Field esting of a Prototype Residential Gas-Fired Heat Pump Water Heater Paul Glanville, PE Hillary Vadnal Michael Garrabrant, PE Member ASHRAE Member ASHRAE Member ASHRAE ABSRAC HEADING Approximately half of water heaters sold in the U.S. and Canada for residential applications are natural gas fired storage water heaters, and for these products the maximum steady state thermal efficiency of available products is approximately 96%, with transient rated efficiencies much lower. o move beyond the thermal efficiency limits of standard condensing-efficiency residential gas water heating equipment, this paper describes an effort to develop an economic gas-fired ammonia-water absorption heat pump deployed as a packaged storage water heater. his new class of gas-fired heat pump water heaters are driven by a direct-fired 2.9 kw (10 kbtu/hr) air source absorption heat pump, which like vapor compression (electric) heat pump water heaters (EHPWH) utilize ambient air to heat stored potable water. With a small 2 kw (6.3 kbtu/hr) gas burner, the packaged unit can exceed this efficiency limitation with Coefficients of Performance (COPs) in excess of 1.5. Unlike EHPWHs however, the prototype gas-fired heat pump water heater (GHPWH) heats potable water with a combination of output from the condenser, absorber, and heat recovery from the products of combustion, hence the evaporator load (space cooling effect) is 30-40% that of an equivalent vapor compression system. Concerning system cost, with a thermal input of 16% of storage GWHs and 3% of tankless GWHs, the cost of GHPWH installation is minimized, requiring only small diameter gas and plastic vent piping, and standard electrical service. his GHPWH represents a step-change in energy efficiency at a projected competitive cost, a critical goal of both government agencies and utility energy efficiency programs. In this paper the authors report on data and findings from a preliminary field evaluation of this prototype GHPWH. INRODUCION With improving building envelopes reducing space heating loads and the continued growth of condensing efficiency warm air furnaces, estimated at over 50% of the U.S. furnace market, water heating represents a growing portion of the residential gas load on the west coast at 35% and growing, nearly 50% in California (EIA, 2009 and Seto, 2013). Despite this, of the approximately half of all residential water heaters sold in the U.S. and Canada that are natural gas-fired, the majority are minimum efficiency gas-fired storage water heaters with an average Energy Factor (EF) of 0.60. Highlighting that the majority of products sold are low-efficiency, recent data from the U.S. EnergyStar show that for the 4.3 million residential gas water heaters sold in 2013, 161,000 were high-efficiency storage-type (condensing and non-condensing) and 397,000 were high-efficiency tankless-type, with the remaining 87% of gas Paul Glanville is a Senior Engineer and Hillary Vadnal is a Principal Engineer, both with the Gas echnology Institute in Des Plaines, IL. Michael Garrabrant is the President of Ston e Mountain echnologies Inc., in Johnson City, N.
Estimated Delivered Efficiency (Output/Input) products low-efficiency. Similarly, of the 4 million electric residential water heaters sold, only 43,000, or 1% of 2013 shipments were high-efficiency electric heat pump water heaters (EHPWH) (EnergyStar, 2013). In general, residential hot water consumption has been on a slight decline over the past few decades, due in part to declining occupancies, broader deployment of water-efficient fixtures, and migration from colder to warmer climates (a trend that has had a more pronounced effect on space heating loads). Similarly, domestic water heating patterns were found to be more distributed throughout the day, with fewer, shorter duration draws, than previously documented. his has broadly been discussed as it relates to the changing U.S. method of test for rating residential water heaters. In a recent meta-analysis of residential domestic hot water consumption, over 10 U.S. studies, Lutz et al. found that the actual daily median quantity is 61.6 hot water draws per day, versus six in the current rating method; and that the average daily median hot water draw volume is 50.6 gallons (196 L), versus the 64.3 gallons(249 L) in the current method (Lutz, 2011). hese have competing effects on electric and gas storage water heaters, both heat pump or conventional: Smaller daily draw volumes yield shorter water heater runtimes, thus the load from standby losses are a larger fraction of total output, yielding lower delivered efficiencies. Specific to gas products, this was recently illustrated in a field and laboratory evaluation of gas water heating products in California, with data summarized in Figure 1. Distributed draw patterns versus clustered draw patterns impact delivered efficiency through increased outlet water temperatures. For slower recovery HPWHs, this is more important and will be a focus of this study. 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Output (L/day, 19 C Rise) 0 50 100 150 200 250 300 350 400 450 Condensing Storage Non-condensing ankless Condensing ankless Non-condensing Storage 0 20 40 60 80 100 120 Output (Gal/day, 67 F rise) Figure 1: Delivered Efficiency vs. Output for Gas Water Heating (Kosar, 2013) Opportunity for Gas Heat Pump Water Heaters Focusing on the retrofit market of the common, low-efficiency, gas-fired storage-type water heaters, an effort is underway to demonstrate a gas-fired heat pump water heater (GHPWH) with a projected EF of 1.3 over twice that of standard gas-fired storage water heaters (Garrabrant, 2013). his work looks beyond existing high-efficiency options for gas water heating customers, which have their own drawbacks and limitations, such as the known gap between the rated versus installed efficiency of gas tankless water heaters (Kosar, 2013). he packaged GHPWH heats the approximately 75 gallon (285 L) of stored water with a nominal 10,000 Btu/hr output (2.9 kw) ammonia-water absorption heat pump, driven by a small 6,300 Btu/hr (1.9 kw) low-emission gas burner, and exceeds the thermal efficiency limitation of standard gas-fired products with Coefficients of Performance (COP) in excess of 1.5. he GHPWH represents a similar leap forward in water heating efficiency to the recent generation of residential EHPWHs, that have demonstrated delivered efficiencies at least twice that of standard electric resistance water heaters
(Glanville, 2012). Like the packaged EHPWHs, the GHPWH is comprised of three major components: a) storage tank, b) sealed system (set of heat exchangers containing the refrigerant), and c) supporting components such as the evaporator fan, combustion system, and controls. With the advent of the new U.S. minimum efficiency requirements (DOE, 2010) and the revised method of test (DOE, 2014), a technology like the GHPWH will benefit from these regulatory-driven market transformation as follows: a) Large volume storage tanks, above 55 gallons storage (213 L), have a higher allowable minimum efficiency than smaller storage tanks. With a required EF of approximately 0.75 for a 60 gallon tank (234 L), versus 0.62 for a 40 gallon tank (156 L), the large volume storage tank market will shift to condensing storage-type. With an installation cost of 2.5 to 3 times that of the 0.62 EF unit and even higher for retrofits these condensing storage-type units will struggle to justify the 20% increase in operating efficiency. With a lower projected installation cost, end users will likely favor the GHPWH with a 110% operating efficiency increase. b) he new method of test, with a distributed draw pattern, including more, shorter duration, hot water draws, may result in a convergence between the current EF ratings of tankless water heaters and those reductions employed by various authorities throughout the U.S., by up to 9%. c) he modified scope of the new method of test adds a new Gas Heat Pump Water Heater product category, in the recognition that a commercialized product is likely to be introduced during this regulatory cycle. his allows for ready categorization of the product, with a Uniform Energy Factor (UEF) and thus eligibility for EnergyStar. DESCRIBING HE GAS HEA PUMP WAER HEAER he GHPWH is based on the vapor absorption refrigeration cycle, using the ammonia-water working fluid pair, which an absorbent (water) is used as a carrier for the refrigerant (ammonia). While the refrigerant is still compressed by an electromechanical pump like EHPWHs, unlike a more typical vapor compression cycle, it is compressed as a liquid in solution with the absorbent. Lifting the pressure of a liquid versus a vapor requires significantly less energy. For example, comparing a 1.3 COP heating ammonia-water heat pump to an 8.2 HSPF vapor compression heat pump, the absorption cycle solution pump requires less than 1.0% of the total energy input to the electric compressor (Herold, 1996). Condenser Heat out - Hydronic Heat Out Flue CHX Desorber Heat in - Combustion Expansion Valve High Pressure Low Pressure hermal Compressor Solution Pump Absorber Heat out - Hydronic Refrigerant (NH3) Refrigerant/Absorbent (NH3/H2O) Evaporator Heat In - Ambient Figure 2: Simplified Diagram of the GHPWH Absorption Heat Pump Figure 2 shows a simplified diagram of the absorption heat pump, highlighting the common components with
vapor compression heat pumps (condenser, evaporator, expansion valve) and the components that comprise the thermal compressor. Like vapor compression-based EHPWHs, this refrigeration effect moves heat from ambient air, at the evaporator, to the stored water (via a hydronic loop), at the condenser. While compression of the liquid refrigerant/absorbent solution is performed by the solution pump, thermal energy from the single-stage, 6,300 Btu/hr (1.9 kw) gas burner is required to drive the refrigerant vapor from its absorbed state in the desorber (sometimes referred to as a generator ). his desorption process occurs at an elevated temperature, 250-300 F (121-149 C), thus exiting flue gases still have useful heat, which is recovered in a separate condensing heat exchanger (CHX), submerged within the storage tank. As adequate refrigerant/absorbent pairs require high affinities and stability over a wide range of temperatures and pressures, they have significant heats of absorption. As such a portion of the desorption heat input is recovered at the absorber, from the heat of absorption of ammonia to water, providing a second heat output to the stored water (via a hydronic loop). hus, the GHPWH heats the stored water via three inputs: condenser heat from the heat pump, recovered heat of absorption in the absorber, and heat recovery of the flue gases via the CHX. Like many of the EHPWHs currently available, the total output of the GHPWH is approximately 10,000 Btu/hr (2.9 kw) and the portion of the heat delivered to the storage tank from the refrigeration effect is a fraction of total heat delivered, which combined yield the following unique features of this GHPWH design as compared to gas water heaters and EHPWHs: With a small combustion system, the gas piping does not require an upsizing from ½ piping (13 mm), typical for converting a lowefficiency storage water heater to a condensing storage or tankless unit. In fact, in new construction applications, smaller ¼ piping (6.4 mm) may be feasible for the GHPWH. As gas-fired equipment with maximum firing rates at or below 10,000 Btu/hr (2.9 kw) are not common, neither is ¼ gas piping (6.4 mm). However, through the 2003 version of the International Fuel Gas Code (IFGC, 2003), sizing guidance exists for ¼ piping (6.4 mm) for pressures of 0.5 psi (3.4 kpa) or lower. While subsequent versions of the IFGC only have guidance for 2.0 psi (13.8 kpa) operating pressures and below, with sizing guidance down to ½ piping (13 mm), numerous authorities having jurisdiction retained this guidance for ¼ (6.4 mm) from this older version, such as for New York City did through the end of 2014 (NYC, 2014). Per this guidance, a run of 150 ft. (45.7 m) of ¼ (6.4 mm) Schedule 40 pipe would suffice for the GHPWH, assuming an operating pressure of 0.5 psi (3.4 kpa) or less and a pressure drop of 0.3 W.C (0.07 kpa). Also, the smaller GHPWH firing rate yields proportionately smaller output of flue gases, allowing for smaller diameter venting of 1 or less. As a condensing efficiency water heater, lower cost venting materials such as PVC are suitable. Prior field installations of GHPWHs used ¾ PVC venting (see Figure 3). With a small heat pump, within the sealed system, the total Figure 3: Prototype GHPWH at N Demo Site ammonia charge is about 1.5 lbs (0.7 kg), lower than the 6.6 lb (3 kg) limit placed on its indoor use by ASHRAE Standard 15. he safe use of ammonia as a refrigerant for indoor equipment has been well demonstrated since the first widespread use of absorption refrigerators in the early 20th century to current times where the quieter absorption mini-refrigerators are preferred by large hotels for guest rooms. he absorption heat pump only partially heats stored water using ambient energy, cooling the room in which it is installed, however this cooling effect is between one third and one half that of EHPWHs, with the other heat outputs to the water coming from the absorber and heat recovery of the flue products.
Compared to the most typical gas-fired water heater, the atmospherically-vented non-condensing storage water heater, the GHPWH and its smaller combustion system have a reduced impact on the home s overall HVAC. Without the central flue common to these non-condensing gas storage water heaters, the standby heat loss from storage tank is not removed through the flue vent but rather warms the surrounding space, a minor benefit during the heating season. With a much smaller combustion air requirement and no draft hood for flue gas dilution (the GHPWH is a power vent water heater), this added ventilation load on the HVAC system is minimized and combustion safety issues with depressurization within tighter envelopes are eliminated. Like EHPWHs, the efficiency of the GHPWH is both a function of the storage tank temperature, specifically the return temperature from the internal hydronic loop, and of the ambient air temperature. As a result, efficiency and heating capacity will vary seasonally, as ambient air temperatures vary in the conditioned or semi-conditioned installation space, and inlet water mains temperatures vary over the course of a typical heating cycle as hydronic return temperature increases until the thermostat is satisfied. Figure 4 shows this latter effect, with COP versus heat pump supply temperature for various generations of previous laboratory GHPWH prototypes (Garrabrant, 2013). hese seasonal and cyclical effects will be investigated during the GHPWH field evaluation. Hot Water Supply emperature [F] 77 95 113 131 149 Figure 4: Laboratory GHPWH Prototype Steady-State Performance and 20 C (68 F) Ambient Conditions (Multiple Generations of Prototype) FIELD ES DESCRIPION AND RESULS In late 2014 a pre-commercial GHPWH prototype was built and installed at a residential field test site in ennessee, in the basement garage of a 3-4 occupant home (variable occupancy for college-aged child). his initial field evaluation of the pre-commercial GHPWH focused on the following: 1) quantifying the performance and efficiency of the GHPWHs in a real world application, 2) understanding the reliability of critical components, and 3) the participating manufacturer developed refined strategies for system controls. Data collection and monitoring points are summarized in Figure 5, indicating measurement of hot water usage (water flow, inlet/outlet temperatures), GHPWH energy consumption (gas/electricity inputs), and numerous heat pump cycle temperatures. For the latter, GHPWH heat pump cycle temperatures are used in the disaggregation of heat pump cycle COP, ratio of heat delivered to the hydronic loop to the net heat input from natural gas combustion to the heat pump, from the system
COP, which is the ratio of heat delivered to the storage tank to the total energy input (natural gas and electricity). Going one step further, the delivered efficiency (DE) is defined as the ratio of useful hot water output to the total energy input (natural gas and electricity) over a defined period, incorporating standby heat loss and energy consumption. For these three efficiency metrics, over any period of time-averaging, COP cycle COP system DE. Over approximately 10 months of operation in the field, the GHPWH performed with cycle COPs near that of prior laboratory testing for much of the time. In Figure 6, the average cycle COPs are shown as a function of ambient temperature ( F), with each data point representing an individual cycle. hese data are separated into two phases, whose separation indicates a major repair to the GHPWH. During the first 2.5 months, Phase 1, the GHPWH s electronic expansion valve and associated controls resulted in numerous cycles with off-design performance. he Phase 2 indicates the monitoring period after the repair of these associated hardware and control software issues. he improvement in COP is marked, with the second phase containing data for 243 firing cycles and highlighting the relative insensitivity of the ammonia-water absorption cycle to changes in operating temperatures in comparison to that of hydronic return/supply temperatures (Figure 4). Average Cycle Heat Pump COP Cold water in Hot water out F Ambient & RH Evaporator Mechanical P F Natural Gas Power meter GHPWH Detail Figure 5: Diagram of Measurements during GHPWH Field esting ( = emperature, F = Flow, P = Pressure) Ambient emperature (C) 4.4 9.4 14.4 19.4 24.4 1.8 1.7 1.6 1.5 1.4 1.3 1.2 GHPWH - Second Phase 1.1 GHPWH - First Phase 1 40 45 50 55 60 65 70 75 80 Ambient emperature (F) Figure 6: Average Cycle COP per GHPWH Cycle Pre/Post-Repair
Estimated Delivered Efficiency (Output/Input) o compare to results to that of conventional gas water heaters, the Input/Output method is used, which posits that the daily energy input versus output of a heating system can yield a delivered efficiency from their linear relationship of the transient energy input to the energy output (Bohac, 2010 and Butcher, 2011). When data are plotted on this I/O chart the slope (m) and y-intercept (b) they can be used to estimate the DE, as follows: Input = m Output + b; Output Input = DE = (m + b Output ) 1 (1) With this relationship and the data for the GHPWH field testing, the daily input energy and output energy are used to generate a DE and these are plotted versus hot water output in Figure 7. Note that the input energy is measured natural gas consumption and an assumed electricity consumption based on runtime, 110 W consumed while active and 5 W consumed while in standby, actual measured power consumption is higher by approximately 2 times due in part to the prototype controls that will not be present with a commercialized product. For the entire monitoring period, including pre/post-repair data, the DE versus output data are overlaid onto those data for conventional gas equipment previously shown in Figure 1, for non-condensing and condensing storage (NCS/CS) and non-condensing and condensing tankless (NC/C). Using these data, one can project the DE for a given hot water output, for example the projected DE of the GHPWH at the High Usage pattern within the new U.S. method of test, 88 gallons/day at a 67 F rise (334 L at a 37 C rise) or approximately 49,000 Btus/day (14.4 kw), is 1.35 (assuming that the commercialized GHPWH has the design electric parasitic loads). his is compared to 0.85 for condensing tankless and 0.60 for non-condensing storage, thus the GHPWH may realize significant energy savings for gas water heating applications. At the time of writing, there are several additional GHPWH field tests ongoing to add to these datasets, seeking to confirm these early performance numbers in different climate zones, residence sizes, and other variants. 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 Output (kwh/day) 0 5 10 15 20 GHPWH CS NC C NCS Log. (GHPWH) 0 10000 20000 30000 40000 50000 60000 70000 Output (Btu/day) Figure 7: Comparison of GHPWH Delivered Efficiency to Convention Gas Water Heating
CONCLUSION Following a recent laboratory prototype development programs, the performance of a gas heat pump water heater installed in a ennessee residence over a 10-month period is described. he laboratory-validated performance and preliminary field test data suggest that the 1.3 UEF target is feasible and if achieved during commercialization, it can be competitive in with other available high-efficiency gas water heating options. Cycle COPs of the GHPWH are on par with prior laboratory results and, as is expected for this type of technology, are relatively insensitive to ambient temperatures. he performance of the GHPWH is shown to be robust over a range of operating conditions, including usage patterns, ambient temperatures, and water mains temperatures, by both laboratory testing and field evaluation of the GHPWH in a U.S. residence. As described, the GHPWH is projected to have a reduced retrofit cost with a small input capacity, requiring no upsizing of gas piping and minimal accommodation of small diameter plastic venting, thus if competitively priced, the GHPWH is an important introduction to the high-efficiency water heating market. REFERENCES Bohac, D. et. al. Actual Savings and Performance of Natural Gas ankless Water Heaters. Minnesota Center for Energy and Environment, 2010. Butcher. et al. Application of Linear Input/Output Model to ankless Water Heaters, Brookhaven National Laboratory. Presented at ASHRAE Winter Meeting, Las Vegas, NV, 2011. Department of Energy, Energy Information Administration. Residential Energy Consumption Survey, 2009 ed. http://www.eia.doe.gov/emeu/recs/contents.html Department of Energy, 10 CFR Part 430, Energy Conservation Program: Energy Conservation Standards for Residential Water Heaters, Direct Heating Equipment, and Pool Heaters, Final Rule, Docket No. EE-2006-B-SD-0129, 2010. Department of Energy, 10 CFR Parts 429, 430, and 431 Energy Conservation Program for Consumer Products and Certain Commercial and Industrial Equipment: est Procedures for Residential and Commercial Water Heaters, Docket No. EERE-2011-B-P-0042-0082, 2014. EnergyStar, via the US Environmental Protection Agency, EnergyStar Unit Shipment and Market Penetration Report Calendar Year 2013 Summary, 2013. Garrabrant, M. Development and Validation of a Gas-Fired Residential Heat Pump Water Heater - Final Report. 2013, prepared for the U.S. Department of Energy, Contract DE-EE0006116. Glanville, P., Kosar, D., and Suchorabski, D. Parametric Laboratory Evaluation of Residential Heat Pump Water Heaters, rans. of ASHRAE v. 118 pt. 1, Chicago, IL. (2012). Herold, K., Radermacher, R., and Klein, S. Absorption Chillers and Heat Pumps. CRC Press, aylor & Francis Group, 1996. International Fuel Gas Code, 2003 Edition. Kosar, D., Glanville, P., and Vadnal, H. Residential Water Heating Program - Facilitating the Market ransformation to Higher Efficiency Gas-Fired Water Heating - Final Project Report, CEC Contract CEC-500-2013-060, 2013. Lutz, J. D., Renaldi, Lekov, A., Qin, Y., and Melody, M. Hot Water Draw Patterns In Single Family Houses: Findings from Field Studies, Lawrence Berkeley National Laboratory, Report LBNL-4830E, 2011. New York City Fuel Gas Code, 2014. Seto, B. et al. California Energy Commission Energy Efficient Natural Gas Use in Buildings Roadmap, presented in Sacramento, CA, 2013.