA NA -NANANANA-NA-NA-NA1636 2 424 Nv\ -NA1601 2 401 1656 2404 1,605'

stdev (cfm) NA NA NA -NA-NA-NA-NA-NA-NA-NA56 44 -NA-NA44 56 54 46 119

3

RMS (cfm) NA NA NA -NA-NA-NA-NA-NA-NA-NA66 49 -NA-NA44 56 78 46 119 3

mean (cfm) GB? 1407 2 174 1544 3 279 1546 3 225 1637 3 263 2 436 1593 2403 1640 2428 1535 2 393 1600 2 400 -NA

stdev (cfm) 564

RMS (cfm) 1048 1 199 ' i l l 01 118 >1 '3 bl 1 91 1 C 37 58 50 71 36 31 25 -NA-

Fixed Damper Position Fixed Damper Position Fixed Damper Position Plenum Pressure Control Plenum Pressure Control Direct Control with Economizer Durt Direct Control with Economizer DJ-t Direct Control with Economizer Duct Direct Control with Economizer Duct Volume Tracking Drect Control with Dedicated Duct Direct Control with Dedicated Duct Direct Control with Dedicated Duct Direct Control with Dedicated Duct Injection Fan Injection Fan Injection Fan Injection Fan Direct Control
, D^tnt iiha An
2 3

iA i n 11 2-A zC 3A

141" 2178 1630 3<tt8 1631 3 I V 1695 3 228 2 427 1GJS 2 410 1643 2404 1621 2429 1622 2418 1632

Feo
8(0 70 88 46 47 55 50 458 62 37 43 42 28 35 31 25 -NA-

!>.
4-A 4-C 5^ 6 A 6B 7A 7-B B-A 8-B 9-A 9B -NA-*
n

Concentration Balance

A different system setup was used for testing the concentration balance measurement technique, s e e Section 5.2.6 Value is for C 0 2 concentration balance measurement technique, not electronic thermal anemometry

A comparison of the two direct airflow measurement techniques for the systems tested in the laboratory is shown in Table 13. Values listed in Table 13 are based upon the absolute volumetric difference of the reference measurement (e.g. the averaging Pitot-tube array with auto-zeroing differential pressure transmitter) and the other measurement techniques (e.g. electronic thermal 50

anemometry, averaging Pitot-tube with a differential pressure transmitter (0-0.1" range), or CO2 concentration balance):

preference Pilot tube array value)[control measurement value j.

Table 13: Absolute difference between control airflow measurements and reference values.
System Description Case Set-point (cfm) 1 >^3 7 400 J?. 0 1G00 J 200 1,600 3.200 1.600 3 200 1C03 r,"" 7 400 1 roi 7 100 1 M0 2 400 * fif* 7 4C3 1 n't mean (cfm) 1T 71 G7 8!i 2b -NA-NA58 3f 2G 14 17 '2 2b 19 stdev (cfm) 12o In 57 71 20 -NA-NA24 16 18 10 12 10 16 13 14 73 1/ C) mean% error 1 b 0To 10'. 2 66' 57 0 8% -NA-NA 3 4% 1 1% 10% 0 9% 0 7% 0 7% 1 1% 12% 0 7% 2 8% 0 9% 7 7% max% error 1 10// !>b . 17 1 ,'10 3"o 3 2% -NA-NA 8 7% 2 8% 3 8% 3 5% 2 2% 2 8% 3 2% 4 5% 3 8% 7 6% 3 7% 22 2 %

F-jtCM Lumper Pcsitic 1 riK-d niinrx-r Position Fixud [Jjinrxr Position Piunum P r e . .uro Control Plenum Pressure control Direct Control with Economizer Duct Direct Control with Economizer Duct Direct Control with Economizer Duct Direct Control with Economfrer Duct Volume Trac'-iP] [ i r e t C-jn'rol with r.rd calc-i Duct Dirrct Control with DeJ catcd Duct D i i i - j f o n l r 1 with [JP. calnd D i r t Diirct Control with Dedicated Du-a
1 IJLCI nn Tan

1A 1 B 1 0 2A

2-C
3-A 3-C 4-A 4C 5A B-A 6B 'A 7B IM 8B
A

Injection Fan Inir-'ini Fan Iniccticn Fan bi Let Co itic w' C O - Bdl.11 ce"
1 2

in
41 72 170

9B NA

Error is percentage of reading using the Averaging Pitot-tube Array as the reference A different system setup was used for testing the concentration balance measurement technique, see Section 5.2.6. Values shown here are concentration balance measurement value

As shown in Table 13, the accuracy of the two direct airflow measurement techniques (relative to the reference measurement technique which is the averaging Pitot-tube array with auto-zeroing differential pressure transmitter) is extremely good for all of the tests completed. It should be mentioned that this result may be partially due to the fact that the airflow profiles within the JCEM Laboratory are very uniform and that the measurement devices were installed according to manufacturer's recommendations. 5.1.2 VAV System Control Results Table 14 summarizes the results of the outside airflow intake rates for the tested systems. Specifically, Table 14 provides the average value, the standard deviation, the root mean square of the outdoor air intake flow rate, and the validity of each measurement and control method.

51

Table 14: Error Analysis for all the systems tested.

Results of Outside Air Intake Rate Measurements

System Description

Measurement Control 1 -NA -NA-NA -NA-NA P P E I E P P E E P P E

Case
1-A 1-B 1-C 2-A 2-C 3-A SI. 1A JC 5-A 6-A 6-B 7-A 7-B B-A 8-B 9-A 9-B -NA- 2

Set-point (cfm) leoo 7 4C0 3 200 1000 J70C 1.600 3iTJC

Validity

mean (cfm) 002 1407 7 174 1544 3 279 1.635 3 15? 1637 3Z.--3 2 43G 1,635 2.424 1,640 2 428 1601 2.401 1,600 2400 1,632

stdev (cfm) FG4 680 B'O 70 BB 38 0 5 SO 4*8 56 4 43 42 44 56 31 25 137

RMS (cfm) ".OIB 1 199 1M3 9 118 52 5 07 20 951 b6 59 58 50 44 56 31 25 141

Fixed Damper Position Fixed Damper Position Fixed Damper Position Plonum Pressure Control Plenum Pressure Control Direct Control with Economizer Duct [ rcctControl with Tern.}Tii'ir Cur* Dirurt Co ltrol with Ecoromizcr Duct Direct U -ltrol * th E r i m i / e r Cu 1
VoliiTL TISCK. g

KV 7V . 2u' 100 , 100r 100% Wr"j 91 c 10" 0 100% 100% 100% 100% 100% 100% 100% 100% 75%"

icno
J2CC

1600 1600 2400 1,600 2,400 1600 2,400 1600 2400 1,600

U n x t Contiul wilii DuliLdtud D u d Direct Control with Dedicated Duct Direct Control with Dedicated Duct Direct Control with Dedicated Duct
Inlaw* *in C i n

Injection Fan Injection Fan Injection Fan Direct Control

1 2

E C

P = Averaging Pitot-tube Array, E - Electronic Thermal Anemometer, C = C O 2 Concentration Balance A different system setup was used for testing the concentration balance measurement technique, see Section 5.2.6

The percentages listed in Table 14 and specifically in the column labeled "validity" were calculated from Equation 30: validity = n
where: nv = the number of valid data points

Equation 30

in Table 14, the value of the root mean square (RMS) was found from Equation 31:

RMS = . . - ( * , - Set Point)2

Equation 31

52

where: Xi

= measured value for a specific time i (a 10 second time step was considered for the duration of each test) Set Point = target outside air intake value n = total number of data points over the duration of the test

Each test presented in Table 14 is subject to errors from the airflow measurement and the control technique used. Each 10 second data point, Xj, recorded during testing was considered valid if it met the following two conditions: 1) \x{ - set point\ < (set point 10%) and

2)-^-<15%
where: e, = the predicted error for the airflow measurement in the JCEM Laboratory

The first condition attempts to account for the accuracy of the control technique by requiring the data point to be within 10% of the set point. The second condition attempts to account for the accuracy of the airflow measurement technique by requiring the predicted error of the data point to be less than 15%.

53

5.2 Data Analysis Appendix D contains graphical results for all the systems tested and listed in Table 12, Table 13, and Table 14 except for the direct control using the concentration balance airflow measurement technique. The results for that test are presented and analyzed in Section 5.2.6. 5.2.1 Fixed Minimum Position Outside Air Damper As indicated in Table 14, the base case control methods did not maintain the required minimum outside air intake rates during times of reduced flow in the VAV system. Outside air intake rates were much closer to a constant percentage of the supply airflow than to a constant flow rate as illustrated in Figure 27. As was predicted in Section 3.3.1, a fixed minimum outside air damper position in VAV systems does not maintain minimum outside air intake rates and is therefore not a recommended control technique.

Figure 27: Outside airflow % of deisgn for fixed minimum outside air damper position test. 5.2.2 Direct Control with Economizer Duct As shown in Table 14, this control technique was able to maintain minimum outside airflow rates during all tests completed at the JCEM Laboratory. These results were independent of the direct outside airflow measurement technique used. However, these results are dependent upon the fact that the averaging Pitot-tube array in the JCEM Laboratory was used with a properly selected differential pressure transmitter and that airflow profiles in the laboratory were very uniform. Without the use of the proper differential pressure transmitter, the airflow rates in the economizer duct would have been too low to measure accurately with the averaging Pitot-tube array. Figure 11 illustrates the effect the range of the differential pressure transmitter has on the error of the 54

measured airflow rates. The electronic thermal anemometer requires no special additional apparatus to measure the low airflow rates typically found in economizer systems during minimum outside air intake mode. In systems where direct airflow measurement devices can be installed according to manufacturer's recommendations, direct control of the outside airflow rate using electronic thermal anernometry or an averaging Pitot-tube array with a high-accuracy differential pressure transmitter is an adequate control technique. 5.2.3 Direct Control with Dedicated Duct with and without an Injection Fan As indicated in Table 14, this control technique was able to maintain minimum outside airflow rates during all tests completed at the JCEM Laboratory. These results were independent of the direct outside airflow measurement technique used and of whether or not an injection fan was used. It is important that the duct be sized such that outside airflow rates will be high enough to measure accurately using an averaging Pitot-tube array or electronic thermal anemometer. Additionally, the proper selection of the differential pressure transmitter is essential for reliable airflow measurements using an averaging Pitot-tube array. The transmitter's range should be selected to match the expected airflow velocities in the dedicated duct. 5.2.4 Volumetric Fan Tracking The results for the volumetric fan tracking control technique in Table 14 indicate that this technique does not provide adequate control of outside air intake rates in VAV systems under typical building operating conditions. This inadequacy is due mainly to the following: Damper Positioning Limitations - In the tests conducted at the JCEM Laboratory, no combination of fixed damper positions for the outside, return and exhaust air dampers allowed for the minimum outside air intake rates to be met under all VAV percentages tested. It is expected that these limitations would be common to most building HVAC systems. Neglecting Exhaust Airflow Rates - As mentioned in the error analysis included in Section 3.3.2, when the exhaust flow in the system is not zero, the actual outside air intake rate will be increased.

For these reasons, and from the test results, the use of volumetric fan tracking is not recommended as a method to maintain minimum outside air intake rates for VAV systems.

55

5.2.5 Plenum Pressure Control As shown in Table 14, the plenum pressure control technique was able to maintain the minimum outside air intake rates 100% of the time during testing. Essential to the success of this control technique is the proper selection of the differential pressure transmitter used to measure the pressure drop across the outside air damper. See Section 3.3.3 for additional details regarding the selection of the differential pressure transmitter. Also important for the proper use of this control technique is to set the pressure drop large enough to be measured accurately, but no so large as to impose an excessive energy penalty upon the system. A potential drawback to the use of this control technique is the careful commissioning which must take place for proper control. For each desired outside airflow rate, an accurate measurement of both the pressure drop and the actual airflow rate must be performed. A Pitot-static tube or electronic thermal anemometry traverse could be used to measure the outside airflow rates during commissioning. Plenum pressure control works best in systems where the installation of direct airflow measurement devices is not possible and only one or two minimum outside air intake rates are required for typical building operation. 5.2.6 Direct Control using Concentration Balance Measurement Technique Due to the difficulty in maintaining a large difference in the return and outside air COz concentrations at high flow rates, a different system setup was used for this system than that described in Section 4.3. For this test, the supply airflow rate was varied from 2,000 cfm to 4,000 cfm. The outside airflow rate was then controlled to 1,600 cfm, or 40% of the design airflow rate for this test. Actual measured flow rates from the test are illustrated in Figure 28. Since air temperatures were not found to be a factor in previous tests, the outside air temperature was controlled to a constant 65F for the duration of the test. CO2 was injected into the return airflow at a rate to maintain a difference between the return and outside concentrations of approximately 200 ppm. The predicted error for the concentration balance airflow measurement technique was below 15% during the entire test.

56

4,500 4,000 ,0 3.500 3.000 2,500

DC S o

a Supply Airflow Outside Airflow A Outside Airflow Setpoint

2,000 rnm"pP 1,500 1,000 500

$*

-** **",***. ***fl

2

*****
fi

A * * ^ * * * * * * * * * * * 6 * * * 0 * *gfvt f t * * f l * * * * * * * * * * * * M f l H H A * * ' * * M | | f l f t * ow* o* ** ***o^w M ftAoqpOg* 'S6SSSWWOB*

0:00

0.18

0:36

0:54

1:12

1:30 Time [hr]

1:48

206

224

242

3:00

Figure 28: Supply and outside airflows for the concentration balance control test. In order to measure the various C0 2 concentrations with the required repeatability (see Appendix C), each airflow was sampled for a duration of 3 minutes. Because the outside air CO2 concentration changes very slowly with respect to the return and supply C0 2 concentrations, it was not sampled every cycle. During the test, concentrations were sampled in the following ordersupply, return, supply, return, and outside. Figure 29 illustrates the measured outside air percentage of design versus the supply airflow percentage of design. All measurements for the actual outside airflow rate were made with the averaging Pitot-tube array with auto-zeroing differential pressure transmitter.

57

Figure 29: Outside airflow % of design with predicted error bars. As indicated in Table 14, the validity of the test was about 75%. While the ability of the concentration balance airflow measurement technique to control the minimum outside air intake rates was not as good as other direct measurement methods, it does show promise as an adequate control strategy. Figure 30 illustrates the percent error between the measured reference outside airflow rate (using the averaging Pitot-tube array) and the one calculated from the concentration balance. During this test, the CO2 concentration balance airflow rate was calculated every three minutes, while the averaging Pitot-tube array value was measured every 10 seconds. This time difference is responsible for the noisy variation of the data shown in Figure 30.

58

25%
20% 15% 10%

*1A

,'

**
o

?
Jt UJ 0% -5% -10% -15% -20% -25% a00
I ' 1

**

'

0:18

0:36

0:54

1:12

1:30 Time [hr]

1:48

206

254

242

3:00

Figure 30: Comparison of airflow measurement techniques for concentration balance control test. The results presented in Table 13 show that the average absolute error between the two methods is approximately 8%. Over the duration of the test, however, the average of the two measurement techniques differed by less than 2%. This implies that with refined control loop tuning (i.e. slowing down the system update time to match that of the calculated outside airflow rates), using the concentration balance airflow measurement technique to control the minimum outside air intake rate could be an adequate control technique. Additionally, it is expected that the difficulty in maintaining a stable CO2 concentration in the return air in the laboratory made accurate control harder to achieve. In a typical office building, the return air CO2 concentration would be more stable, allowing a more accurate airflow measurement to be made. However, this requirement for relatively stable CO2 concentrations limits the applicability of the concentration balance technique. In spaces where large, abrupt changes in occupancy (and hence C0 2 levels) can occur, this method may prove unreliable. This fact may rule out the use of this control strategy in spaces such as conference rooms and auditoriums, or any building where large transient effects are possible. Typical office space should present a suitable application of the control technique using CO2 balance. As shown in the error analysis of this technique in Section 3.2.2.2 and as indicated by the laboratory results included in Appendix C, the accuracy of the concentration balance technique is reduced as the difference in the return and outside air CO2 concentrations becomes small and as the outside air becomes a smaller fraction of the supply airflow. For these reasons, direct control using the concentration balance technique to measure the outside airflow should not be the

59

only method used to maintain outside air intake rates in VAV systems. During periods when the predicted errors indicate that the technique will be unable to accurately maintain minimum outside airflow rates, another control technique should be used.

60

SUMMARY AND CONCLUSIONS

Included in this chapter are the final recommendations based upon the error analyses and laboratory test results for each measurement and control technique. A brief description of possible areas for future research is also presented. 6.1 Minimum Outside Recommendations Air Intake Control and Measurement

Control of minimum outside air intake rates is critical to meet the requirements set by the ASHRAE Standard 62-1989 and building codes for maintaining adequate indoor air quality within conditioned spaces. VAV systems present additional complications compared to CAV systems. Pressure in the mixed air plenum can fluctuate with changing supply air volumes, making commonly used control strategies for CAV systems inadequate for use in VAV systems. Recommendations listed here are based upon findings from both the error analysis and laboratory test results presented in this report. In particular, findings of both the theoretical and experimental analysis indicate that the use of a fixed minimum outside air damper position and volumetric fan tracking are inadequate control strategies to maintain minimum outside air intake rates in VAV systems. These strategies are unable to provide the required outside airflow under all operating conditions. The best control techniques are those based on direct measurement of outside airflow rates. The use of direct airflow measurement devices such as averaging Pitot-tube arrays and electronic thermal anemometry, however, is often limited by physical constraints. An adequate length of unobstructed ductwork is required for uniform flow profiles. Minimum lengths are usually specified by the manufacturer's of the measurement devices. The expected outside airflow velocity in systems having the required lengths of unobstructed ductwork must be considered to select the proper direct airflow measurement technique. In systems sized for use with economizer cycles, outside airflow rates for minimum outside air intake mode are typically too small to be measured with an averaging Pitot-tube array with properly selected differential pressure transmitter (e.g., if the transmitter is selected to have a range that is low enough to measure accurately minimum outside airflow rates, it will not be able to be used for

61

economizer rates of 100 percent outside air intake). Electronic thermal anemometry may be a good alternative in this case. Another alternative is the use of a dedicated duct to provide outside air when only minimum outside air intake is needed. The dedicated duct can be sized such that airflow velocities will be high enough to measure accurately with either an averaging Pitot-tube array or electronic thermal anemometry. An alternative control technique is to use a plenum pressure control strategy. Here, the pressure drop across a fixed orifice such as the outside air damper and louver is measured and maintained at a constant, predetermined value. Generally, this requires that a dedicated minimum outdoor air damper be used in order to create a reliable fixed orifice. For systems with economizers, using a minimum damper position to create the fixed orifice is usually not accurate due to lack of repeatability of the damper assembly (damper, actuator, and linkage).

Finally, another control alternative is to use the C0 2 concentration balance technique to indirectly measure the outside air intake rate and use this value for direct control of the system. Due to the current CO2 sensor limitations, this technique only works accurately when a single sensor is used to measure outdoor air, return air, and mixed air CO2 concentrations; and it will not provide reliable and accurate control when rapidly changing CO2 concentrations occur in the return air. Additionally, this technique should not be used exclusively to control the outside air intake rate. When the CO2 concentration difference between the return air and the outside air is low, or when the outside air is a small percentage of the supply air, large errors may result by using this control technique. For these conditions when the predicted error is large, another control technique should be used. 6.2 Future Work This report has attempted to provide a thorough theoretical and experimental analysis of current and emerging measurement and control techniques for maintaining minimum outside air intake rates in VAV systems. The culmination of this effort is a set of guidelines that can be used by both designers and building operators to assist in providing adequate indoor air quality for conditioned spaces. One limitation of the experimental work presented in this report is the lack of an "absolute" reference of the outside airflow rate during laboratory testing. To minimize possible errors, an averaging Pitot-tube array with a high-accuracy, auto-zeroing differential pressure transmitter has been used as the reference airflow measurement technique. Exceptionally uniform flow profiles in the laboratory as well as the direct measurement of all parameters necessary for the calculation of the airflow rate (see Equation 5) help to further minimize errors. Additionally, the overall agreement between this measurement technique and 62

others testing in the laboratory reduces the possibility that the selected reference airflow measurement method could be inaccurate. One possible solution to this problem would be to use Laser Doppler Anemometry (LDA) as described briefly in Section 0 of this report. Further research would also be helpful in clarifying specific guidelines for the proper use of direct airflow measurement methods such as the averaging Pitot-tube array and electronic thermal anemometry which are limited by the requirement of sufficient unobstructed duct lengths both upstream and downstream of the measurement device to allow for uniform airflow profiles to develop. Very little data was found during the literature review related to the effects of perforated plates and honeycomb airflow straighteners to reduce turbulence and create more uniform airflow profiles in ducts. Such items could reduce the required length of unobstructed duct required for the proper installation of direct airflow measurement devices. Quantitative data regarding the effectiveness of these devices as well as the related energy penalty due to increased pressure drops would be quite valuable. Finally, as with any laboratory research project, field verification of experimental results is important. Specifically, a more thorough understanding of the specific conditions under which the CO2 concentration balance measurement technique is a reliable method for determining outside air intake rates is necessary. Additionally, field verification of control using the CO2 balance method and using plenum pressure control in packaged roof-top units where a duct for the outside air intake does not exist would provide further validation for these control techniques.

63

REFERENCES

Air Monitor Corporation, Product Specifications, P.O. Box 6358 Santa Rosa, CA 95406. ANSl/ASME MFC-2M-1983, "Measurement Uncertainty for Fluid Flow in Closed Conduits," American Society of Mechanical Engineers, New York. ASHRAE, 1988. ANSI/ASHRAE 111-1988, "Practices for Measurement, Testing, Adjusting, and Balancing of Building Heating, Ventilation, Air-Conditioning, and Refrigeration Systems." Atlanta: American Society of Heating Refrigerating, and Air-Conditioning Engineers, Inc. ASHRAE, 1989, "Ventilation for Acceptable Indoor Air Quality," Standard 621989, June (ISSN 1041-2336) ASHRAE 1996, "Ventilation for Acceptable Indoor Air Quality," ASHRAE Standard 62-1989R, Public Review Draft, August. ASHRAE, 1997a, Handbook of Fundamentals, Section 14, 1997. ASHRAE, 1997b, Handbook of Fundamentals, Section 32,1997. ASHRAE, 1997c, Handbook of Fundamentals, Section 6,1997. ASME, 1989, Measurement of Fluid Flow in Pipes Using Orifice, Nozzle, and Venturi. Standard MFC-2M-85. Avery, G., 1989, "Updating the VAV Outside Air Economizer Controls," ASHRAE Journal, Vol.31, No.4, pp. 14-16. Avery, G., 1992, T h e Instability of VAV Systems," Heating, Piping, and Air Conditioning, Vol. 64, No. 2, pp. 47-50. Beckwith, T.G., Marangoni, R.D., and Lienhard, J.H., Mechanical Measurements, 5th Ed., Addison-Wesley, New York, 1993. Dols, W.S., and Persily, A.K., "A Study of Ventilation Measurement In an Office Building," Airflow Performance of Building Envelopes, Components, and 64

Systems, ASTM STP 1255, Mark P. Modera and A.K. Persily, Eds., ASTM, Philadelphia, 1995, pp. 23-46. Drees, K., Wenger, J., and Janu, G., 1992, "Ventilation Airflow Measurement for ASHRAE Standard 62-1989," ASHRAE Journal, Vol. 34, No. 10, pp. 4045. Ebtron Inc., Product Specifications, 1663 Hwy. 701 S., Loris SC 29569. Elovitz, D.M., 1995, "Minimum Outside Air Control Methods for VAV Systems," ASHRAE Transactions, Vol. 101, Pt. 2, pp. 613-618. Engelhard Corporation, Product Specifications, 6489 Calle Real, Goleta, CA 93117. Graves, L.R., 1995, "VAV Mixed Air Plenum Pressure Control," Heating, Piping, and Air Conditioning, Vol. 67, No. 8, pp. 53-55. Haines, R.W., 1986, "Outside Air Volume Control in a VAV System," Heating, Piping, and Air Conditioning, Vol. 58, No. 10, pp. 130-131. Haines, R.W., 1994, "Ventilation Air, The Economy Cycle, and VAV,", Heating, Piping, and Air Conditioning, Vol. 66, No. 10, pp. 71-73. Janu, G., Wenger, J.D., and Nesler, C.G., 1995, "Strategies for Outdoor Airflow Control from a System Perspective," ASHRAE Transactions, Vol. 101, Pt. 2, pp. 631-643. Ke, Y., and Mumma, S.A., 1997a, "Using Carbon Dioxide Measurements to Determine Occupancy for Ventilation Controls," ASHRAE Transactions, Vol. 103, Pt. 3. Ke, Y., Mumma, S.A., and Stanke, D., 1997b, "Simulation Results and Analysis of Eight Ventilation Control Strategies in VAV Systems," ASHRAE Transactions, Vol. 103, Pt. 3. Kettler, J.P., 1988, "Field Problems Associated with Return Fans on VAV Systems," ASHRAE Transactions, Vol. 94, Pt. 1, pp. 1477-1483. Kettler, J.P., 1998, "Controlling Minimum Ventilation Volume in VAV Systems," ASHRAE Journal, Vol. 40, No. 5, pp. 45-50. Kettler, J.P., 1995, "Minimum Ventilation Control for VAV Systems: Fan Tracking vs. Workable Solutions," ASHRAE Transactions, Vol. 101, Pt. 2, pp. 625630.

65

Kreider, J.F., and Rabl, A., Heating and Cooling of Buildings, McGraw-Hill, New York, 1994. Levenhagen, J., 1992, "Control Systems to Comply with ASHRAE Standard 621989," ASHRAE Journal, Vol. 34, No. 9, pp. 40-44 Maki, K.S., Zhenhai, L, Chamberlin, G.A., and Christianson, L.L., 1997, "VAV System PerformanceField Characterization of Airflow System Diagnosis Tools, and Operational Design Implications," ASHRAE Transactions, Vol. 103, Pt. 3. Mamac Systems, Inc., Product Specifications, 10913 Valley View Road, Minneapolis, MN 55344. Mease, N.E., Cleaveland, W.G. Jr., Mattingly, G.E., and Hall, J.M., 1992, "Air Speed Calibrations at the National Institute of Standards and Technology," Proceedings of the 1992 Measurement Science Conference, Anaheim, CA. Meckler, M., 1994, "Demand-Control Ventilation Strategies for Acceptable IAQ," Heating, Piping, and Air Conditioning, Vol. 66, No. 5, pp. 71-74. MKS Instruments, Inc., Product Specifications, 6 Shattuck Road, Andover, MA 01810. Montgomery, D. and Runger, G., 1994, Applied Statistics and Probability for Engineers, John Wiley and Sons, Inc., New York. Mumma, S.A., and Wong, Y.M., 1990, "Analytical Evaluation of Outdoor Airflow Rate Variation vs. Supply Airflow rate Variation in Variable-Air-Volume Systems When the Outdoor Air Damper Position is Fixed," ASHRAE Transactions, Vol. 96, Pt. 1, pp. 1197-1208. Ower, E., and Pankhurst, R.C., 1977, The Measurement of Airflow, 5th Ed., Pergamoh Press, New York. Persily, A., 1993, "Ventilation, Carbon Dioxide, and ASHRAE Standard 62-1989," ASHRAE Journal, Vol. 35, No. 7, pp. 40-44. Solberg, D., Dougan, D., and Damiano, L, 1990, "Measurement for the Control of Fresh Air Intakes," ASHRAE Journal, Vol. 32, No. 1, pp. 45-51. Sterling, E.M., Collet, C.W., and Turner, S., 1992, "Commissioning to Avoid Indoor Air Quality Problems," ASHRAE Journal, Vol. 34, No. 10, pp. 2832.

66

APPENDIX A - JCEM LABORATORY AIR SYSTEM DESCRIPTION A.1 System Components The Airside System includes the following major components: Main Air Handling Unit (with variable speed drive) OACS (with variable speed drive) Return Fan (with variable speed drive) Fan-Powered Mixing Boxes Dampers ZSIM-1 & 2 FSZ-E&W A.2 Component Descriptions A.2.1 Main Air Handling Unit The main air handling unit provides cooling for the temperature control zones, it is a horizontal draw-through unit built with, in order, outside air and return air mixing dampers, a filter bank, a cooling coil, a reheat coil, and a supply fan. To meet the combined loads in the four temperature control zones an airflow rate of 9600 CFM at 55 F would be required. Design Data Description Manufacturer Model No.

Climate Changer Trane 25E

Mixing Dampers The AHU has a combination filter and mixing box which has two opposed blade dampers externally interlocked and driven by a direct-coupled electronic actuator (see below). Filters The combination filter and mixing box contains six 20"x25" UL Class 2 approved filters.

67

Cooling Coil The cooling coil has a sensible cooling capacity of 240 MBH when 100% water solution is used as the heat transfer fluid. This capacity is based on a supply airflow of 10,000 CFM, a coil leaving air temperature of 53.5 F, and a mixed air dry bulb temperature of 80 F at 20% outside air on a simulated design summer day (100 F DB/80 F WB). The pressure drop across the coil given below is based on 100% water solution. A condensate drain located on the bottom of the air handler beneath the coil is piped to a nearby floor drain. Entering Air Temp Entering Water Temp Water System Temp Rise Water Flow Rate Water Side Pressure Drop Face Velocity Capacity Rows 80 F/60 F DB/WB 44 op 12 F 40GPM 2.1'water 500 fpm 240 MBH 4

Reheat Coil A hot water coil provides reheat and morning warm-up capabilities. The coil has a 150 MBH capacity which will provide a 15 F temperature rise at design airflow. Entering Air Temp Entering Water Temp Water System Temp Drop Water Flow Rate Water Side Pressure Drop Face Velocity Capacity Rows Supply Fan Airflow Ext. Static Pressure Fan Diameter Speed Motor Variable Speed Drive A.2.2 Outside Air Conditioning Station The OACS offers unique experimental flexibility by providing control of ventilation air conditions entering the main air handling unit. The unit has the capability to control up to 100% ventilation air over a wide range of conditions. Moreover, in the Larson lab, it has been installed with additional ducting to allow 68 60 F 120 F 20 F 30GPM 1.5' water 500 fpm 150 MBH 1

12,000 SCFM 2.5" Water 22 inches FC 880 RPM 15 hp/480 V/60 hz/3 phase Allen Bradley

the discharge air to bypass the main air-handler and be diverted directly to the exhaust trunk. In this full bypass mode, the unit can be operated as an additional plant load parallel to the main air-handler. Dampers have been installed in the bypass branch downstream of the unit leading to the exhaust trunk. Dampers have also been installed in the exhaust trunk directly upstream of the exhaust grill. The unit includes, in order, a filter bank, a freeze protection damper, a preheat coil, an evaporative cooler, an internal face and bypass damper, a cooling coil, a reheat coil, a variable speed fan and a steam injection humidifier. These components are sized to generate the worst case outside air conditions (based on cooling load) given the extremes of actual intake air conditions. Worst Case Summer Design Day Outside Air Conditions: Dry bulb 100 F Wet bulb 80 F Humidity Ratio 0.0175 lba/lbw Actual Winter Design Day Outside Air Conditions: Dry bulb -10 F Humidity Ratio 0.0005 lba/lbw Design Data Description Manufacturer Model No. Filters The AHU has a combination filter and damper box, which contains eight 16"x25" UL Class 2 approved filters. Freeze Protection Damper The combination filter and damper box has an opposed blade damper that is driven open or closed by a fail-safe pneumatic actuator. Electric Resistance Preheat Coil The preheat coil is an electric resistance coil. Description Open Coil Duct Heater Manufacturer Markel Model Number 3HF44 Capacity 44kW/480V/3-phase Design Airflow Rate 10,000 CFM @ 100% Face Velocity 510 FPM @ design airflow Temperature Rise 17 F design airflow Staged Control 3 - 24VAC Contactors Climate Changer Trane 31A

69

Direct Evaporative Cooler A direct evaporative washer is used to cool and humidify outside air. The unit is installed in a basin which is fitted with a submersible pump, a float valve assembly, and an overflow tube which is piped via the air handler's condensate drain to a nearby floor drain. Description Manufacturer Media Face Velocity Air Side Pressure Drop Effectiveness Direct Evaporative Washer SPEC AIR 12" paperboard 500 FPM 0.2" water 75%

Internal Face and Bypass Dampers Upstream of the cooling and heating coils are a set of opposed blade dampers that allow airflow to bypass the chilled water and reheat coils. These dampers are internally interlocked and driven by an electronic direct-coupled damper actuator (see below). Cooling Coil The unit has a 360 MBH cooling coil when 100% water is used as the heat transfer fluid. This provides the capability to sensibly cool approximately 10,000 CFM of outside air to 52 F on a design summer day (as specified above). The pressure drop across the coil given below is based on 100% water solution. A condensate drain on the bottom of the air handler beneath the coil is piped to a nearby floor drain. Entering Air Temp Entering Water Temp Water System Temp Rise Water Flow Rate Water Side Pressure Drop Face Velocity Capacity Rows 91/63 DB/WB 44 op 12 F 60 GPM 4.5' water 500 FPM 360 MBH 4

Electric Resistance Reheat Coil The reheat coil is a modulating electric resistance coil and It can generate a 13 F temperature rise at 10,000 CFM and a 66 F temperature rise at 2000 CFM. Using silicon-controlled rectifier (SCR) technology the output of the heater can be throttled from 0-100% continuously. Description Manufacturer Model Number Airflow Rate Open Coil Duct Heater Markel 3HF35 10,000 CFM 70

Face Velocity Temperature Rise SCR Control Supply Fan Airflow Static Pressure Rise Fan Diameter Speed Motor Variable Speed Drive

480 FPM 13 F 900 Series Power Switch (2-10 VDC)

12,000 SCFM 1.0" Water 22.5 inches 560 RPM 7.5 HP/480 V/60 hz/3 phase Allen Bradley

Steam Injection Humidifier To simulate latent loads on the air conditioning system the OACS has a 10 psig, 150 Ib/hr steam injection humidifier which is designed to raise the humidity ratio from 0.0005 \bjlb* to 0.0175 lba/lbw at 2000 CFM (20% Outside Air). Description Manufacturer Model Number Design Steam Pressure Design Steam Rate No. Dispersion Tubes Dispersion Tube Length A.2.3 Return Fan Description Airflow Static Pressure Fan Diameter Speed Motor Variable Speed Drive A.2.4 Fan-Powered Mixing Boxes The main air-handling unit supplies medium pressure conditioned air to the parallel VAV fan-powered mixing boxes serving each zone. Each box is outfitted with electric resistance reheat elements. ZSIM-1 & 2 To provide local control of conditions powered mixing boxes have been installed zone. These boxes have built-in dampers resistance reheat coils downstream of the in the Zone Simulators, parallel fanin the air system upstream of each on the primary air inlet and electric mixing section. ZSIM-1 is outfitted In-Line Centrifugal Fan 12,000 SCFM 1.0" Water 24.5 inches 1750 RPM 5HP/480V/3-phase Allen Bradley Steam Injection Humidifier Dri-Steem ML6-60 10psi 150 Ib/hr 4 30 inches

71

with an additional damper and duct section to allow the fan to be operated either in parallel or series mode. Description Manufacturer Model Number Design Primary Airflow Max Secondary Airflow Fan Motor Power Fan Speeds Heater Capacity Face Velocity Temperature Rise Staged Control A.2.5 Dampers All the electrical damper actuators are built by Belimo. Description Direct-Coupled Damper Motor Location OACS (2), Main Air-Handler, ZSIM-1 (Series), Exhaust Trunk Model Number SM 24-SR Power Supply 24 VAC Power Consumption 2 Watts Torque 15N-m(133in-lbf) Control Signal 0-10 VDC or 0-20 V Phase-Cut Operating Range 2-10 VDC Output Signal 2-10 VDC (max 0.5 mA) Running Time 100-200 sec @ 0-15 Nm torque & 95 degree stroke Description Direct-Coupled Damper Motor Location ZSIM-1 & 2, FSZ-E&W Model Number NM 24-SR Power Supply 24 VAC Power Consumption 1.3 Watts Torque 8 N-m (75 in-lbf) Control Signal 0-10 VDC or 0-20 V Phase-Cut Operating Range 2-10 VDC Output Signal 2-10 VDC (max 0.7 mA) Running Time 0-150 sec @ 0-35 degree stroke 150 sec @ 35-95 degree stroke Parallel Fan-Powered Mixing Box with Electric Resistance Reheat Environmental Technologies WFEH-ll-18 4000 CFM 2600 CFM @ high speed and 0.25" exiting static pressure 2 @ 1/2HP/277V/1-phase 3 10kW/480V/3-phase 1700 FPM @ design airflow 9.5 F @ design airflow 2 - 24VAC contactors

72

A.2.6 Zone Simulators 1 & 2 The two zone simulators are expanded sections of ductwork containing, in order, a steam injection humidifier, a chilled water coil, and an electric resistance heater. Steam Injection Humidifier Manufacturer Model Number Design Steam Pressure Design Steam Rate No. Dispersion Tubes Dispersion Tube Length Chilled Water Coil Coil Manufacturer Model No. Entering Air Temp Entering Water Temp Water System Temp Rise Water Flow Rate Face Velocity Capacity Rows Electric Resistance Heater Manufacturer Model Number Heater Capacity

Dri-Steem ML5-60 10psi 50 Ib/hr 5 36 inches

Temtrol 5WC16-36x36x4-08A-RH 95/63 DB/WB

44 op

12 F 10GPM 450 FPM @ design airflow 60MBH 4

Face Velocity
Temperature Rise SCR Control Input Control Signal Approx.

Markel 3HF25 25kW/480V/3-phase 450 FPM @ design airflow 24 F @ design airflow

900 Series Power Switch

2-10VDC

73

APPENDIX B - ANALYSIS OF THE EFFECTS OF HUMIDITY AND PRESSURE ON THE CALCULATED AIRFLOW RATE As shown from Equation 2 rewritten below, the value of the calculated airflow rate from an averaging Pitot-tube array is a function of the density of the measured airflow. V = 1096.7 , K Equation 2

Equation 3 gives the density as a function of the atmospheric pressure and the temperature of the air stream: rp

p=

Equation 3

R-T
This relationship is exact when there is no moisture in the air. Equation 3 only provides an approximate value of the density for actual air conditions. The true value of the density of moist air is given by Equation 32 (ASHRAE 1997c): p = T ^YT \ Equation 32

0.7543 (t+459.6) (l +1.6078 W)

where p = total pressure [in. Hg] t = dry-bulb temperature [<F] W = humidity ratio [Iby/lbfial

The absolute error associated with the calculation of the airflow velocity from Equation 2 when the effects of humidity are neglected can be found from Equation 33:

% Error =

Equation 33

IP. where: pa = actual density from Equation 32 pc = density calculated from Equation 3

The average and maximum absolute errors introduced by using Equation 3 in place of Equation 32 were calculated for several U.S. locations using hourly 74

data. These values are shown in Table 15. The results show that the difference in the calculated airflow velocities is extremely small (maximum errors of less than 0.65% for all locations considered). This implies that the measurement of moisture levels in the air is not required for accurate calculations of the airflow rates from averaging Pitot-tube arrays. Table 15: Errors calculating air density while neglecting the effects of humidity.
Location Denver, CO Madison, Wl Miami, FL New York, NY PhoenixTAZ San Francisco, CA Average Error 015% 018% 0 42% 020% 0 19% 0.21% Maximum Error 0,48% 0 57% 063% 0 56%

0.51%
0.33%

It has been shown that Equation 3 is accurate enough to use without considering the effects of the moisture in air. However, most typical building installations do not measure nor have real-time access to the atmospheric pressure that is also needed in Equation 3. The absolute errors associated with the calculation of the airflow rate when the annual average atmospheric pressure is substituted for the actual atmospheric are also shown in Table 16. The average absolute error for the locations considered is less than 0.30%. Table 17 shows the absolute errors when the effects of humidity are neglected and the annual average atmospheric pressure is used to calculate air density. Table 16: Errors calculating air density using the annual average atmospheric pressure.
Location Denver CO Madison Wl Miami H New York, NY Phoenix, AZ San Francisco, CA Average Error 011% 0 79 Vu 013% 0 30% 0 19'X 0.19% Maximum Error 1 38% 1 37% 0 31% 1 65% 0 83r/o 1.00%

Table 17: Errors calculating the air density neglecting the effects of humidity and using the annual average atmospheric pressure.
Location Denver CO MadibOi Wl Miami F L ~ Nr*York NY
Phnpnivi A /

Average Error 0 28% 0 38% 0 43% 0 37% 0 29% 0.28%

Maximum Error 1 52% 1 73% 1.28% 1 77% 0.91% 1.28%

San Francisco, CA

As shown in Table 17, the average error introduced by using Equation 3 to calculate air density while using the annual atmospheric pressure is less than 75

0.50%. Given thesefindings,the use of Equation 3 to calculate the density of the air stream will not result in large errors for calculated airflow rates.

APPENDIX C - CONCENTRATION BALANCE AIRFLOW MEASUREMENT TECHNIQUE TEST RESULTS C. 1 C02 Sensor Repeatability Errors

As stated in Section 2.1.2.2, manufacturer's listed repeatability errors for CO2 sensors are typically based upon a one-year period. On average, these errors are +/- 30 to 50 ppm. For a proper error analysis of the concentration balance airflow measurement technique, only a small time frame needs to be considered for the repeatability errors. Tests were run in the JCEM Laboratory using a commercially available CO2 balance measurement apparatus that included a C0 2 sensor. Two known concentrations of CO2 were used to test the repeatability of this sensor. Figure 31 shows the results of the testing. The dashed lines in Figure 31 represent the stable concentration readings from the two CO2 sources. On average, a sampling rate of 3 minutes for each airflow was required in order to achieve a repeatability of +/- 3 ppm.
600

550.

500

1"
3450 c
0 B <0

1 f 1

f f
I

b 400.

(3 350
M

8
300.

v.
0:03 0:06 0:09 0:12 0:15 0:18 Time [min]

2>0 200

V_
0:21

I
(*24

a00

a27

Figure 31: Repeatability tests for a commercially available CO2 sensor.

77

C.2

Measurement Technique Verification

Predicted measurement error for the concentration balance technique was presented in Section 3.2.2.2. Figure 32 shows the predicted errors for the JCEM Laboratory using the concentration balance airflow measurement technique.
100% 90% 80% 70% 60% 50% 40% 30% 20% 10% flaOQiaoflflflaffl 0% 0 50 100 150 2 0 0 2 5 0 3 0 0 3 5 0 4 0 0 4 5 0 co2RA- co,OA 10% OA Fraction 30% OA Fraction e 50% OA Fraction

Figure 32: Predicted measurement errors as a function of OA% for JCEM Laboratory (5% error in SA measurement, 1 ppm acquisition error, repeatability errors of 3 ppm). Tests were conducted at the JCEM Laboratory to confirm the errors predicted in Figure 32. Outside air fractions of 10%, 30%, and 50% were used. Figure 33 shows the measured errors using the concentration balance measurement technique. The slightly higher measured errors are most likely due to a supply airflow measurement error greater than the 5% assumed in Figure 32 because of the low supply airflow rates used for test.

78

Figure 33: Measured errors in the calculated outside airflow rate using theconcentration balance method. As illustrated in Figure 33, the outside air fraction has a significant effect on the predicted accuracy of this measurement technique. Use of the concentration balance measurement technique may not be valid for lower outside air fractions unless large differences in the return and outside air CO2 concentration levels exist. Further recommendations for this measurement technique are included in Section 3.2.3.

79

APPENDIX D - LABORATORY TEST RESULTS Included in this appendix are graphical results for the systems described in Section 4.3.4. For each system, three graphs are shown. The first shows the supply and outside airflow rates during the duration of the tests. Both airflow measurements were made with averaging Pitot-tube arrays. Outside airflow measurements were made with the high-accuracy, auto-zeroing differential pressure transmitter. Calibration of the supply airflow measurements is described in Section 4.2. The second graph shows the outside airflow percentage of design vs. the supply airflow percentage of design. For all systems tested, a constant outside air percentage of 100% of design would be ideal. Finally, the third graph shows the percentage error between the two direct airflow measurement techniques; the averaging Pitot-tube array and the electronic thermal anemometer. Throughout all analyses, the outside averaging Pitot-tube array with an auto-zeroing differential pressure transmitter was used as the reference airflow measurement technique.

80

D.1

Fixed Damper Pos., 20% OA, Case 1-A

OA T e m p e r a t u r e [ D e g F ] 60 00 1000 j Sjf jj; O O * <
0i

65.00

70.00

75.00

80.00

85.00

90.00
twin

900 800 700 600 500 400 300 200 tro 30 0 4J

o Outside Air Velocity + OA Temp

5
M

o
m

^-AVI3IE* T *

" " W l W i W I I I T T T I I 1 FTPI1 Hill l l l l M i l r*= 600 800 D00 1200 Supply A i r V e l o c i t y ( F P M )

ff ^

. tuo . tooo . 900 . 800 . 700 600 . 500 . 400 . 300 . 200 . DO 1500

1 1400

Figure 34: Outside air velocity versus OA temperature and supply air velocity.
110% 100% S.

m>

" ^

ft*&

-V^

r**

Q 70% o 60% 50% 40% 2 30% 20% 10% 0% 0%

i i A T - * " i i**n**^Mfc*i** * * * * '

~m**

-n~

- i

10%

20%

30%

40%

50%

60%

70%

80%

90%

100% 110%

S u p p l y A i r % of Design

Figure 35: Outside airflow % of design with predicted measurement errors.

15% 12%
ft ft

6% Vft*

t*-

I 3%

/o
Ul -6% -9% -12% -15% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110%

S -3% t

S u p p l y A i r % of Design

Figure 36: Comparison of direct airflow measurement techniques.

81

D.2

Fixed Damper Pos., 30% OA, Case 1-B

OA T e m p e r a t u r e [ D e g F ]

60.00

65.00

70.00

75.00

80.00

85.00

tooo
S 900 o Outside Air Velocity + OA Temp

fc, 8 0
m 700 "5 600 0
"to 500

t 400 < 300 5 200

M

orf^

^<lf*^f-**"

#*^V^
1)00 1200 Supply Air Velocity ( F P M )

0****^
1400

90.00 t200 - XV0 W00 900 I- B O O 700 600 - 500 400 300 - 200 tJO BOO

S O

t)0

400

600

800

Figure 37: Outside air velocity versus OA temperature and supply air velocity.
110% 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110% Supply Air % of Design

ML
J^ r

y^

i+ i

Figure 38: Outside airflow % of design with predicted measurement errors.

15% 12%. 9%.

f 6%.
| 3%.

%
"& - 3 % .
UJ - 6 %

A A A A*A A* * (UJL * * 1 * * *
^ * *A

* A * *

ji

* " * %

"ffi&l.

-9%. -12% . -15% 0" % 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 11 D %

Supply Air % of Design

Figure 39: Comparison of direct airflow measurement techniques. 82

Fixed Damper Pos., 40% OA, Case 1-C

OA Temperature [Deg F] 60.00 65.00
1

70.00
1

75.00
1

80.00
1

85.00
1

tooo

90.00 1200

900 800 700 600 500 400 300 200 t)0 400 Jtocdtf*' " O * " 0 * 0 0 0 j,0~fl " f*"
J J

0.

u.

o Outside Air Velocity + OA Temp

tuo tooo
900 800 700 600 500 400 300 200 DO COO

* >

A
jrft+****+^

'******
^.Ljrfii H*!L

$*&%*^4fr++***#*^
600 1

800 O00 1200 1 r 1 Supply Air Velocity ( F P M )

1400 1

Figure 40: Supply and outside airflows with predicted measurement errors.
110% 100% 90%

80% 70% 60% 50%

>

7^*

8 2 30%

40%

<U 20%
10% 0%. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110%

Supply Air % of Design

Figure 41: Outside airflow % of design with predicted measurement errors.

15% 12% 9% s s-

!" 6% | 3%

-t*~

4A

0%
"S -3% UJ -6% -9% -12% -15% 0% 10% 20% 30% 40% 50% 60% 70%

'V*U f

80%

90%

100%

110%

Supply Air % of Design

Figure 42: Comparison of direct airflow measurement techniques. 83

DA

Plenum Pressure Control, 20% OA, Case 2-A

OA Temperature [Deg F]
60.00 t400 Q. 1200 U^ ff tOOO 65.00

70.00

75.00

80.00 -+-

85.00

-+o Outside Air Velocity + OA Temp

" 5 o
"3 800

-I- V|-^ + - N N ^ t f ^ ^ *
^_

>
< 600 a > o W 400 9

^ i c ^ f l t ^ B t P J t u I % a ^ 00* .ftp" 0 ,%<* *** 'getfb-go o<s,tg,d>""Wg^h^^lfr-

90.00 t200 tt)0 1000 900 800 700 600 500 400 300 200 L t)0 COO

200 400

600

800

D00

1200

1400

Supply Air Velocity ( F P M )

Figure 43: Supply and outside airflows with predicted measurement errors.
120% 115% 110% 105% 100%

95%
90% 85% 80% 0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100% 110%

Supply Air % of Design

Figure 44: Outside airflow % of design with predicted measurement errors.

15% 12% 9%

g 6%

1 3*
S -3%
ID -6%

, < i * * * * * *

**1

-A*\ V ^

.'.V.'.^.t^

-9% -12% -15% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110%

Supply Air % of Design

Figure 45: Comparison of direct airflow measurement techniques. 84

D.5

Plenum Pressure Control, 40% OA, Case 2-C

OA Temperature [Dog FJ 60.00 t400 0. lu 1200 65.00
1

70.00
1

75.00
1

80.00
1

85.00
1

90.00 1200

o CXitside Air Velocity + 0ATemp

Itro
1000 900 800 700 600 500 400 300 200 L DO
.

*s
o
T5 <
b O

ff tooo
800 600 400 200 400
, 1 , 1 ,

t
-H-M--f-Hrt-H. + + # * # - + + F H h * H n ifli HHHH<iHll ll<?l

600

800

1300

1200

1400

S00

Supply Air Velocity ( F P M )

Figure 46: Supply and outside airflows with predicted measurement errors.
120% 115% 3 110%
M O

105% o X 100%

< 95% 1 90%

< 3

85%
80%. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110%

S u p p l y Air % of Design

Figure 47: Outside airflow % of design with predicted measurement errors.

15% 12% 9% I Error % reading)
6% 3% 0% -3% -6%

-4&1

Sm*

% T*arb?** &** K *

&+K

fry*..*'*+ v*
A
A A

\ -9%
-12% -15% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110%

S u p p l y A i r % of Design

Figure 48: Comparison of direct airflow measurement techniques. 85

Direct Control with Econ. Duct, 20% OA, Averaging Pitot-tube Array, Case 3-A
OA Temperature [Deg F] 60.00 t400 65.00
1

70.00
1

75.00
1

80.00
1

85.00
-1

90.00 1200 VD0

o.
u.

t200

r tooo
800 < ft B " R 600 400 200

o Outside Air Velocity + OATemp

tooo

- 900
800

700

A
-H^+i-^m-H- n i+i-H-i-HM-i in I n in11n n imilium*

^
400

noomtm/iitnAO o o o i o v r t o

oo moo Oft> oooo o oaooooaooiODOflDOeoaonoooaoaaaoato

600 500 400 300 200 DO 0

600

100

U00

1200

1400

C00

Supply Air V e l o c i t y (FP M)

Figure 49: Supply and outside airflows with predicted measurement errors.
120% 115%

c J? 110%
01 *5 o

2 105% o X 100% < 8 9 5 %

I'^tlfiirtWi
10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110% S u p p l y Air % of Design

1 90%
85% 80% 0%

Figure 50: Outside airflow % of design with predicted measurement errors.

15.0% 12.0% 9.0%
4

6.0% 3.0% 0.0%

1
fc
o

-3.0%

Ul -6.0% -9.0% -12.0% -15.0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110% S u p p l y Air % of Design

Figure 51: Comparison of direct airflow measurement techniques. 86

Direct Control with Econ. Duct, 40% OA, Averaging Pitot-tube Array, Case 3-B
OA Temperature [ D e g F ] 60.00 1400 0. 1200 U. j? 1000 o 65.00
1

70.00
1

75.00
1

80.00
1

85.00

o Outside Air Velocity + OA Temp -++W++H--H 4 l+fr Willi |-WiH I HI lillI III! I llllllfr

> <

800 600 400 200 400 ^

p o^t^rf* o orf>o PB>ocP oo

[OOIOOI>IJO*J

90.00 1200 1T)0 1000 900 800 700 600 500 400 300 200 t)0 0 (DO) (200) (300) COO

600

800

D00

1200

1400

Supply Air V e l o c i t y ( F P M )

Figure 52: Supply and outside airflows with predicted measurement errors.

Figure 53: Outside airflow % of design with predicted measurement errors.

15% 12% 9% 6% 3%

tf 0%
'. Error
-3% -6% -.9% -12% -15% 0% 10% 20% 30%

flff % t

^*^****&<*E*

40%

50%

60%

70%

80%

90%

100%

110%

S u p p l y Air % of D e s i g n

Figure 54: Comparison of direct airflow measurement techniques. 87

D.8 Direct Control w/Econ. Duct, 20% OA, Electronic Thermal Anemometry, Case 4-A
OA Temperature [Deg F] 60.00 65.00 1 70.00 1 75.00 1 80.00 1

85.00

90.00
t200 ttio

1400

1200 1000
800 600 400 200 400
*

. tooo
o Outside Air Velocity + OA Temp 900 800 700 600 500 400 300 200 TJ0 0 (t)0) . (200) (300) 1400

I
<

4*+

= H ~ i t - H l l M + l l l + 4 - H M I I 4 + l ] l - l l i m H l I I ll|il

W 9 O

600

800

D00

1200

BOO

Supply Air Velocity (FP M )

Figure 55: Supply and outside airflows with predicted measurement errors.
120% 115% 3 110%
0)

2 105% o
X 100%

<
o 95%
g 90% O 85% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110%

Supply Air % of Design

Figure 56: Outside airflow % of design with predicted measurement errors.

15% 12%

6%

5 -3%
u

-6% -9%
-12% -15%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

110%

Supply Air % of Design

Figure 57: Comparison of direct airflow measurement techniques. 88

D.9

Direct Control w/Econ. Duct, 40% OA, Electronic Thermal Anemometry, Case 4~B
OA Temperature [Deg F] 60.00 1400 fc 1200 65.00 70.00 75.00 80.00 85.00 90.00 1200 . 1t)0 1000 900 800 700 600 500 400 300 200 V0 0 COO

-+o Outside Air Velocity + OA Temp

-+-

.9 tooo *S o
*S < 800 600 400 9

-H-+ H0+ * Hr^fH4-HH-H^HH^#+++trt-H*h I -WaOjVoqji>iCPP<x>ctfa> oosoo>oo < < > ooa 0 0 <M>oa>cu>oaPa B%

>

^
400

200

600

800

t)00

1200

1400

Supply Air Velocity ( F P M )

Figure 58: Supply and outside airflows with predicted measurement errors.
120% 115%

.5 110% 2 105%
o >? 100%

< 95% 1 90%

pppspipii
10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110% Supply Air % of Design

80% 0%

Figure 59: Outside airflow % of design with predicted measurement errors.

15% 12% 9%

. 1 6% 1 3% S, *
S -3% fc
(U -6% -9% -12%
i i ' i * i i i i i i i ' i

n y i & *i+f>*;*\ & v y A y ^ j j i

-15%0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100% 110%

Supply Air % of Design

Figure 60: Comparison of direct airflow measurement techniques. 89

D.10 Volume Tracking, 20% OA, Electronic Thermal Anemometry, Case 5-A
OA Temperature [Deg F ] 60.00 t400 ^ 1200 65.00 70.00 75.00 80.00 85.00 90.00 1200 tt>0 1,000 900 800 . 700 600 500 400 300 200 DO 0 COO

-+
o Outside Air Velocity + OA Temp

-+

-+

2> tOOO

o
< o 400 800 600 +
Vtmmo

+ - H # + - | I H t HH-H4
<*n6o o <U0 o W * OOCD

Hm 144 ZZZ - ~ a.~'8,"rfB'

9 200 400

600

800

t)00

1200

1400

Supply Air Velocity ( F P M )

Figure 61: Supply and outside airflows with predicted measurement errors.
220% 200% c
Ol

| 180% Q *o 160% 140% o o 5 120%

t*

MtyfyiH;

jf*t*

jli

#I^U^

100% 80% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110% Supply Air % of Design

Figure 62: Outside airflow % of design with predicted measurement errors.

15% 12% 9%

1 a
8
Ul

3% 0% -3% -6% -9% -12% -15% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110% Supply Air % of Design

Sr

**

>.^*V

**A

_.*^

^* ^ i** .

Figure 63: Comparison of direct airflow measurement techniques. 90

D.11 Direct Control w/Dedicated Duct, 20% OA, Averaging Pitot-tube Array, Case 6-A
OA Temperature [Deg F]
60.00 V400

65.00

70.00

75.00

80.00
-t-

85.00

t200 o. u. S? tooo

-+o Outside Air Velocity + OA Temp

90.00 1200

tuo tooo
900 800 700 600 500 400 300 200 1)0 C00

" 5 o >
<
4>

800

-^7HH++fc"^+%|H^^
Q<*1> <*<fl>seV*>Q9 "oorf " *><%+nn0 tefppj*

600 -^eofrgfri''%&"Nfi
400

9 O

200 400

600

800

TJOO

1200

1400

Supply Air Velocity ( F P M )

Figure 64: Supply and outside airflows with predicted measurement errors.
120% tB%

J? VD%
0)

2 t)5%
0 X t)0% * 95%

3 90%
9 85% 80% 0% % 20% 30% 40% 50% 60% 70% 80% 90% t)0% TD%

Supply Air % of Design

Figure 65: Outside airflow % of design with predicted measurement errors.

6% .. 12% 9%

f
X

6%
0%.

1 3%. *#**:** * V - * *
J* 1
A,

>

n4

-.!./**
, " .V

S* "3%
Uj -6%
X

-9% -12%

-%

O%

t)%

20%

30%

40%

50%

60%

70%

80%

90%

t)0%

1t)%

Supply Air % of Design

Figure 66: Comparison of direct airflow measurement techniques.

91

DA 2 Direct Control w/Dedicated Duct, 30% OA, Averaging Pitot-tube Array, Case 6-6
OA Temperature [Deg F ] 60.00 1400 70.00 75.00 80.00

65.00

85.00

90.00 1200 11)0

-+4+^+HH+-t4^+HH.4HHH4-^-HH4f-HtfHf 4 U H -HI I I H 4

0.

1200 1,000 800 600 400 200 400 600 800 D00 1200 1400

u. ff % > < <o

w o 0

tooo

^wto^Bafto** 0 ****"

oo

dcp moooop oo eoOo.

o Outside Air Velocity + OA Terrp

900 800 700 600 . 500 400 300 200 DO COO

Supply Air Velocity ( F P M )

Figure 67: Supply and outside airflows with predicted measurement errors.

Figure 68: Outside airflow % of design with predicted measurement errors.

6%

12% 9%

13 . %
x 0%.

6%

* * *{>*\/t" >
A

" ** -4*'*1' V**- r* ft***

S* "3%
U -6%
X

-9% -12% -15% 0% TJ% 20% 30% 40% 50% 60% 70% 80% 90% VO% 1t)% Supply Air % o f Design

Figure 69: Comparison of direct airflow measurement techniques.

92

D.13 Direct Cntri w/Dedicated Duct, 20% OA, Electronic Thermal Anemometry, Case 7-A
OA Temperature [Deg F]
60.00

65.00

70.00

75.00

80.00

85.00

WOO t t200 bu

-+o Outside Air Velocity + OA Temp

-+

-+-

90.00 1200 t0

tooo
900 800 700 600 500 400 . 300 200

2P tooo "5 o
1 800

4 ^ M E b H H & +f-HMHH<lHtHHd^44lfcfefeft|| lr+41| *1

>

>
< R 600 400 200 400

m<WQ" 8n> 8t> ga< 0 0 ^ o < . "

i J < i > BSpoof&toBb

no
600
800 DOO 1200

W> H

COO

Supply Air Velocity ( F P M )

Figure 70: Supply and outside airflows with predicted measurement errors.
120%
115% 110%

8
105% o X 100%

<
a, | D 95% 90%

fefeMlf Vty
10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110% S u p p l y A i r % of D e s i g n

80%
0%

Figure 71: Outside airflow % of design with predicted measurement errors.

15% 12% 9%

fi 6%
1 X 3%
0% -3%

f^* i . 54*x 4 4 *' . ' ft {******+? A

** M * * ' A * A '

til -6

fc

-9%
-12% -15% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110% S u p p l y A i r % of D e s i g n

Figure 72: Comparison of direct airflow measurement techniques. 93

D.14 Direct Cntrl w/Dedicated Duct, 30% OA, Electronic Thermal Anemometry, Case 7-B
OA Temperature [Deg F] 60 00 t400 ? t200
+ + f #HHJ-HH-HH^+HNH^I

65.00

70.00

75.00

80.00

85.00

90 00 t200 . fO0 tooo . 900 800 . 700 600 . 500 400 . 300 200 . 130 0 BD O

"3
0 "5

2? tooo 800 600 400 200 D 41 O

1 l + i l lHHH#ilt4 >
b*%eB><ja) % *S<#e**"to> o Outside Air Velocity + OA Temp

<

<W8<fl*poO08*

# 0 8*.

>
< 1
9

SI

600

800

D00

1200

1400

Supply Air Velocity ( F P M )

Figure 73: Supply and outside airflows with predicted measurement errors.
120% 115%

& 110% 3 S 2 105%

o X 100% o
o

95%

^tw^wiff
10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110% Supply Air % of Design

90% O

0%

Figure 74: Outside airflow % of design with predicted measurement errors.

15% 12% 9%

. 1 6%
| 3%.
A

"* V "3%
UJ -6% -9% -12% -15% 6 0S 10% 20% 30%

it&i<V& *- >^ \V&!

ai

40%

50%

60%

70%

80%

90%

100%

110%

Supply Air % of Design

Figure 75: Comparison of direct airflow measurement techniques. 94

D.15 Injection Fan w/Dedicated Duct, 20% OA, Averaging Pitot-tube Array, Case 8-A
OA Temperature [Deg F] 60.00 t400 65.00 70.00 75.00 80.00 85.00 90.00 t200 tt)0

ff

o. *5 o Z >
5

t200

o Outside Air Velocity + OATerrp -T4HHti^ft.ii i^4--H=i ii il i M'W'n+iHt'lHiW44> \m

u "gB^tlrfV-tfg>p"Vo"ooq'0 ouMeTOO*Q-tTT u

tooo
. 900 800 700 600 . 500 400 . 300 200 t)0 0
BOO

tooo
800 600 400 200 400 600 800 BOO 1200 1400 Supply Air Velocity ( F P M )

Ar^-

"rtf"-< X C B I * ^

Figure 76: Supply and outside airflows with predicted measurement errors.

16%
sign
10%

D D5% 3
X t% 95% 90%
85% 80% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% TD% Supply Air % of Design

* <
8

jftlttltllWt'flW^

2 3 o

Figure 77: Outside airflow % of design with predicted measurement errors.

% _
12% 9%

? 6 % X %

1 **
* AM

S "3%.

-6%
X

""

-9% -12%

-%
0%

v%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1t )%

Supply Air % o f D e s i g n

Figure 78: Comparison of direct airflow measurement techniques.

95

D.16 Injection Fan w/Dedicated Duct, 30% OA, Averaging Pitot-tube Array, Case 8-B
OA Temperature [Deg F] 60.00 t400 65.00 70.00 75.00 1 80.00 1 85.00 90.00 1400 1,300 . t200 fDO tooo 900 800 700 600 500 400 300 200 DO
BOO

t200

1 : 1 o Outside Air Velocity + OA Temp

tooo
800

;ss%igaa?Bniiity w attfaiftuaa.'4*

600

400 200 400 600 800 D00 1200 1400 Supply Air Velocity ( F P M )

Figure 79: Supply and outside airflows with predicted measurement errors.

80% 0% U% 20% 30% 40% 50% 60% 70% 80% 90% t)0% 1C% Supply Air % o f Design

Figure 80: Outside airflow % of design with predicted measurement errors.

*% 12% 9% ? * X 6% 3% 0%

I -3% lS -6%
X -9% -12% -% 0%
I 1 1 1 ' 1 1 1 ' 1 ' 1 1 ' 1 -

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1t)%

Supply Air % of Design

Figure 81: Comparison of direct airflow measurement techniques.

96

D.17 Injection Fan w/Ded. Duct, 20% OA, Electronic Thermal Anemometry, Case 9-A
OA Temperature [Deg F] 60.00 WOO 65.00 70.00 75.00 80JOO 85.00 90.00 t200 tUO

1200
1

-+o Outside Air Velocity + OATerrp

-+-

-+-

-+-

tooo
900 800 . 700 600 500 400 . 300 200 tlO -B00

S tooo "5 o B00 "5 > < 600

400 200 400

11 in a i H -W HH-I4IK-DH i iiiniiimnii i i iniilil info. > H

W9 n f i p o o l e qateoorco o ntfi % ocnno<Twnftna noP^Mfr,

600

800

D00

1200

1400

Supply Air Velocity ( F P M )

Figure 82: Supply and outside airflows with predicted measurement errors.
120% 16%

85% 80% 0% D% 20% 30% 40% 50% 60% 70% 90% t)0% 1t>% Supply Air % of Design

Figure 83: Outside airflow % of design with predicted measurement errors.

12% . 9%. c" "8 M 5 6% . 3% . 0% -3%

V:.**.**'A'* 1 - v . " . : * - ' ..'*'. s y/-

******

** *t*

It*

|2 -6% . ** - 9 % -12%

-%.
0%

'

'

>

tr-

f>%

20%

30%

40%

50%

60%

70%

80%

90%

1)0%

TD%

Supply Air % of Design

Figure 84: Comparison of direct airflow measurement techniques.

97

D.18 Injection Fan w/Ded. Duct, 30% OA, Electronic Thermal Anemometry, Case 9-B
OA Temperature [Deg F] 60.00 1400 IL o 1200 65.00 70.00 75.00 80.00 85.00 90.00 1200 1B0 1000 900 800 700 600 500 400 300 200 DO C00

-+-

-+-

-+-

-H-H++III H i l l III lllll I I Hill III I Hill III Hill IHiHMi H|| ||U=>
l%oabfoOa oo #oco co a> tf> OD9> o oo oacaa

J? 1000 800 600

4> O 9

o Outside Air Velocity + OATemp

t
1200 1400

400 200 400 600 800 D00 Supply Air Velocity ( F P M )

Figure 85: Supply and outside airflows with predicted measurement errors.
120% 16% 5 10%

e
f

2 tK%
o X -00% * 9 95%

mm
V% 20% 30% 40% 50% 60% 70% 80% 90% 1)0% 1B% Supply Air % of Design

5 90%

85%
80%
0%

Figure 86: Outside airflow % of design with predicted measurement errors.

%
12% 9%

f
X
X

6% 3% 0%

5 -3%
UJ -6% -9% -12%

^att

,-.. f_.-. :\^\

Tr

*...'"A^ , :

>'&A

-%
0% t)%

'

20%

30%

40%

50%

60%

70%

80%

90%

V0%

1f>%

Supply Air % of Design

Figure 87: Comparison of direct airflow measurement techniques.

98

APPENDIX E - PRESSURE AND TEMPERATURE DATA Included in this Appendix is tabular data pertaining to the control of the HVAC system during testing. Information regarding the duct static pressure and air temperatures is specifically included. E. 1 Duct Static Pressures

Table 18 provides a summary of the duct static pressures for all the tests performed in the experimental analysis. See Section 5.1.2 for information regarding the analysis. Case numbers are the same as those listed in Table . Table 18: Duct static pressure data for systems tested.
Supply Duct Static Pres. (set point = 1.85 InWG)
System Description Fixed Damper Position Fixed Damper Position Fixed Damper Position Plenum Pressure Control Plenum Pressure Control Direct Control with Economizer Duct Direct Control with Economizer Dud D rect C u ' o> wJ- E< or J i i i " l ' u l CiCLtCoi[islhthL r oi<oi " ^n-umL 1 rautinq Prul orli il n t h 1 JCII - i ' 1 ' J.iJ Case 1 A 1-B 1-C 2A 2-C 3-A mean 1847 1846 1853 1844 1B71 1846 1847 IW1 18-U 1848 1 M. 1R..I 1R4' 1B44 1 IMP 10-18 1 I'll 1.850 stdsv 0058 0 057 0 377 00G0 0530 0 057 RMS 0058 0057 0 377 0061 0530 0057

Return Duct Static Pres. (set point = -1.0 InWG)

mean -0 990 -0980 -0 977 -0991 4)990 -0 990 -nqcn 0J31 0W1 NA -n> M 09*12 01 4991 1 r>ju O**JI stdev 0 024 0448 0 449 0022 0D24 0024 0022 0022 0022 NA 0 027 0022 0024 0023 0 000 0022 0 022 0.021 RMS 0020 0 449 0 449 0024 0026 0 026 0 024 0024 002: NA OtlA' 00^3 0 0.li 0024 OOuC 0 0>? 0 0/4 0.022

Outside Duct Static Pres. (set point = 0.0 InWG) mean

0000 0011 0 011 0000 0000 0000 0 000 0000 stdev 0003 0422 0 422 0003 0 002 0003 0003 0004 0'i1 RMS 0 003 0422 0 422 0003 0 002 0003 0 003 0004 0TO4

i-c
4A 4' 5-A i t
bit

orwi
00* Ou.E 00* 1) ' 9 0 0 >H PO'C 0058 til OC/ 01. '' 0.057

now
00*8 OOC 0058

o&c
0000 l)ll< 0041 0l>TC

onn?
OC" ,'

ono?
uf> i 0T43

rr3
0050 OL'* 0058 J (.'In 00'.? 3 0W 0.057

L i r n . l l 01 hoi wtli C A I I J r-1 f u 1 r .Jl 'julh'iiDudi i l J'.rt

on. eiw
OOLI N1 -it N-NA-

t 7-B F \ RL>
VIA

nr?
0007 t \ N N.% -NA-

Direct Control with Ded catcd Duct Injection Fan Injection Fan Injection Fan Injection Fan

own
NA

N" N.t -NA-

o"ir
-0.992

9-B

E.2

Temperature Proftles

Table 19 contains information showing the status of the air temperatures during testing. See Section 5.1.2 for additional information regarding the analysis. Case numbers are the same as those listed in Table 12.

99

Table 19: Supply and outside temperature data for systems tested.
Supply Air Temperature (set point =65 f)
System Description Fixed Damper Position Fixed Damper Position Fixed Damper Position Plenum Pressure Control Pkii ni Pre u r o Cr n(ii>
r n i l C onhm * tli t MHIIIII zif T I I 1

Space Outlet Temp. (set point = 70 F)

mean 7025 70.26 70 22 7018 7T 28 / 0 30 t S stdev 0341 0.283 0 313 0.271 0 ,1*3 0 <<0 rj^' f 0 7* 0 ?"i7 tii 0 -TV, 0 358 0 347 0 412 0 257 0 PA 0 1W 0 704 0.428 RMS 0426 0384 03B4 0327 Otju Olio 0 J"i

Outside Air Temperature (variable setpoint)

RMS 0 690 1205 0 610 0663 1 V I 0TC8 CG43 0/0?
U >JU

Case 1-A 1-B 1C 2A

mean 5485 S524 54 84 54 05

stdev 0480 0648 0 511 0375

RMS 0503 0689 0534 0 378 " 0 421 0451 0 fc 1 0645 0-2R 0 70 J OPi 0644 1210 0 454 1 ifi

^r
JA

ss 01
15 10. Sill 54 67 5 r i"
r

n :^i
0 410 1 14/ 0 633 0T'1 0 503 CWi 0UJ8 l bJJ 0Jd2 1704 0 610 P 141 0.516

I) H i ' Cciiiiml ft th 1 txioir i 1 r t 1 i f - l Crnlio1 with LOII nn / * Pu t D i t . d ( Y n i i 4 Willi Ft^>mii^L V r j i n i I n Vini Drcct U x m with Dni ca i l l Lu 1 diJ

(.
4A *l. >/ GA GI> /A 7 EI R >

70/3 70,1 10 7C27 70 15 70 40 70 21 'CZI 70 14 /H 1(1 70.14

o n
"3..U 0 477 0 418 0178 0 5/b 0 33/ 0-/7 0411 0/7/ 0.451

>5 17

0r.w 1 7r8 0 W, i)4J3 1431 i f a P10J J U1 1.850

< t 93 W 54<i| ' 5 9-j 55 71 jABP 54 RB M 14 54.89

D net Ten n i rln Di-d o i i-d 1 i t L1 red Cnr* nl wlh l)i-d o. i-d 1 ik' r n rt Cun'ml wlh Di-d l a ' c l I n 1 1 ihtfi Tin

li" ITt 0 1 1 1 1 1
Ir I T I I U i f n

'III

0b2? 0'-4H 0326

1 \
9-B

Injection Fan

100