1、Reid Hart, PE is Associate Director in the Technical Research Group at Portland Energy Conservation Inc. (PECI), Portland, Oregon. Jack Callahan, PE is Emerging Technology Program Manager at Bonneville Power Administration, Portland, Oregon. Kenneth Anderson and Patrick Johanning are Engineers at PE
2、CI, Portland, Oregon Unitary HVAC Premium Ventilation Upgrade Reid Hart , PE Jack M Callahan, PE Kenneth Anderson Assoc. Member ASHRAE Assoc. Member ASHRAE Assoc. Member ASHRAE Patrick Johanning Assoc. Member ASHRAE Abstract Rooftop packaged HVAC units (RTUs) serve over 40% of commercial building sp
3、ace with relatively reliable air conditioner and furnace operation; however, improper settings and failed controlsespecially economizersnegatively impact indoor air quality and energy use. A field survey of over 300 packaged units found 91% with at least one problem and 64% with two or more problems
4、. To date, utility HVAC programs, codes, and green standards have focused on higher efficiency units or unit tune-ups to the exclusion of ventilation and control upgrade opportunities. Hourly simulation over a range of U.S. climates shows that a comprehensive package of RTU control retrofits produce
5、d HVAC savings between 30% and 48%ten times the savings from incrementally higher efficiency unit replacement. The proposed RTU control retrofits make economizers more reliable, improve ventilation control, and reduce energy use. The retrofits include stand-alone direct digital controls, demand cont
6、rolled ventilation with integrated fan cycling or fan speed control, premium economizer control, and ventilation lockout during warm up. Field testing showed that for RTUs, a single return air carbon dioxide sensor was as effective as multiple-room sensors in allowing demand controlled ventilation t
7、o improve indoor air quality. Field testing of four RTUs with variable speed fans and solid-state DCV controllers was extended to include six additional units that are using DDC controls with fan cycling rather than fan speed reduction. The fan cycling control algorithms are demonstrated to comply w
8、ith ASHRAE Standard 62.1. INTRODUCTION: PROBLEMS UP ON THE ROOF Fortunately unitary HVAC systems, typically rooftop units (RTUs), run quite reliably for many years with minimal maintenance. RTUs smaller than 15-tons typically do not receive testing and balancing when installed and service contracts
9、rarely include more than filter changes. Because comfort is generally maintained, owners assume RTUs must be working properly. However, multiple field studies have found many performance problems that result in poor ventilation and excess energy use. Based on a compilation of four RTU field studies
10、covering more than 500 units in the Western U.S. (Cowan 2004), Figure 1 shows that more than 60% of the outside air economizersmany recently installeddid not work effectively. Refrigerant charge was incorrect for more than 40% of RTUs. Most units had fans either on when not needed or off when they s
11、hould be on. A major problem was found with fan operation. While some errors were fixed with a simple thermostat adjustment, the more profound problem was air flow well below manufacturers minimum recommendation. The combination of high fan power and low air-flow was due largely to excessive pressur
12、e drop in the duct systems. A frequency distribution of field tested external static pressure is shown in Figure 2 (Jacobs 2003). The average system external pressure drop of 0.48 inches WC (120 Pa) was more than double the AHRI Standard 210/240 test point of 0.20 inches WC (50 Pa) for median-sized
13、units. LV-11-C064 2011 ASHRAE 5172011. American Society of Heating, Refrigerating and Air-Conditioning Engineers, 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
14、 is not permitted without ASHRAES prior written permission.This data from the rooftop means that fan energy use and potential fan energy savings are much greater than unit ratings such as EER and SEER indicate. A common theme from the field studies is that RTUs have multiple in situ control problems
15、 and energy waste. Figure 1. Unitary HVAC Problems Figure 2. Actual Field Static Pressure PREMIUM VENTILATION PACKAGE IMPROVEMENTS The “premium ventilation package” grew out of a premium economizer utility program in Oregon (Hart et al. 2006). It is a comprehensive controls upgrade for existing RTUs
16、 that include multiple measures that have synergies. Broad application of the package approach in a utility retrofit program reduces customized audit costs. Low administrative costs per unit is important to maintain cost effectiveness when per RTU savings small. Items like optimum start thermostats,
17、 economizer controls, and demand controlled ventilation can be independently prescribed, but a package of measures provides the synergy of addressing common problems together and building installer expertise and efficiency. The measures include: Optimum start. Most programmable thermostats have an o
18、ptimum start option that slowly increases or reduces the setpoint rather than moving immediately to the occupied setpoint. Delaying heating or cooling until needed saves energy. Resistance heat lockout. Resistance heat is locked out when it is warm enough for the heat pump to meet load. Ventilation
19、lockout during morning warm-up. HVAC units typically start 2 to 3 hours before occupancy with full ventilation provided, using unnecessary heating. A separate thermostat relay closes the ventilation dampers during warm-up. Outside air damper seals. When closed, outside air dampers for RTUs have leak
20、age of 5% to 25%. New low-leakage dampers can be installed or adhesive-backed closed-cell insulation foam can be added to existing dampers. Outside air economizer. Economizers typically have low quality (snap disc) outside air temperature sensors and are operated with simple dry bulb change over in
21、the Western U.S. and enthalpy sensors in the East. Premium ventilation uses an integrated economizer with differential temperature changeover control (Hart et al. 2006). Based on recent testing of solid-state economizer controllers (Robison et al. 2008), an improved outside air sensor with an aggres
22、sive setting (at least 68F) can achieve savings close to the differential type of changeover. Demand controlled ventilation. Demand controlled ventilation (DCV) has traditionally been applied to larger RTUs serving a zone with dense and variable occupancy. RTUs with a properly operating economizer l
23、imit the benefit of DCV in zones with consistent occupancy assuming the RTU was installed with proper system testing, adjusting, and balancing 46%64%42%58%27%0% 20% 40% 60%Refrigerant CircuitEconomizerAir FlowThermostatSensors% of units tested with problem0%2%4%6%8%10%12%14%0.05 0.15 0.25 0.35 0.45
24、0.55 0.65 0.75 0.85Unit external static pressureFrequencyMedian Unit ARI Test PointPa 12 62 112 162 212In. WG518 ASHRAE Transactions(TAB). Unfortunately, RTUs do not normally receive proper TAB so ventilation is often significantly higher than required (Davis et. al. 2002). RTUs that have excess min
25、imum ventilation can benefit from a DCV system even without occupancy swings. DCV adjusts ventilation to meet actual load, which is usually less than design. In addition DCV provides ventilation reduction during morning warm-up. Installation requires a DCV controller and a CO2 sensor. Premium Ventil
26、ation Package with Fan Cycling Based on field testing (Hart 2009) there were difficulties configuring solid-state economizer/DCV controllers and matching low-cost Variable Speed Drives (VSDs) for some motor types, therefore an alternative approach is currently being field tested. This alternative in
27、corporates a DDC controller combined with a programmable thermostat. By integrating economizer and unit control, the DDC option has the ability to run several other energy saving modes and improve operation of the modes already discussed. Control accuracy is enhanced by using digital setpoint entrie
28、s rather than setpoint screws on the solid state economizer controller. Figure 3 shows how setpoints and ventilation vary throughout a typical summer day as the premium ventilation with fan cycling sequence moves through several additional modes, including: Night Flush. During the cooling season, th
29、e economizer is used in the early morning hours to pre-cool the space. Stand-Alone Demand Response. Demand response based on outside conditions is achieved by adjusting the cooling setpoint up during peak afternoon periods. An advanced version can also pre-cool the building before the peak period. O
30、ccupancy Sensor Standby Mode. When the space is empty, setpoints are relaxed and ventilation is suspended.1Advanced Optimum Start. Extends the start period when outside temperature indicates conditions exceed design. Fan Cycling DCV. Using fan cycling rather than a VSD provides energy savings from t
31、he fan motor being turned off instead of low speed VFD operation. Pre-purge cycles can enhance the ventilation over simple analog control. Summer Typical Mode Setpoints4550556065707580859012:00AM2:00AM4:00AM6:00AM8:00AM10:00AM12:00PM2:00PM4:00PM6:00PM8:00PM10:00PM12:00AMSpace Temperature /Setpoint0%
32、20%40%60%80%100%120%140%160%180%OSADamperOA DamperSpace TempCool SetpointEcono SetpointHeat SetpointUnoccupiedNightFlushOptimum Start(Purge)OccupiedStandbyOccupiedPre-DemandDemandRecoveryUnoccupiedC F3224167Figure 3: Premium Ventilation Typical Sequence Modes and Setpoints 1Section 5.3 of Standard 6
33、2.1-2010 requires the ventilation fan system to be enabled only when the spaces served are actually occupied as further described on page 5-11 and 6-41 and example A-A of the 62.1-2007 Users Manual(ASHRAE 2007). 2011 ASHRAE 519Field Testing the Premium Ventilation Package The results in this paper a
34、re based on field testing of four RTUs with variable speed fans and solid-state DCV controllers (Hart 2009). Testing is currently underway of an additional six RTUs using stand-alone DDC controls from three manufacturers with the additional control modes described above and fan cycling rather than f
35、an speed reduction. Future testing is anticipated for a DDC version with fan VSD for situations where continuous fan operation is desired. More advanced demand response strategies and fault diagnostics can be incorporated as well. PREMIUM VENTILATION SAVINGS POTENTIAL DOE 2.2 was used to simulate pe
36、rformance for eight cities covering a range of U.S. climate zones.2Energy savings for the Premium Ventilation Package ranged regionally from 18% to 44% of HVAC use, comparing favorably with a 1.5% to 6.7% savings from upgrading RTUs from SEER 13 to 15. The analysis demonstrates a large potential for
37、 package savings. Once the fan cycling DDC field test is complete, savings for that revised package will be analyzed using an expected value approach (Hart 2008). Baseline conditions are based on field studies of the typical as-found condition of smaller RTUs. For example, baseline ventilation rates
38、 were based on the field observation that minimum outside air is 13% of supply rather than the code required 7% (Davis et al. 2002). To estimate savings, a simulation was developed of heat pump RTUs on a small office building using the VSD version of the Premium Ventilation Package with solid-state
39、economizer/DCV controls. The total height of bars in Figure 4 represents baseline office RTU HVAC energy use in eight varied United States climate regions. Interactive energy conservation measure (ECM) savings from the analyses are shown in the sections at the bottom of the bars. The remaining HVAC
40、energy use after all measures are completed is shown in the top portion of the bars. Figure 4: Rooftop Unit Savings in Representative Climates. The premium ventilation package results in a range of regional savings that is 5 to 25 times the savings of an upgrade from SEER 13 to 15. 2The baseline bui
41、lding for savings analysis is a 20,000 square foot two-story office building using primarily the Title 24 eQuest defaults with an increase in unoccupied lighting and equipment loads to reflect reality and higher than required ventilation (31 cfm/person or 13%) to partially reflect field observation
42、of typical ventilation minimums at 20%. Packaged single-zone units with a SEER rating of 13.0 were simulated. Heat pump systems on a typical small office building were analyzed so results would all be electric for easy comparison. Gas heating results and detailed basis for analysis is included in Ap
43、pendix B of the original savings study (Hart 2008). 02,0004,0006,0008,00010,00012,000PhoenixAZSactoCAEugene ORBoise IDBurlton VTChicagoILMemphisTNHoustonTXHVACkWhper1000 sq.ft.or93sq.mBase Total HVACTotal ECM SavingsOptimum StartStrip Heat lockoutWarmup CycleIntegrated EconomizerDCVVSD fanRemaining
44、HeatingRemaining CoolingRemaining Fan however, standard RTU DCV controllers typically operate with a single CO2sensor. For the large VAV systems that have relatively small zones with high occupancies, there is no question that a return air CO2sensor is an inadequate indicator of critical zone occupa
45、ncy. RTUs typically serve one or at most a few rooms. In the field test at the senior center, the metered RTU served two rooms that have moderately high and varying occupant density. Figure 7 shows readings of three monitoring CO2sensors: one in the return air, one in the billiard room, and one in t
46、he computer lab. The data was selected for a time period with the biggest room-to-room CO2concentration differential. Both rooms have CO2readings that never exceed the threshold level of 1,100 ppm indicating adequate ventilation. In this situation the maximum difference between the CO2concentration
47、in the more fully occupied room and the return air is never more than 150 ppm. Figure 7. Ventilation for AC-1 Diverges in Two Rooms Served due to Imbalanced Occupancy 4006008001,0001,2001:00PM2:00PM3:00PM4:00PMMon, Jun 15, 2009PPMCarbon DioxideCO2 Return A ir CO2 Computer Lab CO2 Billiard Rm CO2 Tar
48、get522 ASHRAE TransactionsField data shows that effective control for the subject RTU was achieved with a single CO2sensor in the return air streameven when there was an imbalance in occupant density between rooms served. Based on results in this study, RTUs serving up to a few rooms have difference
49、s in CO2concentration that are much less drastic than conditions experienced in large VAV systems where the critical zones are small relative to the total area served. Acceptable air quality can be maintained for most RTUs with a single return air CO2sensor set below the target threshold based on a multiple space analysis. A simpler and more reliable approach may be to use a supply air CO2sensor (Warden 2004) (Nassif et al. 2005); however, a supply air sensing approach is not compatible with threshold-type DCV contr
