1、Brad Acker is Researcher, Kevin Van Den Wymelenberg is a director and assistant professor at the University of Idaho, Integrated Design Lab, Boise, Idaho Demand Control Ventilation: Lessons from the Field- How to Avoid Common ProblemsBrad Acker, PE Kevin Van Den Wymelenberg Member ASHRAE Member ASHR
2、AE Abstract Demand control ventilation (DCV) has the potential to save energy by reducing ventilation rates in accordance with occupancy levels provided by the surrogate indication of CO2levels. However, improperly installed, designed, or operated systems may save energy at the expense of Indoor Air
3、 Quality (IAQ) or enhance IAQ at the expense of energy. These outcomes may have the potential to foul the image of an otherwise viable energy efficiency measure. This paper reports what the authors believe to be common problems in the design, installation and operation of DCV systems which use CO2as
4、 a surrogate for occupancy levels. Six HVAC systems were investigated: two commercial offices, two medical offices, and two school environments. The design drawings, air balance reports, and current equipment set up were investigated. Four systems were controlled locally through roof top unit contro
5、l logic and two systems were controlled by central building energy management systems. Functional testing of equipment was carried out and system parameters were logged including CO2levels, fans states, and air stream temperatures. Functional testing was broken up into three system aspects. First, C
6、O2control signal functional testing was conducted to confirm that the control link between CO2sensors and outside air damper positioning was in place. Second, sensor placement functional testing was conducted to confirm that the sensors placement could accurately report the CO2levels of the controll
7、ed zone. Third, the Outside Air (OSA) level test was conducted by inspecting the air balance reports to determine the OSA rates and to confirm that the system was balanced in accordance with DCV standards. The study found that no systems were functioning properly for a number of reasons, some of whi
8、ch were overlapping. Reasons for non-functionality included poor sensor placement, improper information provided in mechanical schedules or design documents, fan cycling issues, and poor installation. Details on failure modes will be presented. Proper engineering documentation requirements will be e
9、xplained. Test, Adjust, Balance (TAB) specifications and DCV specific requirements for TAB along with information that building operators need to know about system operation will be presented. LV-11-C062502 ASHRAE Transactions2011. American Society of Heating, Refrigerating and Air-Conditioning Engi
10、neers, Inc. (www.ashrae.org). Published in ASHRAE Transactions, Volume 117, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAES prior written permission.INTRODUCTION Demand control ventilation (DCV) is
11、 a building ventilation control strategy in which the quantity of mechanically supplied outdoor air intake is regulated by some type of occupant density sensing. DCV is intended to save energy by means of supplying design ventilation air t o occupants during periods of h igh occupancy and supplying
12、the minimum required ventilation to dilute building related c ontaminants during low occupancy periods. Reducing the amount of outdoor air that needs to be conditioned can save energy. If air side economizers are in use specific outdoor conditions will have an effect on the amount of energy savings.
13、 Carbon dioxide (CO2) sensors are the industry standa rd to determined space occupancy for DCV. It has been shown that CO2levels are a good determination of space occupancy (Turpin, 2001). It should be noted that CO2is not considered an indoor air quality (IAQ) concern at levels found in typical bui
14、ldings (400-2000ppm) but is used solely as an indication of occupancy level (Emmerich, 2001). DCV systems can be incorporated into existing HVAC equipment and often times operate in conjunction with existing economizer controls, sharing the same outdoor air damper (OAD). Savings from DCV systems are
15、 achieved by the reduction in outdoor supply air (OSA) that requires conditioning as compared to a fixed OSA flow rate during all occupied hours. FIELD RESEARCH Six spaces were randomly chosen from a list of spaces provided by the funding agency, which supplied incentives for the use of DCV. All stu
16、dy spaces were in ASHRAE climate zone 5B. Table 1 below shows the description of the study spaces. Two control types were encountered in the study and listed in Table 1. The control types were roof top unit (RTU) or control with a building energy management system (EMS). DCV is intended for spaces w
17、hich have variable occupancy rates, this was the case in the spaces for building codes 04-09 but buildings 01 and 10 had incorrectly located CO2 sensors which well be ad dressed below. All buildings were fully occupied except for building 10 which was estimated by building management to be 70% lease
18、d out. Specific occupancy patterns within spaces was not studied. Table 1 Study Space Functional testing was broken up into three system aspects. First, CO2 control signal functional testing was conducted to confirm that the control link between CO2 sensors and OA damper positioning was in place. Se
19、cond, se nsor placement functional testing was conducted to confirm that the sensors accurately reported the CO2levels of the controlled zone. Third, the OSA level test was conducted by inspecting the air balance reports to determine the OSA rates and to confirm that the system was balanced in accor
20、dance with DCV standards such as presented in ASHRAE 62.1-2007. Only when DCV systems passed all three functional tests could energy savings estimates be carried out. Building Code Year Installed Control Type Building Type Building Size, SF (m2) Study Space Description Study Space Size SF (m2) 01 2
21、006 EMS Office 68,000 (6317) One third of entire building 22,600 (2099) 04 2008 RTU Office/Medical 21,104 (1961) Break room 300 (28) 06 2 007 BAS Elementary School 63,400(5890) C lassroom 1,000 (93) 08 2007 RTU High School 65,000 (6039) Classroom 800 (74) 09 2 007 RTU High School 102,000 (9476) Clas
22、sroom 78 0 (72) 10 2 008 RTU Office 15,750 (1463) Half of entire building 7,875 (732) 2011 ASHRAE 503Functional testing of DCV systems was done in accordance with the basic control type and differed whether it was managed through an energy management system (EMS) or at the roof top unit (RTU). EMS b
23、ased control systems were tested by manually lowering the CO2setpoints to be below the current levels within a space and watching (on the EMS screen) for the OA damper to open in that space. RTU base d control systems were tested by havi ng one person exhale directly onto the CO2sensor and having an
24、other person located at the RTU watch for the damper to respond and for the DCV indicator light to illuminate (if available) on the logic controller. In addition to physical functional testing, trendlog data and other direct measurement data were analyzed. Care was taken when analyzing data to find
25、periods during which only the DCV signals were dictating the OA damper position. Periods of economizer use were avoided. Logged data included air stream temperatures of the outdoor air, m ixed air, return air, supply fan state (on/off), fan Amperage, CO2sensor signals, compressor state (on/off), and
26、 compressor Amperage. In this field study, no systems were found to pass the functional tests. Failure modes will be explained below. FAILURE MODES System Design All systems studied failed the functional test related to air balancing and DCV standards for OSA rates. From working with local design an
27、d TAB professionals the authors believe that a good design specification schedule of a single zone DCV system will include minimum outside air flow rate, the zone CO2concentration at that flow rate, the zone maximum outside air flow rate, and the corresponding CO2concentration at the maximum flow ra
28、te. All these values are easily calculated using ASHRAE 62.1-2007. If these exact values are not given other values could be supplied to inform the TAB contractor of these same four setting which are minimum airflow, control signal at minimum, maximum airflow, and control signal at maximum. Table 2
29、below shows the given system specifications for each system studied. It can be seen from this table that none of the design engineers specified all the information that a TAB professional would need to install and balance a functional DCV system. All drawings noted the use of CO2 sensors in the desi
30、gn. Many designs provided notes which described in general terms outside air dampers opening and closing from minimum to maximum positions with regard to zone population or CO2levels. No actual CO2concentration levels, control signal values or outside air flow rates were given as guidance to TAB con
31、tractors for the proper balancing of a DCV system. Some interviews with TAB contractors revealed that in some cases minimum OSA levels would be set to zero or some other arbitrary flow which could lead to building pressurization issues if they did not consider exhaust and exfiltration airflows. Tabl
32、e 2 System specification per design documents Building Code System Total cfm (lps) Min. OSA cfm (lps) 01 28,000 (13,215) 4,000 (1,888) 04 1,500 (708) 300 (142) 04 1,500 (708) 300 (142) 08 2,400 (1,133) 690 (326) 09 6,000 (2,832) 1,500 (708) 10 8,500 (4,012) 1,300 (614) System Installation Table 3 be
33、low shows results of the functional testing related to system installation. One RTU failed CO2control signal functional testing. The exact reason for this failure could not be determined but it is suspected that it was caused by a wiring problem given that the physical output from the CO2sensor did
34、not signal damper movement despite functional performance 504 ASHRAE Transactionsof both the sensor and the damper individually. This failure is assumed to be a workmanship issue and not a systemic problem. Two units failed the sensor placement functional test. This issue is considered systemic and
35、needs to be addressed by the design community. These units had CO2 sensors placed in c ommon mixed ai r returns from multiple zones. This configuration disallows control of the critical zone. Similar errors have been identified in previous research (Carrier, 2001. Emmerich, 2001). CO2concentrations
36、for sensors placed in this manor tend to be below typical control levels due to this zonal averaging. Figure 1 below illustrates an example from Building 10 where the CO2 sensor was located in a common return. It s hows CO2levels rising only 100 ppm above ambient levels. This finding not only shows
37、the senor is in an incorrect location, but it also shows that the building is being over ventilated due to the fact that return air CO2levels are so close to ambient levels. It should also be noted that typical sensor accuracy is +/- 50 ppm. Table 3 System Installation Test Code Control Type Sensor
38、Placement Control Signal 01 EMS Fail Pass 04 RTU Pass Pass 04 RTU Pass Fail 08 RTU Pass Pass 09 RTU Pass Pass 10 RTU Fail Pass Figure 1 Common Return Sensor Location, CO2levels Ambient levels = 450ppm 2011 ASHRAE 505System Operation Intermittent fan operation was noted at two of the sites (both high
39、 schools). Intermittent fan operation is a problem in many aspects of HVAC control systems and greatly affects DCV performance. Fan cycling was found to be a problem in 38% of systems studied as part of a RTU field study (Jacob, 2003). With intermittent fan operation, OSA dampers only function when
40、the supply fan is running, otherwise they move to a fully closed position. When supply fans are set to run only during calls for cooling or heating, CO2 levels go unchecked when the fan is not on. This result can be seen in Figure 2 below. Fan cycling was also raised as an issue that affects the per
41、formance of DCV systems. Fan cycling is perceived to be an energy savings measure by building operators. This is an incorrect perception that requires education to overcome. Operators need to understand that fan cycling reduces indoor air quality, occupant comfort, diminishes accuracy of sensor sign
42、als, and reduces DCV functionality. Table 4 System Fan Operation Code C ontrol Type Fan Operation 01 EM S Continuous 04 RTU Continuous 04 RTU Continuous 08 RTU Intermittent 09 RTU Intermittent 10 R TU Continuous Figure 2 Fan Cycling 00.20.40.60.811.2050010001500200025007:12 8:24 9:36 10:48 12:00 13:
43、12 14:24 15:36SupplyFan (On=1,Off=0)CO2(ppm)Building 08-High School,- CO2 and Supply Fan RelationshipCO2 Supply Fan State506 ASHRAE TransactionsDESIGN CONSIDERATIONS Best practices for the design and installation of DCV systems are constantly evolving. In addition DCV systems interact with other sys
44、tems and setpoints such as econ omizers and b uilding pressurization setpoints. Designers must be knowledgeable of interactions and take them into account when specifying a DCV Systems. As stated above, from working with local design and TAB professionals the authors believe that a good design speci
45、fication schedule of a single zone DCV system will include minimum outside air flow rate, the zone CO2concentration at that flow rate, the zone maximum outside air flow rate, and the corresponding CO2concentration at the maximum flow rate. The reason the authors believe that CO2levels should be spec
46、ified is because straight forward equations are available in ASHRAE 62.1-2007 and it is not dependent on the sensor range or control signal type which may not be under the control of the HVAC designer. It could be argued that the use of two CO2sensors, one measuring ambient levels and one in the con
47、trolled zone, will provide for better control of ventilation systems due to the fact th at CO2 levels changed with the seasons or by other factors such as pollution levels. Device accuracy also needs to be considered, many devices currently on the m arket have an acc uracy in the range of 50-100ppm
48、and only some can be recal ibrated in the field. These fact ors need to be considered on a case by case basis. In general CO2 sensor are less expensive and more reliable than they were 10 years ago but issues still exist and sensors should be evaluated based on the requirements of the application (M
49、axwell, 2009). CONCLUSIONS This study ha s exposed several issues with regard to the d esign, installation, and operation of DCV systems that prohibited the association of measured energy savings with the systems monitored. While not looked at in this study, it is important to understand that DCV can sometimes increase energy consumption during cooling periods when not used in conjunction with OSA economizer functionality. All of the buildings in this study were built under ASHRAE 62-1989. While this standard needs to be considered when asking why systems were installed as they we
