1、Field Performance of Demand Control Ventilation in VAV Systems Scott Hackel, PE Saranya Gunasingh Member ASHRAE Associate Member ASHRAE ABSTRACT Demand control ventilation (DCV) systems use sensors to control ventilation air to a space or spaces based on the estimated number of people present. Thoug
2、h the technology has been around for some time and has multiple theoretical benefits, less is known about its performance in real systems, especially complex multizone ones. A field study was implemented to fill this knowledge gap. Study results show that DCV does save significant energy (primarily
3、the fuel used for heating), and is generally cost-effective for most owners. It also demonstrated improvement in systems through recommissioning, but showed that greater deficiencies are found in the design phase for this technology. Based on field observations, best practices for both design and co
4、mmissioning are also shared. INTRODUCTION Demand control ventilation (DCV) systems use sensors generally either CO2 or occupancy sensors to control ventilation air, or outside air (OA), to a space or spaces based on the estimated number of people present. The technology has the potential to save sig
5、nificant energy usage, especially in extreme climates like the upper Midwest where winters are cold and summers are humid. DCV has been in use for many years, and its theoretical impacts have been well established. But little is known about its performance in real systems, especially complex multizo
6、ne ones. A field study of such systems was initiated to fill that knowledge gap by 1) quantifying impacts of DCV implementation, 2) demonstrating improved system operation through commissioning, and 3) adding to DCV best practices. The study focused specifically on multizone, non-packaged systems be
7、cause they serve a large portion of DCV floorspace but are both less understood and substantially more complex than single zone rooftop units. First, information was gathered on a number of actual DCV systems installed in the upper Midwest. A subset of DCV systems was selected for more detailed meas
8、urement, to analyze performance. Following this initial period of measurement, the systems were recommissioned according to best practices in order to optimize performance. This two-step approach allowed for quantification of both the impact of DCV as well as the additional impact of system commissi
9、oning. Finally, lessons learned were collected from system designers, operators, and commissioning agents throughout all these steps. RESEARCH METHOD As part of the initial characterization, 32 buildings with 96 HVAC systems were identified, consisting of CO2-controlled, multizone DCV systems. DCV w
10、as characterized for each system through a combination of visiting these buildings or in-depth interviews and as-built documents. From there, 6 buildings were shortlisted for in-depth monitoring based on building automation system (BAS) data availability, ease of further monitoring, and verification
11、 of OR-13-C001available information. Utility bills were also collected to identify and avoid any outliers in terms of energy performance. In the end, five buildings with a standard variable air volume (VAV) system (non-unitary) were selected, and one building with a multizone VAV dedicated outdoor a
12、ir system (served by a ground source heat pump) was selected. Table 1 summarizes the buildings selected for monitoring. Table 1. Sample building set for monitoring. Name Building Type Age Owner Type System Type CO2 Sensor Location Design OA, cfm (m3/s) Office / Art Gallery Office / Assembly 12 Publi
13、c VAV Return + Occ. 6,240 (2.94) Library A Library 5 Public VAV Zone 3,503 (1.65) Library B Library / Office 11 Public VAV Return 4,235 (2.00) Performing Arts Ctr. Assembly 2 Owner VAV Return 3,500 (1.65) Office Office 2 Leased DOAS (VAV) Zone 2,933 (1.38 Higher Ed. Perf. Arts Education 4 Owner VAV
14、Zone 5,700 (2.69) Monitoring Following this characterization, monitoring process began with owner/ operator interviews, validation of existing BAS systems, and installation of monitoring equipment to provide data points not available through BAS. The monitoring utilized five minute data collection a
15、t both system and zone levels. At the system (central air handling unit) level, the following points were monitored: Outdoor air temperature, relative humidity and CO2 concentration Flow rates: outdoor air, supply air, return air Damper position: outside air, return air Temperatures: mixed air, retu
16、rn air, supply air (both upstream and downstream of the fan) Valve positions: hot water, chilled water Fan power: supply fan, return or exhaust fan Points monitored at the zone level included: Discharge air temperature VAV damper position and/or air flow rate Reheat valve position Zone temperature,
17、relative humidity, and CO2 concentration Occupancy sensor status, if available In addition to data monitoring, an occupant survey was completed based on ASHRAE Standard 55-2013 to assess overall comfort and whether occupants were satisfied with temperature, humidity and air quality in these systems
18、with DCV. Recommissioning The second phase of data collection included a system recommissioning. Recommissioning was executed based on industry best practices (see Commissioning Best Practices section). Our recommissioning process was aimed at optimizing system control sequences and setpoints, CO2 s
19、ensor design, airflow measurement, and handover to operators. The effort included the following: Virtual performance pre-checks focused on OA flow including all parameters shown in Figure 1, as well as CO2 concentrations Staff interviews for understanding systems adjustments, design intent, controls
20、 sequencing Validation of BAS measurements for air temperature, valve these were mostly accurate, though some corrections were required, especially for some air flow rate points. Air flow rate points were confirmed or adjusted based on measurement from a certified balancing professional. Once data w
21、as collected, it was further checked using: energy balances on the AHU; flow balances between outside, supply and return flows; recreation of fan curves; and correlations between dependent variables such as energy flow and valve position, and airflow and damper position. In addition to the above dat
22、a was checked visually, for reasonableness, including daily and weekly snapshots of temperatures, airflows, damper positions, fan operation, and valve position. The analysis of energy savings from measured results is based on an energy balance of the air handling unit as shown in equation 1: + + + =
23、 ( 1 ) Where is the energy contained in the supply air, is the energy in outside air is the energy in the return air stream, is the power consumption of the fan, is the heat transfer to or from the coils. Each variable is the product of the mass flow rate and the specific enthalpy of a specific air
24、stream. This equation is solved for coils, first based on the measured results and then based on a theoretical case with zero OA (air,outside=0). The difference in the two values of coils is the energy saved. This was repeated for data collected both before and after recommissioning. As a final step
25、, savings results were weather normalized based on a typical meteorological year (TMY) weather for Minneapolis, MN. Significant additional detail on this calculation method are given in Hackel 2015. 2016 ASHRAE Winter ConferencePapers 3MODELING Detailed energy models were built and calibrated to val
26、idate energy savings calculated analytically and to quantify scenarios that were not directly measureable - like alternate operating scenarios, the impact of future code, and a commissioned scenario for one of the buildings. EnergyPlus was chosen due to its advanced air side modeling capabilities an
27、d availability of hourly input/ output data. Certain EnergyPlus objects were instrumental in executing proper DCV operation in the model. The Controller:MechanicalVentilation was used to implement multizone DCV; this object works with the outdoor air objects to determine OA control when DCV is on. T
28、he System Outdoor Air Method was set to InddorAirQualityProceedure to capture DCV operation based on CO2. Other significant model inputs included specifying mechanical ventilation controllers in Controller:OutdoorAir objects and specifying the controlled zone name and CO2 setpoint schedule in ZoneCo
29、ntrol:ContaminantController object. FIELD RESULTS A large variety of approaches were encountered in the initial broad observations of DCV in Minnesota buildings. There were some commonalities to many of the systems. Over 80% of them had economizers, which is worth noting because any resulting saving
30、s must occur outside of economizer mode, leading to much fewer run-hours for the DCV sequence in this northern climate. Also 60% of systems were less than five years old and 82% were less than 10 years old; DCV is still a fairly new approach in the state. This is likely driven by recent code require
31、ments for DCV, and increased emphasis on energy savings in general. A lack of preventative maintenance was also pervasive in the private buildings surveyed only 11% of the CO2 controls had regular preventative maintenance (public buildings fared better). A wider variety was found in system sequence/
32、configuration; there was no clear one or two best practice sequences. 19% of systems used a return sensor to directly control OA damper. 26% used a zone sensor to directly control the OA damper. 37% of systems used zone sensors along with a ventilation reset strategy (considering all VAV box positio
33、ns see Murphy 2004), though with varying approaches. And 19% of the systems used a large number of zone sensors (usually occupancy) to vary VAV minimums AND implement a ventilation reset. Energy Savings After monitoring and analyzing a representative sample of the systems observed, several quantitat
34、ive results were compiled. The median annual savings of the systems monitored was $0.09 per square foot ($0.97/m2) of area served, which averaged 34% of air handling unit energy consumption. But energy savings from DCV was most directly correlated to the magnitude of design OA. When normalizing for
35、this metric the median savings equates to $0.50/cfm of OA ($1060 $-s/m3). The breakdown of savings by building and fuel is given in Table 2. Its worth noting that a large majority (80%) of the savings is from natural gas (and therefore sensitive to its price), due to the heavily heating dominated cl
36、imate of the upper Midwest, where heat most often is fueled by with natural gas. Table 2. Energy savings from DCV normalized by design OA flow rate. Savings per Design OA Flow Gas, therms/cfm (GJ-s/m3) Elec., kWh/cfm (kWh-s/m3) Total Cost, $/cfm ($-s/m3) Office / Art Gallery 1.38 (310) 1.68 (3560) 1
37、.14 (2420) Library B 0.96 (210) 0.08 (170) 0.69 (1460) Library A 0.73 (160) 0.21 (450) 0.54 (1140) Performing Arts Center 0.53 (120) 0.82 (1740) 0.45 (950) Office 0.0 (0) 4.01 (8490) 0.39 (830) Higher Ed Performing Arts 0.34 (80) 1.10 (2330) 0.35 (740) Median 0.63 (140) 0.96 (2033) 0.50 (1059) The e
38、conomic impacts of energy savings were also considered. Based on life cycle cost analysis it was 2016 ASHRAE Winter ConferencePapers 4determined that owners could afford to spend up to $7,000 per 1000 cfm ($14,800 per m3/s), the cost at which the owner would break even) on a system controlling the m
39、ajority of their OA, and $1,700 per 1000 cfm ($3,600 per m3/s) on a system controlling only a portion of that air such as that which goes to heavily occupied zones like conference rooms. More detail is given in Table 3. Table 3. Economics of DCV for 1000 cfm of OA CO2 and Occupancy Control Typical C
40、O2 Control Typical Partial CO2 Control Break-even cost $16,412 $6,658 $1,643 Simple payback not available 4-5 years 7-8 years After measuring each of the six systems across all seasons in their as-found state, the frequency of deficiencies as well as the corresponding potential for additional saving
41、s were tested by recommissioning each system. Systems were recommissioned according to established commissioning practices for DCV (see EDR 2007, Dougan 2004, and CEC 2006, as well as best practices in the next section). The common deficiencies are shown in Figure 2. Figure 2. Deficiencies found in
42、recommissioning the six monitored systems. Recommissioning had a mix of impacts. Three systems yielded a significant increase in savings, ranging from almost negligible to greater than 80%. Two systems saw no change. And the savings for the sixth system (Library A) were negligible due to a change in
43、 CO2 setpoint required to correct inadequate indoor air quality (IAQ) at certain times. Per unit of OA flow, the median savings after recommissioning was then $0.63/cfm ($1,300 per m3/s. Break-even costs for the savings increase just due to recommissioning were $2,900 per 1000 cfm ($6,100 per m3/s),
44、 equating to a payback of about one year based on the costs incurred in the recommissioning process. Additional field results for all phases of the project can be found in (Hackel 2015). CONCLUSIONS AND BEST PRACTICES Perhaps the most important conclusion from this study is that DCV in large VAV sys
45、tems can cost-effectively save energy for building owners, operators, and tenants. Based on the results above, for every cfm of OA in a typical system DCV would save $0.50 each year. If a mechanical engineer is designing a large AHU for a new office with an estimated ventilation requirement of 5,500
46、 cfm (2.6 m3/s), based on this study they could estimate the expected savings to be about 5,500 cfm x $0.50/cfm = $2,750 per year. This does assume that a significant majority of the 5,500 cfm is E n e r g y s a v i n g sN o e n e r g y s a v i n g s ; i mp r o v e d IA QS tr i p i n g d e n o te s
47、i te m i d e n ti f i e d , b u t n o t a d o p te d b y o w n e r2016 ASHRAE Winter ConferencePapers 5controlled by CO2 sensors. Systems that use CO2 sensors in a fraction of the spaces only often just the heavily occupied conference and training rooms will save considerably less. And savings were
48、by far highest for the one system that used both occupancy sensors and CO2 sensors to modulate airflow in essentially every space. Other than the amount of OA controlled, savings scaled most primarily with how aggressive the sequence was in terms of the absolute lower limit for OA flow (the choice o
49、f this variable will be discussed in more detail below). One system with a near zero lower limit for OA flow reached savings of $0.69/cfm ($1,500 per m3/s), while a system that used a lower limit of about 2/3 of its design flowrate only reached savings of $0.35/cfm ($740 per m3/s). The results also show that DCV is relatively cost effective. Economic results show break-even costs between $2,000-$7,000 per 1000 cfm of OA ($
