The Next Generation of
Lubricants for Improved System Performance
Refrigeration lubricants can have a significant beneficial
impact on two major performance parameters of critical importance to the
refrigeration and air conditioning industry:
-
Improved energy efficiency
-
Noise reduction
Energy efficiency improvements are driven by the need to reduce
green house gases. On average, it has been estimated that during the life of an
refrigeration system (1) some 84% of green house emissions are due
to the energy consumed during the life of the fridge. This can rise to as much
as 97% for appliances(1). Governments have responded this issue with
a variety of schemes
-
USA:
·
DoE 30% improvement by 2001, 15% by 2003
·
Energy Star label (20% better than minimum for
RAC)
-
Europe: CEN classification system
-
Japan:
·
Voluntary steps towards Kyoto protocol (25-30%)
·
Energy labeling scheme to be introduced 21 August
2000
§
Orange: failed to meet standard by X%
§
Green :
achieved standard by X%
Recently, George Bush mentioned appliance standards as an important
element of the new administration national energy plan(2). Further, legislation
or tax credit schemes could be forthcoming.
In the Industrial commercial sector, there has also been movement towards
improvements in energy efficiency standards to be achieved by 2006(3). These are:
-
DoE has proposed
a 20% increase in federal minimum efficiency standards
-
The institute urged adoption of targeted tax
incentives
-
If this was adopted today 84% of central air
conditioning and 66% of heat pumps would not comply with the proposed
legislation
Correct selection of the compressor lubricant can have a
significant impact on the energy efficiency performance of a system. Lubricant
selection can impact compressor energy efficiency in a number of ways. This
paper will focus on the energy losses arising from frictional losses in
lubricated contacts and heat transfer processes. However, it is important to
recognise that lubricant properties can also impact energy efficiency in other
ways. For example, use of a lubricant
with too low viscosity, or too high refrigerant gas solubility, can reduce
volumetric efficiency in some compressor types.
Frictional energy losses due to viscous drag between the
lubricant and bearing surfaces in an appliance compressor can contribute
significantly to the overall efficiency of a refrigeration system. In order to
establish the expected relationship between lubricant viscosity and energy
consumption, refrigerator efficiency tests based on the conditions of ISO 5155
were carried out on standard commercial refrigerator and freezer cabinets.
Figure 1: The Effect of
Viscosity on the Energy Efficiency of a Domestic Fridge and Freezer Cabinet.
The graph clearly shows that by reducing lubricant viscosity
energy efficiency can be improved by at least 10%.
Friction modifiers or lubricity additives can also give marked
reduction in friction. To illustrate the frictional loss behaviour of
lubricated contacts across the range of lubrication regimes, the friction
coefficient data was obtained using a conforming-pin-on-ring tribometer of
in-house design(4).
Figure 2: The Effect of Friction Modifiers/Lubricity
agents on the Coefficient of Friction
Use of friction modifier additives will contribute to energy
efficiency only if a significant proportion of power losses occur at boundary
or mixed film lubricated contacts. (It should be noted that even if a contact
operates in the hydrodynamic regime at standard operating conditions, it must
pass temporarily through the boundary regime at start up and shutdown. This is
particularly important at start-up, when the contacts may be lubricant-starved
due to drainage occurring during while the equipment is stationary.
It has been
previously supposed that lubricants generally have a negative impact on heat
transfer. However, it is possible that lubricants can improve, rather than
reduce, the heat transfer efficiency. Improvements
in heat transfer can also lead to marked improvements in energy efficiency.
It has been
recognised for nearly forty years that a lubricant-rich fluid layer exists near
the tube walls in evaporators(5). Recently,
Kedzierski has proposed a logical link between the pool boiling
enhancement/degradation associated with lubricant contamination and the
accumulation of the lubricant at the boiling surface (6,7). The excess concentration (excess surface
density) arises from the relatively low vapour pressure of the lubricant
compared to refrigerant. At the heated wall, the refrigerant preferentially
evaporates, leaving behind a relatively thin liquid layer enriched in
lubricant. It has been hypothesised that the lubricant excess layer establishes
the bubble size and site density for boiling, which are known to have profound
effects on the magnitude of the heat transfer(6,7).
In joint
collaboration with Kedziersk several polyol esters have been screened using the
NIST pool boiling apparatus. It has been found that:
-
Larger lubricant mass fractions promote smaller
bubble departure diameter, which in turn, leads to poorer heat transfer.
-
Proximity of the bulk fluid temperature to the
Critical Solution Temperature of the mixture benefits heat transfer by the
formation of additional excess liquid films that draw superheated liquid into
the bubbles sides
-
High Lubricant viscosity benefits pool boiling by
promoting a thick thermal boundary layer.
Heat transfer of POE lubricant gave results of 45% - 120% of pure R134a. Although very system dependent, a very crude rule of thumb is,
1% improvement in energy efficiency can be achieved for every 2% improvement in
heat transfer. Further work is ongoing to optimise lubricant properties to give
improved heat transfer characteristics.
In summary, major improvements in energy efficiency for
appliance compressor can be achieved through reducing the lubricant viscosity.
For larger industrial commercial systems, improvements in energy efficiency can
be obtained through a combination of viscosity optimisation, reduction in
friction coefficient and improvements in the heat transfer properties of the
lubricant.
Manufacturers of domestic appliances are increasingly seeking to reduce
noise as a route to competitive advantage. It is widely believed that foaming
contributes to noise reduction in refrigeration lubricants(8), and lubricant
formulations designed to induce foaming have been reported(9). The presence of foam
on the surface of the lubricant undoubtedly acts to reduce noise generation due
to lubricant droplet impacts. However, the mechanisms by which foams could
actually absorb acoustic energy are not clear and it is likely that entrained
vapour bubbles actually make a much greater contribution to acoustic
attenuation in refrigeration lubricants(10).
Vapour bubbles form readily in a refrigeration lubricant because it is
saturated with dissolved refrigerant gas. The presence of such vapour bubbles
at the metal – lubricant interface will significantly reduce the acoustic
transmission coefficient of the interface. The presence of vapour bubbles also
amplifies the effects of the relaxation attenuation processes occurring in the
liquid. Because the vapour is more compressible than the bulk liquid, the bubbles
contract disproportionately more than the liquid in response to the acoustic
pressure wave. Consequently, the surrounding liquid responds with a greater
amount of viscous flow and absorbs much more energy than in a homogeneous
liquid(11,12). Gas bubbles also refract and scatter acoustic energy.
Foam generation and vapour entrainment are related, but not identical,
properties of a lubricant formulation, which are partly dependent on bulk
properties such as viscosity, but more sensitive to surface properties
including surface tension, Gibbs elasticity and interfacial shear and
dilational viscosities (13). These properties can be modified both by control of the chemical
structure of the basefluids and by use of appropriate additives in the
lubricant formulation. However, this presents a particular challenge in
lubricant formulation as both foaming and vapour entrainment can have
detrimental effects on compressor lubrication and system energy efficiency. For
example, high foam levels can enter the suction port leading to excess
lubricant transport, whereas entrained vapour bubbles may reduce the
hydrodynamic lubricant film thickness. Lubricant formulations for improved
noise and vibration performance must maintain stable, controllable levels of
foam and vapour entrainment in order to deliver improved acoustic performance
without compromising reliability or efficiency.
Standard foam stability and vapour entrainment test methods(14,15) can be used to select candidate basefluids and acoustically active
additives and to optimise dose rates. Light scattering analysis can be used to
compare bubble size distribution under standard conditions. In one such test,
the lubricant is stirred and sparged with gas for a defined period. After the
gas flow is stopped, the bubble size distribution is determined as a function
of time. Figure 3 shows results illustrating how a candidate acoustically-active
additive stabilises entrained vapour bubbles in a model lubricant basefluid. In
the pure basefluid, the mean bubble size is initially high and rapidly
decreases to zero as the largest bubbles rapidly rise to the surface and burst.
In the formulated fluid, the initial bubble size is lower and the mean bubble
size changes much more slowly over time, showing that the bubbles are retained
within the bulk of the lubricant.
Figure 3: Effect
of Acoustically Active Additives on Mean Bubble Size using Laser Diffraction
Control of the bubble size distribution is important, because this
determines the frequency dependence of the excess acoustic attenuation due to
the entrained bubbles(11). Acoustic attenuation is particularly high at frequencies corresponding
to resonant modes of the bubble cavity. It is therefore possible to tune the
absorption characteristics of a fluid to match the noise spectrum of the
equipment, by control of the bubble size distribution.
Previous workers have found that the
spectrum of radiated noise from a compressor is very similar to the vibration
spectrum of the compressor shell, and that this is determined by the resonant
frequencies of the compressor shell(16). These resonant modes each
have a characteristic distribution of intensities across the shell, so the
measured vibration spectrum is highly dependent on the measurement position. It
is therefore necessary to measure vibration at a number of locations in order
to assess the overall potential for noise radiation.
Accelerometer attachment studs were fixed to the shell of a
representative compressor at a range of locations, including one that was
designated as the reference point. The
compressor was then charged with a standard amount of refrigerant gas and a
test lubricant and operated until steady state temperature and pressure
conditions were established. Shell acceleration was then recorded sequentially
at each location, over a designated time interval typically ranging from 5s to
60s, for post processing. Acceleration at the reference point was recorded
simultaneously and monitored to confirm that the operating conditions were
constant during the entire measurement sequence.
Acceleration data were recorded and post-processed using a commercial
vibration analysis software package. For each measurement point, the raw
acceleration data were integrated to give velocity as a function of time, then
Fourier transformed to give the vibration velocity frequency spectrum. The
A–weighting was applied to the spectrum, and the mean vibrational velocity
calculated for the frequency range 0 – 8 kHz and expressed as dB(A) re 5x10-8m/s.
The process was repeated for each lubricant formulation. Acoustic performance
of each formulation was ranked according to the comparative value of the
A-weighted mean vibrational velocity.
Figure 4: Instrumented
Noise Evaluation Test Stand
It should be emphasized that this is a screening test to assess the
relative performance of different lubricants. The absolute numerical value of
the mean vibratory velocity in dB(A) re 5x10-8m/s will not necessarily be close to the sound
power level radiated by the compressor in dB(A) re 1x10-12 W,
because the compressor shell surface is not flat and the radiation ratio s is not unity. However,
as part of an overall screening technique the method has the advantage of being
quick, cost effective, and is a useful screening tool before more in-dept test
are carried out.
Results using this technique can be seen in Figure 5.
Figure 5: Noise
Evaluation on various Polyol Ester Appliance Grades
Noise reducing lubricant compositions will not be effective in all
compressor designs. The reasons for this behavior are complex and are dependent
on a large number of variables. Obviously, compressors, which already have high
levels of entrained refrigerant in the lubricant, may have excessive foaming if
noise reducing compositions are used. In addition, the optimum bubble size to
damp a particular frequency in a specific compressor design may have not been
the correct one for an alternate compressor design.
Lubricants can now be specifically designed to deliver targeted
performance advantages in the areas of improved energy efficiency and noise
reduction. Delivering these effects via the lubricant, rather than the more
expensive system design route, can results in significant cost savings.
However, a detailed knowledge of the structure property relationships of the
lubricant is required to achieve the optimum balance of properties
(1)
March Consulting Group, Final Report 30/09/98
(2)
Assosiation of Home Appliance Manufacturers (18th
May 2001). http://www.aham.org
(3)
E.Dooley. (13th April 2001). Good News
for Consumers on Minimum Efficiency Standards for Central Air Conditioners and
Heat Pumps, ARI, http://www.ari.org
(4)
S.Boyde, S.Randles, P.Gibb, S.Corr, P.Dowdle,
A.McNicol and R.Bailey. (July 2000). Effect of Lubricant Properties on
Efficiency of Refrigeration Compressors. International Compressor Engineering
Conference at Purdue, West Lafayette, Indiana, USA.
(5)
K.Stephan. (1963). Influence of Oil on Heat Transfer of Boiling
Refrigerant 12 and Refrigerant 22. XI International Congress of Refrigeration,
Vol. 1, pp. 369-380.
(6)
M.Kedzierski. (2001). The Effect of Lubricant Concentration, Miscibility
and Viscosity on R134a Pool Boiling. International Journal of Refrigeration,
Vol. 24, No. 4.
(7)
M.Kedzierski and M.Kaul. (1993). Horizontal Nucleate Flow Boiling Heat
Transfer Coefficient Measurements and Visual Observations for R12, R134a, and
R134a/Ester Lubricant Mixtures,” NISTIR 5144, U.S. Department of Commerce,
Washington, D.C.
(8)
ASHRAE handbook (1998).
Refrigeration, 8. 7.20. American Society of Heating, Refrigeration and
Air-Conditioning Engineers.
(9)
K S Sanvordenker .US Patent US
3792755.
(10)
S.Boyde. (2001) The role of Lubricant
in Reducing Vibration and Noise in Refrigeration Compressors. To be presented
at ImechE Compressors and their Systems conference to be held in November.
(11)
L E Kinsler and A R Frey. (1962).
Fundamentals of Acoustics, 2nd Edition, John Wiley.
(12)
J R Allegra and S A Hawley.
(1972). J. Acoust. Soc. America, 51,
1545.
(13)
Kirk-Othmer Encylopaedia of Chemical
Technology, 4th Edn. Vol 11, p. 783.
ASTM D892-98 Standard Test Method for Foaming Characteristics
of Lubricating Oils
(14)
ASTM D3427-99
Standard Test Method for Air Release Properties of Petroleum Oils
(15) T Sisson and F Simpson. (1984).
Proceedings 1984 International Compressor Engineering Conference at Purdue,
p273 - 284.