1、2010 ASHRAE 147ABSTRACTThis paper describes the basis for an energy efficientrefrigerant-based thermal energy storage system design forcommercial DX systems. A description of a system is given, anddesign aspects contributing to energy efficiency are identified.Similarly, operating characteristics an
2、d strategies contribut-ing to energy efficiency are discussed. Challenges in measur-ing and predicting energy efficiency are identified, along witha proposed method for collecting certified performance data,and sample test results. Field data is then analyzed to deter-mine if it is consistent with e
3、nergy savings predictions. Thebenefits of using such energy efficient thermal energy storagesystems are then considered for several climate zones. Resultsshow that refrigerant-based thermal energy storage for com-mercial DX systems is energy neutral or better given specificdesign considerations and
4、operating strategy.INTRODUCTIONBackgroundHistorically, thermal energy storage (TES) has beenapplied to large chiller-based systems in the form of eitherchilled water or ice storage. The system analyzed in this manu-script is a packaged refrigerant-based ice-on-coil storagesubsystem, designed for use
5、 with Standard Direct Expansion(DX) equipment. “Standard DX Equipment” means unitary,direct expansion A/C equipment, including but not limited to,split, mini-split, packaged, and single package vertical unitsystems. As such, this design is termed a “Unitary ThermalStorage System,” or UTSS. It shares
6、 many of the same benefitsas the chiller-based systems, but is designed to work with stan-dard DX equipment in both new and retrofit applications, andin a manner that can improve the combined systems overallefficiency. As with the performance of standard DX equip-ment, specific climates and applicat
7、ion conditions will impactthe actual amount of net energy used. Measuring the perfor-mance of systems in the field is challenging due to the broadrange of operating conditions, indoor and outdoor environ-mental conditions, as well as the range of uncontrollable vari-ables present in the field, such
8、as building design, equipmentmaintenance, and allowable variations in manufactured OEMequipment itself. For this reason, performance ratings of stan-dard DX systems are lab based, and tested over a well-definedrange of controlled operating conditions. The lab data can thenbe used to evaluate the sys
9、tems operation over a broad rangeof conditions, both static and dynamic. This in turn can be usedto predict performance and efficiency in real-world applica-tions. As a final step, actual field data can then be analyzed tosee if it is consistent with predictions, and to provide indica-tions for furt
10、her investigation.Defining “Energy Neutral or Better”What is meant by “energy neutral or better?” Storagetechnologies cannot be 100% efficient. However, with theaddition of a UTSS system, a DX systems efficiency can beimproved. The improvements offset the UTSS storage losses,for a net improvement in
11、 overall efficiency of the hybridUTSS/DX system. This is what is meant by energy neutrality.In modeling the impact of UTSS, one must compare the sameDX system with and without storage. This comparison shouldbe conducted with a year-long (8760 h) analysis in order toassess the full impact of charge a
12、nd discharge cycles for theUTSS system. Using typical meteorological year (TMY) data,results show that the annual energy consumption of standardEnergy Efficient TES Designs for Commercial DX SystemsRobert Willis Brian ParsonnetAssociate Member ASHRAERobert Willis is a senior development engineer and
13、 Brian Parsonnet is CTO at Ice Energy, Inc., Windsor, CO. OR-10-016 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction, distribution, or transm
14、ission in either print or digital form is not permitted without ASHRAEs prior written permission. 148 ASHRAE Transactionscommercial DX systems is unchanged or slightly improvedwith the application of refrigerant-based UTSS, withinreasonable specified application guidelines and operatingstrategy. Whi
15、le use of an actual years data or data from anextreme year may be useful to demonstrate the range of behav-ior, TMY data is most appropriate for determining typicalenergy efficiency.SYSTEM DESCRIPTIONThe TES system discussed in this paper is a UTSS whichuses refrigerant for charging and discharging.
16、 The two maincomponents of the UTSS system are the storage section andthe charging module, which houses the refrigeration equip-ment for building ice. These components are connected by arefrigerant management system that contains key featuresresponsible for efficient operation. Charging module compo
17、-nents, including the compressor, expansion device, and con-denser fans, are active during the charge cycle. When theUTSS provides cooling, these components are inactive andrefrigerant is supplied to the evaporator with a small refriger-ant pump.The UTSS is always used in conjunction with a DXsystem
18、, to create a “hybrid” cooling solution as shown inFigure 1. The DX system provides the airflow for cooling fromeither system.SYSTEM OPERATIONThe UTSS has two main modes of operation - chargingand cooling. During a scheduled on-peak service time period,commonly Noon 6:00 P.M., all or part of the bui
19、lding load isserved by the UTSS via an evaporator coil inserted in the DXsair stream. Remaining load, if any, is served by the DX system.(A common configuration, for example, would be for 50% ofa 10-ton (35.2 kW) load to be served on-peak by the UTSS,and the other 50% by the DX.) At all other times,
20、 such asduring the UTSS charge cycle, the DX system provides all thecooling required.SYSTEM EFFICIENCYFactors leading to high efficiency can be categorized aseither inherent to the UTSS system itself, or emergent as aresult of integration to the DX system. Inherent factors includea refrigerant-based
21、 design, use of liquid overfeed for both icecharging and ice cooling, gravity feed, narrow approachtemperatures, and control optimizations. Emergent factorsinclude reduced cycling, improved dehumidification, avoid-ance of rooftop temperatures, reduced coil freeze-up, andcooling capacity that is deco
22、upled from daytime ambienttemperature, enabling right-sizing of the DX system.Refrigerant-Based DesignThe UTSS was designed to operate with refrigerant as theheat transfer fluid for both charge and discharge modes. Theparticular system discussed here uses R-410A. The use ofrefrigerant allows a key d
23、esign strategy contributing to effi-cient charge cycles. Using gravity alone, the refrigerant is effi-ciently fed into the storage section of the UTSS, without theuse of actively managed valves, pumps, or other controldevices. Condensed refrigerant in the refrigerant managementsystem forms a liquid
24、column which increases pressure at theinlet of the storage section and promotes refrigerant flow intothe ice storage heat exchanger. The bottom of the heatexchanger is flooded with liquid refrigerant and evenly distrib-uted throughout. As the refrigerant vaporizes, it rises in theheat exchanger, car
25、rying liquid with it. This mixed phaserefrigerant is carried to the top of the heat exchanger and backinto the refrigerant management system. Here the vapor andliquid are separated. The liquid is immediately sent back to thestorage section heat exchanger by gravity while the vapor iscondensed by the
26、 charging module. This design eliminates theneed to pump refrigerant through the heat exchanger, andavoids the associated power draw. In addition, since the refrig-erant is gravity fed, refrigerant velocities are low, minimizinglosses associated with pressure drop in forced flow situations.The simpl
27、e refrigerant management system design allowsmaximum flow and avoids the need for isolation valves and therestriction they would impose on the system. Ultimately, theonly active components during the charge cycle are thecompressor, expansion device, and condenser fansno differ-ent than the component
28、s of a typical condensing unit.Figure 1 “Hybrid” cooling solution combining a UTSS andpackaged DX system. 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional
29、 reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission. ASHRAE Transactions 149Liquid OverfeedAs mentioned above, the storage section operates withliquid overfeed. The refrigerant management system (RMS)design effectivel
30、y separates liquid and vapor, eliminating theneed to generate superheat in the storage section heatexchanger. The entire internal surface of the ice storage heatexchanger is wetted with liquid refrigerant, allowing utiliza-tion of the full surface area of the heat exchanger for buildingice. The resu
31、lt is a heat exchanger that operates with an aver-age approach temperature of 5F (2.8C) during the chargecycle. Minimizing the approach temperature reduces the effi-ciency degradation that results from excessively low compres-sor suction temperatures.The RMS use of liquid overfeed also extends to th
32、e cool-ing cycle, during which the buildings evaporator coil alsooperates more efficiently, without the need for a thermostaticexpansion valve (TXV), and without any entrained oil. TheUTSS cooling cycle capacity is not a function of ambienttemperature, and this becomes an important source of effi-ci
33、ency gains for the UTSS/DX system as a whole.Charging ConditionsUTSS designed for DX systems benefit from night/daytemperature swings. The nameplate energy efficiency of astandard refrigerant-based air conditioning unit degrades as afunction of rising temperature. As a general rule, DX coolingsystem
34、s using R-22 refrigerant experience energy efficiencydegradation of about 1.2% per degree (Faramarzi 2004) above95F (35C), equating to a 15% or greater nameplate energyefficiency loss at peak temperatures. R-410A (the refrigerantnow universally replacing R-22) exhibits a more aggressivedecay, estima
35、ted at about 1.6% per degree (Wells et al. 1999,Domanksi and Payne 2002). Conversely however, compressorefficiency improves as temperatures fall below 95F (35C)(Wells et al. 1999, Domanksi and Payne 2002). The UTSStakes full advantage of this swing. It keeps the DXs compres-sor from running when its
36、 least efficient (during the hottestpart of the day), and runs the charging module compressor tomake ice when its most efficient (in the cool of the night).Both of these factors contribute to improved overall hybridsystem efficiency.The UTSS control strategy for the charge mode alsoincludes a delaye
37、d start time for charging. This ensures that thesystem realizes the maximum benefit from cool nighttimetemperatures. The systems controller tracks the amount ofstored cooling capacity used in the preceding cooling period,and determines the amount of time required to recharge thestorage section to fu
38、ll capacity. The delayed charge strategy isintended to allow enough time for a full recharge by morning,commonly targeted at 8:00 A.M. As a result, the charge cyclespends as much time as possible in the vicinity of the coolestpart of the night, which is almost universally 6:00 A.M.Another notable ch
39、arge cycle characteristic is that the charg-ing module compressor starts only once per day, avoidingcycle losses. And since the next days cooling is deliveredwithout the use of a compressor, the entire stored capacity ischarged and delivered free of cycling losses on either end ofthe process.The cha
40、rge cycle is controlled to recharge the tank everynight. This improves efficiency by ensuring maximum DXoffset the next day during the scheduled on-peak period. It isimportant to note that a full recharge does not waste moreenergy if it goes unused. This is because the parasitic heatlosses are a fun
41、ction of the temperature difference between theinside and outside of the storage tank. The full volume of waterin the storage tank is not frozen, so the mixed phase temper-ature inside the tank at equilibrium is 32F (0C) for anyamount of storage above a few percent. Therefore, losses arenot increase
42、d by having extra capacity ready to be deployed.Cooling PerformancePerformance penalties associated with high daytime roof-top temperatures are also eliminated with UTSS operation.According to industry norms, rooftop temperatures exceedambient conditions by 6 to 8F (3.3 to 4.4C), accentuating thedeg
43、radation of roof mounted DX equipment performance dueto temperature (Faramarzi 2004). As an example, on a 115F(46.1C) day, the DX condensing unit would be operating at123F (50.6C), degrading efficiency by another 10% (or 12%for R-410A). Ironically, “cool roof” technologies can contrib-ute to this ef
44、fect by reflecting heat directly into the DXcondensing coil surface. However, UTSS systems are unaf-fected by solar heat gain and elevated rooftop temperatures,since the condensing unit runs during the dark and coolevening hours when rooftop nighttime temperatures declinerapidly back to ambient.The
45、UTSS utilizes evaporator coils designed for liquidoverfeed. The combination of using pumped liquid refrigerantand liquid overfeed improves cooling performance andreduces the need to over-size DX equipment for the sake ofmeeting design day. This is further enabled by the fact that thecooling capacity
46、 does not degrade with increased ambienttemperatures, and cycling losses are eliminated by use of asmall refrigerant pump. Daytime rooftop temperatures do notimpact UTSS cooling performance. In fact, the hotter it getsduring the day, the greater the relative site energy efficiencyimprovement over a
47、standard DX unit.UTSS systems reduce the impact of over-sizing on energyefficiency. The efficiency degradation for DX systems asso-ciated with a rise in temperature is due to both a decrease incooling capacity, and a simultaneous increase in energy con-sumption. To compensate for the loss in cooling
48、 capacity, thecondensing unit must be over-sized to serve the load on thehottest days. Also, system performance degrades with age,commonly estimated to be 1% per year(Proctor and Wilson1998). As a result of both factors, systems are routinely over-sized 20 to 50% (Proctor et al. 1995). The most obvi
49、ous con-sequence of over-sizing is that the compressor is larger thanneeded on all but the hottest days. So there is a demand (kW)penalty for about 98% of the years cooling hours (ASHRAE 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission. 150 ASHRAE Transactions2005, Chapter 28). Some
