Performance of Low GWP Refrigerant R-516A in an Air-Cooled Chiller.
Several low-GWP refrigerant candidates have been evaluated as prospective alternatives to R-134a in both screw and centrifugal compressor-based chillers under the Low-GWP Alternative Refrigerants Evaluation Program (AREP) coordinated by the Air Conditioning, Heating and Refrigeration Institute (AHRI, nd). Schultz and Kujak (2013) and Schultz (2014) reported the results from tests of several R-134a-like refrigerants in a water-cooled screw compressor-based water chiller, including R-1234ze(E), N-13a, N13-b (now R-450A), XP10 (now R-513A), and ARM-42a.
N-13a and N-13b produced capacities that were short of R-134a by 12% to 15%. Efficiencies were observed to be lower than with R-134a by up to 3.5%. Poor evaporator and condenser heat transfer coefficients were reported, however, it is now suspected that the N-13a and N-13b temperature glides of ~1[degrees]Fd (0.6[degrees]Cd) led to fractionation within the chiller such that the saturation temperatures calculated from the measured shell pressures were inaccurate, artificially depressing the calculated heat transfer coefficients. While possible, the fractionation associated with zeotropic blends (R-4xx numbered refrigerants) makes their use in chillers problematic because of the difficulty in translating pressure measurements into saturation temperatures due to uncertain or unknown local concentration shifts within the chiller.
R-513A (an azeotropic blend of R-1234yf and R-134a) produced capacities that closely matched those with R-134a while efficiencies were lower by ~4.5%. Schultz et al (2016) reported similar results from tests of R-513A in an air-cooled chiller. Although having an ASHRAE classification of A1 (non-flammable), R-513A has a GWP ~ 600. Although potentially useful as an interim fluid, R-513A's GWP might be too high for long term use under the GWP reduction requirements of the Kigali Amendment to the Montreal Protocol.
R-1234ze(E), with GWP ~ 1, provided only 75% of the capacity obtained with R-134a and therefore would require significant changes to chiller design, such as using compressors with much larger displacement and larger diameter interconnecting piping, to accommodate its lower vapor density.
ARM-42a (an azeotropic blend of R-1234yf, R-134a, and R-152a), with a GWP < 150, produced capacities within 1% of R-134a while efficiencies were lower by ~3.5%, similar to R-513A. Kulankara and McQuade (2013) reported similar results from tests of ARM-42a in an air-cooled screw chiller. The data in both these reports suggested poor condensing heat transfer performance. However, it is now suspected that the saturation pressure-temperature description available at the time for ARM-42a was inaccurate, leading to calculation of higher condensing saturation temperatures from the measured pressures than actual.
The formulation of ARM-42 has been adjusted since the initial tests of ARM-42a. Characteristics of the new formulation, now designated as R-516A by ASHRAE Standard 34-2016 (Addendum b), were described by Abbas et al (2016). ARM-42 is the first ternary (three component) blend to be classified as an azeotrope and receive an R-5xx designation. This paper reports the results of further tests conducted on a 105 RT (370 kW) screw compressor-driven air-cooled water chiller with R-134a as baseline and with R-516A as a potential replacement for R-134a.
The composition, GWP, safety classification, and critical temperature and pressure of the refrigerants considered as potential replacements for R-134a for use in chillers are listed in Table 1. Table 2 lists the thermodynamic performance characteristics of each refrigerant in a simple single-stage vapor compression cycle assuming a fixed compressor volumetric displacement operating between a saturated condensing temperature of 115[degrees]F (46.1[degrees]C) with 15[degrees]Fd (8.3[degrees]Cd) of subcooling and a saturated evaporating temperature of 40[degrees]F (4.4[degrees]C) with no compressor suction superheat. These conditions are typical of an air-cooled water chiller running at the AHRI Standard 550/590-2011 rating conditions (44[degrees]F/6.7[degrees]C leaving chilled water temperature and 95[degrees]F/35[degrees]C outdoor ambient temperature). The isentropic efficiency of the compressor was taken as 0.7. The thermodynamic properties of the single component fluids are based on REFPROP descriptions (Lemmon et al., 2013). The properties of the blends are determined through REFPROP using binary interaction parameters provided by the respective refrigerant suppliers.
The transition to GWP less than 550 entails a corresponding transition from Class 1 (non-flammable) to, at best, Class 2L (flammable with burning velocity less than 10 cm/s). R-152a's flammability classification of "2" essentially removes it from consideration as a replacement for R-134a. The cooling capacity of all the fluids listed here is predicted to be within 6% of that with R-134a, with the exception of R-1234ze(E). All the fluids are subject to at least a small penalty in thermodynamic efficiency (except for R-152a). Interest in R-1234yf for chillers is limited by its relatively poor performance in capacity and efficiency. In spite of its low capacity, R-1234ze(E) is receiving interest as an alternative to R-134a because it is a single molecule refrigerant and because of its inherent thermodynamic efficiency. R-513A offers a good match to R-134a performance, especially being non-flammable; however, its GWP is likely too high for long term deployment. With its close match to R-134a capacity and low GWP, R-516A can be seen as a potentially suitable alternative to R-134a.
The equipment tested was an air-cooled water chiller with a nominal design cooling capacity of 105 RT (370 kW) operating with R-134a refrigerant. The chiller had dual refrigeration circuits driven by matching screw compressors. Only one of the circuits was utilized during the tests reported here. The evaporator was of a flooded style shell-and-tube design. The chiller had previously been used to test a prototype (non-production) microchannel condenser; this was left in place. Refrigerant flow control was done by an electronic expansion valve (EXV) modulated to maintain a target subcooling (as a function of operating conditions) leaving the condenser. The condenser fans were operated at full speed to prevent the algorithms for minimizing total (compressor + fan) power consumption (calibrated for R-134a) from confounding the measurements of overall performance. Capacity was controlled by an adjustable frequency drive down to 50% of full speed followed by compressor slide valve modulation.
The unit was installed in a controlled ambient facility in our laboratory for the tests. Capacity was determined from measurements of chilled water flow rate along with inlet and outlet temperatures (RTDs) and pressures. The average ambient air temperature was determined from a grid of thermocouples placed across the entering faces of the condenser slabs. Electrical power input to the compressor and total power to the unit were also measured. Pressure and temperature (RTD) sensors were mounted at various locations along the refrigerant circuit at the inlet and outlet of the main components. The flow rate of the refrigerant circulating through the chiller was measured by a turbine flow meter installed in the liquid line between the condenser and expansion valve. This allowed an energy balance between the evaporator chilled water and refrigerant flow streams to be computed. The energy balances with R-134a and R-516A were typically of very similar values, being <1.5% in magnitude and within 0.5% of each other at the 44/95 rating condition. Tests were first run with R-134a to establish baseline performance. The R-134a charge was then recovered and the unit charged with R-516A with no other changes to the unit. The same oil charge was used for both tests.
METHOD OF TEST
The method of test followed that in Appendix C of AHRI Standard 550/590-2011 (ANSI/ASHRAE, 2011).
Operating conditions were maintained within tighter tolerances than prescribed by the standard. The cooling capacities reported were calculated from the measured chilled water flow rate and the difference between the entering and leaving chilled water enthalpies. The capacity mentioned here corresponds to the "gross refrigerating capacity" as defined in AHRI Standard 550/590-2011.
Tests in cooling mode consisted of a refrigerant charge sweep at the nominal operating/rating conditions of 44[degrees]F (6.7[degrees]C) chilled water leaving temperature at a flow rate of 265 gpm (16.7 L/s) and an ambient temperature of 95[degrees]F (35[degrees]C). This will be referred to as the "44/95" rating condition. Subcooling at the EXV inlet was controlled to ~18[degrees]Fd (10[degrees]Cd). The flooded evaporator does not produce any compressor suction superheat. The test matrix produced several measurements at the nominal operating condition to check repeatability. The repeatability in capacity and EER was generally within [+ or -]1.5%, trending primarily with small deviations in the ambient temperature from the test run set point.
RESULTS AND DISCUSSION
Overall Chiller Performance
The capacity and EER produced at the nominal 44/95 rating conditions as a function of refrigerant charge are shown in Figure 1 and Figure 2, respectively, for both the baseline R-134a and R-516A. At charge amounts below 80-85 lbm (36-39 kg), performance suffers because the evaporator tube bundle is not fully wetted (as evidenced by evaporator approach temperatures). Once the evaporator bundle is fully submerged, performance is insensitive to the refrigerant charge amount. An operating charge of 90 lbm (40.8 kg) was selected for both R-134a and R-516A.
A set of runs was made with each refrigerant to determine performance from full capacity down to minimum capacity under the 44/95 rating conditions. The results are shown in Figure 3. Full capacity with R-516A fell only about 2% below the capacity attained with R-134a, consistent with the prediction based on thermodynamic properties (Table 2). However, the efficiency (EER or COP) with R-516A at full capacity was about 3.5% lower than with R-134a, about 1.5% lower than predicted by thermodynamic properties alone. The reasons for this will be discussed later.
Figure 4 highlights the replicated data points run under the 44/95 full load conditions. The variation in capacity and efficiency between repeat runs is less than the differences between R-134a and R-516A. The variations correlate closely with small deviations ([+ or -]0.6[degrees]Fd / [+ or -]0.3[degrees]Cd) in the outdoor temperature from the target condition (95[degrees]F / 35[degrees]C).
Capacity and efficiency of the chiller as a function of outdoor air temperature from 65[degrees]F to 125[degrees]F (18[degrees]C to 52[degrees]C) are shown in Figure 5 and Figure 6, respectively. All runs were made with a chilled water leaving temperature of 44[degrees]F (6.7[degrees]C). Capacity with R-516A ran about 2% lower than with R-134a over the full outdoor air temperature range. Similarly, efficiency with R-516A consistently ran about 3.5% lower than with R-134a over the whole range. Typical chiller operating conditions fall well below the critical point for R-134a-like refrigerants, so trends in performance with outdoor air temperature are not expected to differ significantly between refrigerants.
Component Level Performance
The performance of the evaporator was seen to be unaffected by the refrigerant. Figure 7 shows the difference between the chilled water leaving temperature and the refrigerant saturation temperature (computed from the measured evaporator shell pressure). These approach temperatures with R-516A are just slightly larger than those with R-134a. Figure 8 shows the corresponding shell-side heat transfer coefficients (ho'). R-516A falls, at most, only 5% below R-134a. R-516A performed better in this bundle relative to R-134a than the 15% lower single tube pool boiling heat transfer coefficient reported by Abbas et al (2016). R-134a experienced larger approach temperatures and lower heat transfer coefficients at lower heat fluxes (lower capacities). The R-134a refrigerant charge might have been slightly lower than optimal for part load conditions. For the same charge, R-516A's lower density (see Table 2) might have provided sufficient volume for fully wetting the evaporator at part load conditions.
Figure 9(a) shows the approach of the subcooled condenser outlet temperature to the entering outdoor air temperature. The condenser outlet approach temperatures for R-134a and R-516A tended to be within about 0.5[degrees]Fd (0.3[degrees]Cd) of each other. Figure 9(b) shows that the saturation temperatures at the condenser outlet (computed from the measured pressure) for R-134a and R-516A were also very close to each other. (Note that the difference between the temperatures in Figure 9(b) and Figure 9(a) is the amount of subcooling at the condenser outlet.) However, Figure 9(c) indicates that the saturation temperatures entering the condenser with R-516A were about 1.5[degrees]Fd (0.8[degrees]Cd) higher than with R-134a.
As noted in Table 2, the mass flow rate of refrigerant circulating through the unit is about 10% higher with R-516A than with R-134a. Therefore, a higher pressure drop is needed across the condenser with R-516A to move its higher mass flow rate. With similar approach temperatures at the condenser outlet, this higher pressure drop results in a higher pressure and therefore higher saturation temperature entering the condenser. This translates into additional lift that the compressor must overcome, leading to a small additional penalty to efficiency on top of R-516A's inherently lower thermodynamic efficiency. This impact from R-516A's higher mass flow rate can be mitigated by small design changes to the condenser.
An in situ compressor volumetric efficiency can be defined as the ratio of the volumetric flow rate of the refrigerant computed upstream of the compressor suction to the volumetric displacement of the compressor. Figure 10 shows this volumetric efficiency as a function of pressure ratio. The volumetric efficiency with R-516A appears to be slightly lower than with R-134a. This difference can be attributed to uncertainties in the properties of R-516A. A 1% increase in the liquid density (thereby increasing the circulating mass flow rate calculated from the measured volumetric flow rate) or a 1% increase in the vapor specific volume (increasing the volumetric flow rate computed from the mass flow rate) results in the two data sets being coincident. The divergence at high pressure ratios likely comes from use of the slide valve to keep the current drawn by the compressor motor within the allowed limit. R-516A's lower efficiency caused a higher power (current) draw resulting in a larger opening of the slide valve and apparent the lower volumetric efficiencies.
Similarly, an in situ compressor isentropic efficiency can be defined as the ratio of the isentropic enthalpy change of the refrigerant stream passing through the compressor to the power consumed by the compressor. Figure 11 plots the compressor isentropic efficiency against pressure ratio. Compressor efficiency with R-516A appeared to be slightly lower than with R-134a, especially at higher pressure ratios. Again, a 1% increase in the liquid density of R-516A would make the two data sets coincide for pressure ratios less than 4.5. The discrepancy in compressor efficiencies here at higher pressure ratios is not consistent with the trends reported above in overall chiller efficiency where the differences between R-516A and R-134a efficiencies were relatively constant over the range of outdoor temperatures considered. As with the volumetric efficiencies, the isentropic efficiencies at high pressure ratios might be affected by use of the slide valve to limit the compressor motor current draw.
Finally, it should be noted that compressor discharge temperatures with R-516A ran about 10[degrees]Fd (6[degrees]Cd) lower than those with R-134a. This can be of significance to the performance of an oil separator downstream of the compressor. Lower discharge temperatures can result in a higher concentration of refrigerant in the oil collected and returned to the compressor.
Refrigerant R-516A has been tested in a screw compressor-driven air-cooled water chiller as an alternative to R-134a. The cooling capacities with R-516A were within about 2% of R-134a. This is consistent with simple cycle performance predictions based on thermodynamic properties. Overall chiller efficiencies with R-516A ran about 3.5% lower than with the baseline R-134a. This was about 1.5% lower than predicted from thermodynamic properties. This additional reduction in efficiency can be attributed to higher condenser entering saturation temperatures caused by the higher pressure difference needed to drive R-516A's higher refrigerant mass flow rate. This penalty can be mitigated by attention to condenser design.
Beyond the higher condenser pressure drop, the evaporator, condenser, and compressor appeared to be relatively unaffected by the change from R-134a to R-516A. Evaporator saturation temperatures (and approaches to chilled water leaving temperature) were essentially identical between the two refrigerants. Similarly, condenser outlet temperatures (approaches to ambient air temperature) with R-516A were essentially the same as those with R-134a. Very slight differences in compressor performance, measured as in situ volumetric and isentropic efficiencies, were detected. However, the differences don't correlate strongly with overall chiller performance measurements and could be a consequence of slight inaccuracies in computed refrigerant properties.
The results presented here are consistent with other results presented earlier for a similar fluid (ARM-42a). Together, these results indicate that R-516A can be considered as a design-compatible replacement for R-134a with GWP < 150. Work is still needed to complete the assessment of material compatibility, mostly in regard to the R-152a content of R-516A. Evaluation and selection of a compatible lubricant is also needed. In addition, standards and codes will need to be updated to permit the safe use of Class 2L refrigerants before commercialization can occur.
NOMENCLATURE CAP Cooling Capacity [Btu/hr or tons, kW] [DELTA]CAP* Capacity relative to baseline = CAP/base - 1 [-] CDT Compressor Discharge Temperature [[degrees]F, [degrees]C] [DELTA]CDT CDT relative to baseline = CDT - base [[degrees]F, [degrees]C] COP Coefficient of Performance [-] [DELTA]COP* COP relative to baseline = COP/base - 1 [-] EER Energy Efficiency Ratio [Btu/W*hr] ho' shell-side or refrigerant-side heat transfer coefficient [Btu/hr*[ft.sup.2]*R,kW/[m.sup.2]*K] mR* ratio of refrigerant mass flow rate to baseline [-] Pcond condensing pressure [psia, MPa] Pevap evaporating pressure [psia, MPa] Tamb ambient (outdoor) temperature [[degrees]F, [degrees]C] TChWo chilled water leaving temperature [[degrees]F, [degrees]C] Tout refrigerant temperature at condenser exit [[degrees]F, [degrees]C] Tsat saturation temperature [[degrees]F, [degrees]C] [??]L* ratio of liquid density to baseline [-] [degrees]Fd, [degrees]Cd used to signify temperature differences
In consideration of full disclosure, the authors are employed by Ingersoll Rand, which manufactures the type of equipment tested here and is evaluating R-516A and other refrigerants as potential future replacements for R-134a. REFERENCES
Abbas L, Kim S, and Schultz K. 2016. Novel Reduced GWP Refrigerant Compositions to Replace R-134a in Stationary Air-conditioning and Refrigeration. International Refrigeration and Air Conditioning Conference (Purdue). Paper 1707. http://docs.lib.purdue.edu/iracc/1707
AHRI. nd. AHRI Low-GWP Alternative Refrigerants Evaluation Program. http://ahrinet.org/arep.aspx
ASHRAE. 2016. ANSI/ASHRAE Standard 34-2016, Designation and Safety Classification of Refrigerants. Atlanta: ASHRAE.
AHRI. 2011. ANSI/AHRI Standard 550/590 (IP) with Addendum 3, 2011 Standard for Performance Rating Of Water-Chilling and Heat Pump Water-Heating Packages Using the Vapor Compression Cycle. Arlington, VA: AHRI.
Kulankara S and McQuade W. 2013. Test Report #14 - System Drop-In Test of Refrigerant Blend ARM-42a in an Air-Cooled Screw Chiller. AHRI, http://ahrinet.org/App_Content/ahri/files/RESEARCH/AREP_Final_Reports/AHRI%20Low-GWP%20AREP-Rpt-014.pdf
Lemmon, EW, Huber, ML, McLinden, MO. 2013. NIST Standard Reference Database 23: Reference Fluid Thermody-namic and Transport Properties-REFPROP, Version 9.1, National Institute of Standards and Technology, Standard Reference Data Program, Gaithersburg.
Schultz K. 2014. Performance of R134a Alternative Lower GWP Refrigerants in a Water-Cooled Chiller. Presented at the ASHRAE Winter Conference, New York, Conference Paper NY-14-C065. http://www.techstreet.com/ashrae/standards/ny-14-c065-performance-of-r134a-alternative-lower-gwp-refrigerants-in-a-water-cooled-screw-chiller?product_id=1874613
Schultz K and Kujak S. 2013. Test Report #7 - System Drop-In Tests of R134a Alternative Refrigerants (ARM-42a, N13a, N-13b, R-1234ze(E), and Opteon[TM] XP10) in a 230-RT Water-Cooled Water Chiller, AHRI, http://ahrinet.org/App_Content/ahri/files/RESEARCH/AREP_Final_Reports/AHRI%20Low-GWP%20AREP-Rpt-007.pdf.
Schultz K, Kujak S, Majurin J. 2016. Refrigerant R513A as a Replacement for R134a in Chillers. Presented at the ASHRAE Winter Conference, 24-27 Jan 2016, Orlando, Conference Paper OR-16-C017. http://www.techstreet.com/ashrae/standards/or-16-c017-refrigerant-r513a-as-a-replacement-for-r134a-in-chillers?gateway_code=ashrae&product_id=1910117
Kenneth Schultz, PhD
Marcos Perez-Blanco, PhD
Kenneth Schultz is a Principle Engineer and Marcos Perez-Blanco is an Advanced Engineer in the Modeling and Simulation Network of Excellence, Ingersoll Rand, La Crosse, WI.
Table 1. Compositions, GWPs, Class, and Critical Properties of R-134a-like Refrigerants for Chillers Refrigerant Composition (% mass) [GWP.sub.AR4] [GWP.sub.AR5] R-1234yf R-134a R-152a R-134a 100 1430 1300 R-513A (XP10) 56 44 630 573 ARM-42a 82 7 11 115 107 R-516A (ARM-42) 77.5 8.5 14 140 131 R-152a 100 124 138 R-1234yf 100 <1 <1 R-1234ze(E) <1 <1 Refrigerant Class Tcritical Pcritical [degrees]F/[degrees]C psia / MPa R-134a A1 214 / 101 589 / 4.06 R-513A (XP10) A1 208 / 98 554 / 3.82 ARM-42a (A2L) 209 / 98 541 / 3.73 R-516A (ARM-42) A2L 211 / 99 551 / 3.80 R-152a A2 236 / 113 655 / 4.52 R-1234yf A2L 202 / 95 491 / 3.39 R-1234ze(E) A2L 229 / 109 527 / 3.63 Table 2. Simple Thermodynamic Cycle Performance Predictions Refrigerant [DELTA]CAP* [DELTA]COP* Pcond Pevap psia / MPa psia / MPa R-134a 0.0 0.0 173 / 1.19 49.7 / 0.343 R-513A (XP10) +0.004 -0.023 180 / 1.24 54.2 / 0.374 ARM-42a -0.023 -0.023 175 / 1.21 53.6 / 0.370 R-516A (ARM-42) -0.016 -0.021 175 / 1.21 53.5 / 0.369 R-152a -0.052 +0.028 155 / 1.07 44.8 / 0.309 R-1234yf -0.062 -0.047 172 / 1.19 53.1 / 0.366 R-1234ze(E) -0.257 -0.002 131 / 0.90 36.9 / 0.254 Refrigerant [DELTA]CDT [??]L* [degrees]Fd/[degrees]Cd R-134a 0/0 1.00 R-513A (XP10) -10/-5.5 0.94 ARM-42a -12/-6.5 0.89 R-516A (ARM-42) -10/-5.5 0.88 R-152a +20/+11 0.75 R-1234yf -17/-9.5 0.90 R-1234ze(E) -12/-7.0 0.97
|Printer friendly Cite/link Email Feedback|
|Author:||Schultz, Kenneth; Perez-Blanco, Marcos|
|Publication:||ASHRAE Conference Papers|
|Date:||Jan 1, 2018|
|Previous Article:||Investigation of Low GWP Flammable Refrigerant Leak from Rooftop Units.|
|Next Article:||A Simulation Study on the Performance of Radiant Ceilings Combined with Free-Hanging Horizontal Sound Absorbers.|