System Considerations for Refrigerant Blends

with Temperature Glide

Theory and Practical Retrofit Guidance
 Table of Contents
1 Blend and Glide Basics . . . . . . . . . . . . . . . . . . . . . . . 3
1.1 Superheat and Subcooling of Refrigerant Blends�� 4
1.2 Temperature Glide and the Pressure-Enthalpy
Diagram�������������������������������������������������������������� 4
1.3 Superheat Control with Glide ������������������������������ 7

2 Thermostatic Charges and Capacity for Alternate

Refrigerants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1 Thermostatic Charges for High-Glide Alternates�� 9
2.2 Capacity of Alternate Refrigerants���������������������� 11

3 System Retrofits to R407A, R407F, R448A, or

R449A: Liquid Side. . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.1 TEV Capacity when Retrofitting from R404A. . . . 13
3.2 TEV Capacity when Retrofitting from R22. . . . . . 15
3.3 Liquid Line Solenoid Valve Capacity when
Retrofitting from R404A . . . . . . . . . . . . . . . . . . . 16
3.4 Liquid Line Solenoid Valve Capacity when
Retrofitting from R22. . . . . . . . . . . . . . . . . . . . . . 17

4 System Retrofits to R407A, R407F, R448A, or

R449A: Suction Side. . . . . . . . . . . . . . . . . . . . . . . . . 18
4.1 EPR Capacity when Retrofitting from R404A
or R22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.2 EEPR Capacity when Retrofitting from R404A
or R22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.3 Suction Solenoid Valve Capacity when
Retrofitting from R404A or R22. . . . . . . . . . . . . . 21

5 Other Considerations and Retrofit Summary . . . . 22

5.1 System Filtration. . . . . . . . . . . . . . . . . . . . . . . . . 22
5.2 Oil Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.3 Elastomeric Seals. . . . . . . . . . . . . . . . . . . . . . . . 23
5.4 Summary for Retrofits to R407A, R407F,
R448A, or R449A. . . . . . . . . . . . . . . . . . . . . . . . 24

2
System Considerations for Refrigerant
Blends with Temperature Glide
Theory and Practical Retrofit Guidance for Low and Medium
Temperature Refrigeration

BLEND AND GLIDE BASICS

1 Refrigerant choices for refrigeration systems are undergoing

significant change. The phase-out of R22 is well into its final
few years. Severely restricted now, production and importation
of new R22 in the US will cease as of 2020. EPA Significant New Alternatives
Policy (SNAP) Rule 20, issued in July 2015 also sets a timeline for de-listing
R404A, R507, and others as acceptable alternatives for many commercial
refrigeration applications. Some of the first de-listings concerning system
retrofits took effect July 20, 2016. EPA has also approved (or SNAP listed)
several alternative HFC and HFO-blend refrigerants that may be used to
retrofit existing systems, as well as for new systems. Most of these blends
are viable replacements for R22, R404A, and similar refrigerants, but have
properties that require careful consideration to ensure a successful retrofit.
John Withouse
1.1 SUPERHEAT AND SUBCOOLING OF REFRIGERANT BLENDS Senior Principal Engineer
Parker Hannifin, Sporlan Division
The refrigeration industry has been using refrigerant blends for many years.
Some of these are azeotropic, in which the blend constituents form a mixture
email: [Link]@[Link]
that changes phase at a constant temperature at a given pressure. This is the
phone: 636 392 3335
same behavior as a pure refrigerant, made up of only one type of molecule.
In the evaporator, as the refrigerant boils off from a saturated liquid to a
saturated vapor, no sensible heat is gained by the refrigerant until it is 100%
vapor. This is a desirable characteristic for most refrigeration systems. Pure
refrigerants and azeotropes have a single saturation temperature at a given
pressure. Superheat and subcooling are simply calculated from this single-
point relationship. Azeotropic blends are assigned ASHRAE R-numbers in
the 500-series, such as R502 or R507.

Other blends are zeotropic, in which the blend constituents change

temperature with change of phase at constant pressure. This behavior is
commonly called temperature glide. As a zeotrope boils off in the evaporator,
the constituent with the highest vapor pressure will evaporate first. While the
first constituent is evaporating, the remainder of the constituents are gaining
sensible heat. As the temperature of the refrigerant increases, the vapor
pressure of the next constituent will then be reached, and it will evaporate,
while the remaining constituents (including those already evaporated) will
continue to gain sensible heat. This will continue for all constituents, until
all are boiled off. At a constant pressure, the saturated vapor of the blend
will be at a higher temperature than the saturated liquid, thus the term
“temperature glide”. This also means that at a given pressure, there are now
two saturation temperatures to consider: saturated liquid temperature (also
known as “bubble point”) and saturated vapor temperature (also known
as “dew point”). Because of this behavior, superheat and subcooling for
zeotropes must be calculated using the proper point. Zeotropic blends are
assigned R-numbers in the 400-series, such as R404A, R407C, R410A, etc.

3
Because having little or no glide Saturation pressure temperature 1.2 TEMPERATURE GLIDE AND
is favorable for heat exchanger relationships for R507 (an THE PRESSURE-ENTHALPY
design, zeotropes with low glide azeotrope), R404A (a low-glide DIAGRAM
have been more widely used in zeotrope) and R407A (a high-glide
the past. In the case of blends zeotrope) are shown below. This is a Temperature glide and subcooling
such as R404A or R408A that have useful way to visualize temperature and superheat calculations can
only around 1°F temperature glide, and some of the differences also be visualized on a pressure-
glide, there is very little effect on between these refrigerant types. enthalpy (P-h) diagram. Figure 2
heat exchanger performance. The focus is on typical medium is a P-h diagram for R407A, and
Even ignoring glide introduces temperature refrigeration illustrates the refrigeration cycle
only a small error into superheat evaporator range, from 10 to 30°F. and proper calculations for a high-
and subcooling calculation. With glide refrigerant. As with all P-h
the current movement toward On this plot, temperature glide is diagrams, the liquid saturation line
more environmentally-friendly the horizontal distance (meaning at forms the left side of the saturation
refrigerants, many blends with fixed pressure) between the bubble dome, and the vapor saturation line
significantly higher temperature point and dew point lines of the forms the right side of the dome.
glide (on the order of 10°F) are zeotropic blends. Azeotropic R507 The lines of constant temperature
commonly being used, and more has a single saturation line, and has (or isotherms) slope downward
are being introduced to the market. no glide. R404A has just over 1°F left-to-right in the saturated region,
These include but are not limited to glide, while R407A has over 10°F showing that R407A is a zeotrope,
R407A, R407C, R407F, R448A, and temperature glide. having temperature glide. For a pure
R449A. With high-glide zeotropic refrigerant or azeotrope, these lines
refrigerants, care must be taken would have no slope: no change
to use bubble point to calculate in temperature at a given pressure
subcooling, and dew point to during phase change. Subcooling
calculate superheat. and superheat are always calculated
from the nearest saturation state.

85

R507 Saturated
80 R404A Bubble Point R507 – Single P-T Relationship, No Glide
R404A Dew Point
R407A Bubble Point
75 R407A Dew Point
Pressure (psia)

70
R404A Temperature Glide ~1.1°F

65

60 R407A Temperature Glide ~10.2°F

55
10 12 14 16 18 20 22 24 26 28 30
Saturation Temperature (°F)

Figure 1 - Temperature Glide

4
 T = 80°F T = 100°F T = 120°F T = 140°F
T = 20°F T = 40°F T = 60°F T = 160°F
T = -20°FT = 0°F
T = -40°F Liquid Saturation R407A P-h Diagram
(Bubble Point)
Liquid Subcooling =
Bubble Point Temp. @Phigh
Vapor Saturation
– Liquid Temp. (Dew Point)

Phigh Condensing

Discharge Superheat =
Discharge Gas Temp.
Pressure (psia)

– Dew Point Temp. @Phigh

Expansion (∆h=0)
100

Compression
Plow

Evaporating

Suction Superheat =
Suction Gas Temp.
Isotherm’s with slope in the saturated
– Dew Point Temp. @Plow
region shows that the refrigerant is
zeotropic, and thus has temperature glide
q = 0.3 q = 0.4 q = 0.6 q = 0.7 q = 0.9 q=1
q = 0.1 q = 0.2 q = 0.5 q = 0.8

0.000 25.0 50.0 75.0 100. 125.

Enthalpy (Btu/bm)

Figure 2 - R407A P-h Diagram

Figure 2 depicts a R407A refrigeration system operating at medium

temperature conditions. The measured high and low side system pressures
are Phigh = 252 psia and Plow = 57 psia. Using the proper saturation states:

Liquid Subcooling Calculation:

Measured liquid line temperature = 80°F
From NIST Refprop, R407A bubble point @ 252 psia = 100°F
Liquid Subcooling = 100°F – 80°F = 20°F
Suction Superheat Calculation:
Measured suction gas into the compressor = 65°F
From NIST Refprop, R407A dew point @ 57 psia = 20°F
Suction Superheat = 65°F – 20°F = 45°F
Discharge Superheat Calculation:
Measured compressor discharge gas = 165°F
From NIST Refprop, R407A dew point @ 252 psia = 108°F
Discharge Superheat = 165°F – 108°F = 57°F

5
The calculated numbers can be checked with the P-h diagram, by scaling
them with the isotherms. In the case of liquid subcooling, that process
extends from the bubble point, leftward into the subcooled liquid region.
It extends from the 100°F isotherm to the 80°F isotherm, confirming our
calculation of 20°F subcooling.

To show the effect of using the incorrect saturation state,

consider Suction Superheat calculated using bubble point:
Measured suction gas into the compressor = 65°F
From NIST Refprop, R407A bubble point @ 57 psia = 10°F
Suction Superheat = 65°F – 10°F = 55°F

An additional 10°F superheat would be calculated that does not actually

exist. This amount of error in calculating superheat at the outlet of an
evaporator coil would lead to an improperly adjusted TEV and liquid
flooding in the suction line, possibly all the way back to the compressor. We
also readily see this is incorrect by checking our calculation against the P-h
diagram – the closest saturation state is dew point, not bubble point.

6
 1.3 SUPERHEAT CONTROL WITH GLIDE
The thermostatic expansion valve (TEV) has only one function in the
system: to meter the proper amount of refrigerant to the evaporator coil
by controlling superheat at the bulb location. In order to do this, the TEV
is located at the coil inlet. However, conditions at the coil outlet indicate
whether the TEV is working properly. For zeotropic refrigerants, superheat
is calculated based on dew point, the closest saturation condition at the
evaporator outlet. It follows naturally that superheat is also controlled
based on this condition. This means the vapor pressure or dew point of the
refrigerant is used to control the TEV. The following diagram and explanation
of the fundamental operation of TEVs is excerpted from Sporlan Bulletin
10-9, Thermostatic Expansion Valves: Theory of Operation, Application,
and Selection.

In order to understand the principles of thermostatic

expansion valve operation, a review of its major
components is necessary. A sensing bulb is connected
to the TEV by a length of capillary tubing which
transmits bulb pressure to the top of the valve’s Bulb
diaphragm. The sensing bulb, capillary tubing, and 1 Pressure
diaphragm assembly is referred to as the thermostatic
element. The thermostatic element on all standard
Sporlan TEVs is replaceable. 2 Evaporator

3
Pressure

e
2+
The diaphragm is the actuating member of the valve.

rc
Fo

e
rc
Its motion is transmitted to the pin and pin carrier

ng

Fo
si

ng
ni

lo
assembly by means of one or two pushrods, allowing 3 Spring

C
Pressure pe
the pin to move in and out of the valve port. The O Refrig .
Curve 2
superheat spring is located under the pin carrier, and
a spring guide sets it in place. On externally adjustable
valves, an external valve adjustment permits the spring 1
2
Closing Force

pressure to be altered.

There are three fundamental pressures acting on the

valve’s diaphragm which affect its operation: sensing Evaporator Temperature Superheat
bulb pressure P1, equalizer pressure P2, and equivalent
spring pressure P3 (see Figure 1). The sensing bulb
pressure is a function of the temperature of the Bulb Temperature
thermostatic charge, i.e., the substance within the bulb. Figure 1
This pressure acts on the top of the valve diaphragm
causing the valve to move to a more open position. The
equalizer and spring pressures act together underneath the diaphragm causing the valve to move to a more
closed position. During normal valve operation, the sensing bulb pressure must equal the equalizer pressure
plus the spring pressure, i.e.:
P1 = P2 + P3
Equivalent spring pressure is defined as the spring force divided by the effective area of the diaphragm. The
effective area of the diaphragm is simply the portion of the total diaphragm area which is effectively used by
the bulb and equalizer pressures to provide their respective opening and closing forces. Equivalent spring
pressure is essentially constant once the valve has been adjusted to the desired superheat. As a result, the
TEV functions by controlling the difference between bulb and equalizer pressures by the amount of the spring
pressure.

7
Based on the characteristics of high-glide zeotropic blends, we see the
evaporator pressure referred to is in fact vapor pressure or dew point.
Considering the three pressures acting on a TEV, P1 (bulb pressure) and P2
(vapor pressure) are acting on opposite sides of the same diaphragm. TEVs
are designed for these two pressures to be very close to one another. These
will then be the fundamental forces acting to position the pin in the port,
producing the best superheat response. Since each of these two similar
pressures is acting on the same area, only a small amount of additional
force (via spring pressure, P3) is needed to fine tune the balance through
adjustment. This is illustrated in the above figure by the small size of P3 in
comparison to the other two forces. The stroke of the valve is relatively very
small, and the force the spring needs to provide is very small, so the spring
constant is low. The adjustment assembly is typically designed to provide
8-12 full rotations from minimum to maximum design spring force. This
allows for virtually infinite tuning, but only within the design range of the
spring. It follows that the thermostatic element charge should be designed
to be close to the vapor pressure (dew point) of a zeotropic refrigerant to
provide the best TEV performance.

What if the thermostatic element charge is significantly lower in pressure

than the vapor pressure at the evaporator outlet? The opening force on
the TEV will be very low, and the closing forces will keep the pin and port
closed until the bulb gets much warmer, resulting in inadequate refrigerant
feed to the evaporator and high superheat. Adjusting spring pressure lower
will help compensate by balancing the forces at a more open position. If
evaporator outlet pressure is too much higher, even removing all spring
pressure may not be sufficient. The TEV will be operating outside its design
parameters, desired superheat may not be obtained, and the evaporator will
be fed too little refrigerant for its load.

What if the thermostatic element charge is significantly higher in

pressure than the vapor pressure at the evaporator outlet? The opening
force on the TEV will be very high, moving the pin and port more open, and
flooding will occur. Adjusting spring pressure higher will help compensate
by balancing the forces at a less open position. However, if too much
additional spring pressure is needed, maximum design spring force may
be reached, and the valve will be operating outside its design parameters.
Proper superheat adjustment may not be possible, and superheat response
will be compromised. Liquid flood back may occur, particularly at high
condensing pressure conditions. TEVs, like most control systems, operate
best in the middle portion of their design range.

THERMOSTATIC CHARGES AND CAPACITY

2 FOR ALTERNATE REFRIGERANTS

Many alternates to R404A, R507, and similar refrigerants have
been developed. Most of these can also be used as alternates
to R22 in refrigeration applications. At this time, the alternates that are
generating the greatest interest and gaining market momentum include
R407A, R407F, R448A, and R449A. R407A and R407F are HFC blends
containing R32, R125, and R134a. R448A and R449A are blends of HFC
and HFO constituents, containing R32, R125, R134a, and R1234yf. R448A
also contains a small percentage of R1234ze. These alternate refrigerants
all have similar temperature glide, in the range of 10°F, due to the common
constituents among them. Section 2 examines how these alternates

8
 100

90
R404A Dew Point
80 R407A Dew Point
70 R407F Dew Point
Pressure (psia)

60

50

40

30

20

10

0
-40 -30 -20 -10 0 10 20 30 40

Saturation Temperature (°F)

Figure 3 - Vapor P-T Relationships, R404A-like HFC's

compare to R404A and R22 from During the early stage of The vapor pressures of R407A and
the standpoint of TEV thermostatic thermostatic charge development, R407F are similar to one another,
charges and volumetric capacity. Sporlan engineers will compare but both are significantly lower
the vapor pressure-temperature than R404A. Is there an existing
2.1 THERMOSTATIC relationship of the new refrigerant refrigerant that may be a better
CHARGES FOR HIGH-GLIDE to that of a refrigerant with match, for which thermostatic
ALTERNATES existing thermostatic charges. charges have already been
If they are a close match, that designed? Figure 4 adds R22 to the
Let’s relate the design and is a strong indication that an same chart.
operational characteristics of existing thermostatic charge will
TEVs to some of the high-glide also function well with the new The saturation curve of R22 is a
refrigerants. Over the past few refrigerant. Figure 3 compares much closer match to the vapor
decades, many different refrigerants the vapor pressure-temperature pressure curves of R407A and R407F.
have been used. Sporlan has relationships of R407A and R407F Consider typical low and medium
developed thermostatic charges with R404A, across the normal evaporator operating temperatures
to function properly with most ranges of low and medium of -20°F and 20°F:
of these. Since the charge in the temperature refrigeration.
thermostatic element does not
function as a heat transfer fluid,
we are primarily concerned that its TEMPERATURE VAPOR PRESSURE (PSIA)
°F R404A R407A R407F R22
pressure-temperature relationship
is close to the vapor pressure- -20 30.7 23.2 24.4 24.9
temperature relationship of the 20 70.0 57.0 59.8 57.8
system refrigerant.

9
 100

90
R404A Dew Point
80 R407A Dew Point
70 R407F Dew Point
Pressure (psia)

R22 Saturated
60

50

40

30

20

10

0
-40 -30 -20 -10 0 10 20 30 40

Saturation Temperature (°F)

Figure 4 - Vapor P-T Relationships, R404A-like HFC's

R22 saturation pressure varies by less than +/- 1 psi at -20°F, and by +/- 2 psi at 20°F. This indicates thermostatic
charges for R22 will likely be applicable to R407A and R407F. However, the vapor pressure of R404A exceeds those of
R407A and R407F on average by 7 psi at -20°F and nearly 12 psi at 20°F.

Figure 5 is the same chart as above, but the vapor pressure curves of R407A and R407F are replaced with those of
R448A and R449A. It should be noted that in Figure 5, the dew point curve of R448A lies directly under that of R449A.
Though it may appear that it was omitted, these two are instead so similar they cannot be distinguished at the scale
presented.

100

90
R404A Dew Point
80 R448A Dew Point
70 R449A Dew Point
Pressure (psia)

R22 Saturated
60

50

40

30

20

10

0
-40 -30 -20 -10 0 10 20 30 40

Saturation Temperature (°F)

Figure 5 - Vapor P-T Relationships, R404A-like HFO Blends

10
The saturation curve of R22 appears to be an even closer match to the vapor
pressure curves of R448A and R449A. Looking more closely at the -20°F and
20°F points:

TEMPERATURE VAPOR PRESSURE (PSIA)

°F R404A R448A R449F R22
-20 30.7 24.5 24.6 24.9
20 70.0 59.3 59.1 57.8

We find a variation of only +0.4 psi at -20°F, and -1.5 psi at 20°F for R22.
R404A vapor pressure exceeds those of R448A and R449A by 6 psi at -20°F
and by 11 psi at 20°F. These straightforward comparisons strongly indicate
that thermostatic charges designed for R22 will function well for R407A,
R407F, R448A, and R449A throughout the range of low and medium
temperature refrigeration applications. An extensive program of charge
characteristic testing performed in the Sporlan Engineering Lab confirmed
both low and medium temperature thermostatic charges for R22 are also the
best choice for R407A, R407F, R448A, and R449A.

What is the effect of using thermostatic charges designed for R404A with
these refrigerants? We have established that small variations in vapor
pressure from one refrigerant to another, around 5% or less, pose no issue
with TEV adjustment or operation. What if that variation is on the order of
20%? From the TEV balance equation (P1 = P2 + P3), if P1 becomes much
larger than P2, P3 must increase to compensate. However, we also know
that:
a) P3 should be kept small so that P1 and P2 are the fundamental forces
acting to position the pin;
b) By adjusting P3 very high the design range of the spring may be
exceeded;
c) The TEV will not be operating in its effective design range, where it will
provide the best superheat response.

Thermostatic charges designed for refrigerants with much higher vapor

pressure, including R404A, should not be used with R407A, R407F, R448A,
and R449A. Thermostatic charges designed for R22 are the proper choice.
Conversely, lower-pressure thermostatic charges for R22, R407A, R407F,
R448A, or R449A should not be applied with higher vapor pressure
refrigerants such as R404A.

2.2 CAPACITY OF ALTERNATE REFRIGERANTS

The capacity (Q) of a direct expansion heat exchanger is calculated by the
following equation:
Q = ṁ(hout – hin), where:
ṁ = mass flow rate of the refrigerant;
hout = the enthalpy of the refrigerant leaving the heat exchanger;
hin = the enthalpy of the refrigerant entering the heat exchanger

Refrigeration systems are almost always designed to a specific cooling

capacity, so the Q of greatest interest is the capacity of the evaporator. The
(hout – hin) part of the equation is also known as net refrigerating effect, or
NRE. It is closely approximated by subtracting saturated liquid enthalpy
into the TEV from saturated vapor enthalpy at the evaporator pressure.
Refrigerant choice and high and low side operating conditions determine
NRE. So for a given Q, ṁ will be determined by the NRE of the refrigerant.

11
Let’s look at the NRE of some of the alternates, using 90°F saturated liquid,
and saturated vapor at -20°F and 20°F, calculated by NIST Refprop 9.1. We
will compare with R22 and R404A.

EVAPORATOR NET REFRIGERATING EFFECT (Btu/lbm)

TEMPERATURE °F R404A R407A R407F R448A R449A R22
-20 44.1 57.5 64.8 59.0 58.2 66.1
20 49.4 62.6 69.7 64.1 63.3 70.1

Using NRE, we will calculate the required mass flow of each refrigerant per
ton (12,000 Btu/hr) capacity.

EVAPORATOR MASS FLOW PER TON (lbm/min)

TEMPERATURE °F R404A R407A R407F R448A R449A R22
-20 4.54 3.48 3.09 3.39 3.44 3.03
20 4.05 3.19 2.87 3.12 3.16 2.85

There is quite significant variation in mass flow required per ton capacity at
the same conditions for these refrigerants. The mass flow of R407F is very
close to that of R22, and both are ~32% lower than R404A. The mass flow of
R407A, R448A, and R449A are similar, and are ~23% lower than R404A, but
still ~12% higher than R22.

It is also useful to compare the volumetric flow of these refrigerants, in

order to better understand the sizing of liquid and suction control valves. All
valves fundamentally control flow by volume, not mass directly. Density or
specific volume is required to calculate volumetric flow. We will use -20°F
low temperature saturated evaporating, and 20°F medium temperature
saturated evaporating, and 100°F condensing with 10°F subcooling for both.
So as not to overly complicate the analysis, high and low side pressure drops
will be ignored. From Refprop 9.1, densities for -20°F saturated vapor, 20°F
saturated vapor, and 90°F saturated liquid are:

CONDITION REFRIGERANT DENSITY (lbm/ft3)

°F R404A R407A R407F R448A R449A R22
-20 Sat. Vapor 0.69 0.47 0.45 0.48 0.48 0.48
20 Sat. Vapor 1.53 1.11 1.06 1.11 1.12 1.07
90 Sat. Liquid 62.92 69.34 67.64 66.42 66.41 72.52

Mass flow per ton is divided by liquid and vapor density to determine
the liquid volumetric flow and vapor volumetric flow per ton for each
refrigerant, at low and medium temperature system operating conditions.

LOW TEMP VOLUMETRIC FLOW PER TON (ft3/min)

(-20/100) R404A R407A R407F R448A R449A R22
Vapor 6.58 7.40 6.87 7.06 7.17 6.31
Liquid 0.072 0.050 0.046 0.051 0.052 0.042

12
 MEDIUM TEMP VOLUMETRIC FLOW PER TON (ft3/min)
(20/100) R404A R407A R407F R448A R449A R22
Vapor 2.65 2.87 2.71 2.81 2.82 2.66
Liquid 0.064 0.046 0.042 0.047 0.048 0.039

This comparison highlights the differences on the liquid side, and the
similarities on suction side of the system. The lower liquid density of R404A
in comparison with R22, along with higher mass flow requirements means
that an R404A liquid side valve must flow an even greater volume than a
simple mass flow comparison of these two would indicate. R407A, R407F,
R448A, and R449A are again in the middle, around 30% lower than R404A
and 20% higher than R22. On the suction side of the system, the refrigerants
are a much closer match. Suction volumetric flow of R404A and R22 are
very similar, with R407A, R407F, R448A, and R449A being 4 – 10% higher
than both at medium temperature and 9 – 17% higher at low temperature
conditions.

SYSTEM RETROFITS TO R407A, R407F,

3 R448A, OR R449A: LIQUID SIDE

There is much interest and significant activity with these
refrigerants for system retrofits, mostly due to regulatory
actions. These include the ongoing phase-out of R22, which will no longer
be able to be produced or imported into the US as of January 1, 2020, and
the pending removal of R404A and similar refrigerants from EPA’s approved
list of refrigerants for retrofit supermarket systems (July 20, 2016), new
supermarket systems (January 1, 2017), and new condensing units (January
1, 2018). Refrigerant will still be available for service into the foreseeable
future, but prices will rise. Blends like R404A present more difficulties with
reclaiming and recycling than pure refrigerants such as R22.

When considering retrofits of refrigeration systems, we will assume that one

goal is to replace as few system components as possible, in the interest of
reducing both cost and system downtime. Section 3 examines how R407A,
R407F, R448A, and R449A may function with existing valves. The large
differences in mass flow and liquid volumetric flow rate between R404A and
R22 noted in Section 2.2 mean that retrofits of each to alternate refrigerants
must be addressed separately.

The valve sizing and selections in all the subsequent sections of this
document are carried out with Version 5.09.15 of the Sporlan Selection
Program. This is proprietary software that employs the NIST Refprop DLL for
all refrigerant property calculations. Flow and loading calculations are based
on experimentally determined valve characteristics, and results have been
experimentally validated. In the following sections, results that are outside
the normally accepted range are highlighted in red.

3.1 TEV CAPACITY WHEN RETROFITTING FROM R404A

These refrigerants are mostly used in systems having multi-circuit
evaporator coils, which requires the use of a refrigerant distributor. Pressure
drop across the distributor reduces the pressure drop across the TEV, so that
must be determined before selecting TEVs. The Sporlan selection program
will be used to calculate the distributor nozzle and tube pressure drop for

13
use in selecting TEVs. We will use a 1 ton (12,000 Btu/hr) load, and assume
a 3 circuit evaporator with 3/16” OD, 18” long distributor tubes. From this
a nozzle is sized for R404A, then that nozzle size is applied to the alternate
refrigerants. The selection program will calculate tube ΔP, nozzle ΔP, and
distributor total ΔP. Distributor total ΔP is then subtracted from system high-
to-low ΔP to obtain ΔP across the TEV, the driving force for flow across the
TEV.

LOW TEMP REFRIGERANT DISTRIBUTOR

(-20/100) R404A R407A R407F R448A R449A
Orifice # 2-1/2 2-1/2 2-1/2 2-1/2 2-1/2
Tube ΔP (psi) 28 22 19 21 22
Nozzle ΔP (psi) 17 13 11 13 13
Dist ΔP (psi) 45 35 30 34 35
Hi-Lo ΔP (psi) 221 229 240 234 230
TEV ΔP (psi) 176 194 210 200 195

MEDIUM TEMP REFRIGERANT DISTRIBUTOR

(20/100) R404A R407A R407F R448A R449A
Orifice # 1-1/2 1-1/2 1-1/2 1-1/2 1-1/2
Tube ΔP (psi) 14 11 9 10 11
Nozzle ΔP (psi) 23 18 15 17 18
Dist ΔP (psi) 36 28 24 27 29
Hi-Lo ΔP (psi) 182 195 204 199 196
TEV ΔP (psi) 146 167 180 172 167

For a given distributor orifice size, the total pressure drop across it for the
alternates is lower by 10 – 15 psi at low temperature conditions, and 8 – 12
psi at medium temperature conditions. While these lower distributor ΔP’s
are still adequate for the function of the distributor, there is more ΔP across
the TEV for R407A, R407F, R448A, and R449A. Using the R404A distributor
ΔP, a TEV is selected for R404A. The same TEV (same port size, pin angle,
and stroke) is then applied to the alternate refrigerants, employing their
distributor ΔP’s, in order to compare sizing.

LOW TEMP R404A TEV SELECTION: FBSE-2

(-20/100) R404A R407A R407F R448A R449A
Dist ΔP (psi) 45 35 30 34 35
TEV ΔP (psi) 176 194 210 200 195
TEV % loading 64 44 38 43 44

MEDIUM TEMP R404A TEV SELECTION: FBSE-1

(20/100) R404A R407A R407F R448A R449A
Dist ΔP (psi) 36 28 24 27 29
TEV ΔP (psi) 146 167 180 172 167
TEV % loading 65 45 40 44 45

From this comparison, we see that a TEV and distributor sized properly for
R404A may be oversized for the alternates. Percent loading for R407F
reduces approximately 25%, the others reduce approximately 20%. While

14
there are no hard rules that apply to all TEVs on all systems, loadings below
50% begin to become a concern. Mass and volumetric flow requirements for
R407A, R407F, R448A, and R449A are lower, yet there is greater ΔP across the
TEV, to drive flow higher. Keep in mind the TEV is sized at a nominal
condition. Under low-load conditions, the TEV will be even more oversized.
When thermostatic charge is also considered (as discussed in Section 2.1),
there is a high risk that existing R404A TEVs will provide poor superheat
control when retrofitted to R407A, R407F, R448A, or R449A. There is also a
high risk of liquid flood back.

Replacement of TEVs may be the best choice in most retrofits from R404A
to R407A, R407F, R448A, or R449A. Most existing distributors will perform
acceptably well. Sizing of existing R404A distributors and TEVs should be
checked before replacement.

3.2 TEV CAPACITY WHEN RETROFITTING FROM R22

The same exercise can be performed with an R22 system. We will again use
a 1 ton (12,000 Btu/hr) load, and assume a 3 circuit evaporator with 3/16”
OD, 18” long distributor tubes. A nozzle is sized for R22, then applied to the
alternate refrigerants.

LOW TEMP REFRIGERANT DISTRIBUTOR

(-20/100) R22 R407A R407F R448A R449A
Orifice # 1-1/2 1-1/2 1-1/2 1-1/2 1-1/2
Tube ΔP (psi) 17 22 19 21 21
Nozzle ΔP (psi) 21 27 24 27 28
Dist ΔP (psi) 38 49 43 48 49
Hi-Lo ΔP (psi) 185 229 240 234 230
TEV ΔP (psi) 148 180 197 186 181

MEDIUM TEMP REFRIGERANT DISTRIBUTOR

(20/100) R22 R407A R407F R448A R449A
Orifice # 1 1 1 1 1
Tube ΔP (psi) 8 11 9 10 11
Nozzle ΔP (psi) 23 30 26 30 30
Dist ΔP (psi) 31 41 35 40 41
Hi-Lo ΔP (psi) 153 195 204 199 196
TEV ΔP (psi) 122 154 169 159 155

With a distributor orifice sized for R22, total pressure drop for the alternates
is higher by 4 – 11 psi at both low and medium temperature conditions. Yet
because of their significantly greater total high-to-low side ΔP, there is still
as much as 47 - 49 psi greater ΔP across the TEV at both low and medium
temperature conditions. Using the R22 distributor ΔP, a TEV is selected for
R22. The same TEV (same port size, pin angle, and stroke) is then applied to
the alternate refrigerants, in order to compare sizing.

15
 LOW TEMP R22 TEV SELECTION: FBVE-2
(-20/100) R22 R407A R407F R448A R449A
Dist ΔP (psi) 38 49 43 48 49
TEV ΔP (psi) 148 180 197 186 181
TEV % loading 66 70 60 68 69

MEDIUM TEMP R22 TEV SELECTION: FBVE-1

(20/100) R22 R407A R407F R448A R449A
Dist ΔP (psi) 31 41 35 40 41
TEV ΔP (psi) 122 154 169 159 155
TEV % loading 75 76 66 81 83

With a TEV and distributor sized properly for R22, percent loading changes
little for the alternates. Percent loading for R407F is somewhat lower at both
medium and low temperature conditions, but well within an acceptable
range. Loadings for R407A, R448A, and R449A are just 2 – 4% percent higher
at low temperature conditions. At medium temperature conditions, loading
with R448A and R449A increases by 6 – 8%. All these would be considered
within an acceptable range. Coupled with the knowledge that thermostatic
charges for R22 are the proper choice for these refrigerants as well, there is
a high likelihood that existing R22 TEVs and distributors will provide good
superheat control when retrofitted to R407A, R407F, R448A, or R449A. Sizing
should be checked to confirm this.

3.3 LIQUID LINE SOLENOID VALVE CAPACITY WHEN

RETROFITTING FROM R404A
Another commonly used control valve in refrigeration systems is the
solenoid valve. These are most commonly used to control liquid flow, but are
also sometimes used to control suction gas flow. We will first look at liquid
line solenoid valves. Unlike an expansion valve, the purpose of which is to
precisely meter refrigerant to the evaporator, the solenoid valve’s purpose
is to simply provide electrically actuated on/off refrigerant flow control.
Most solenoid valves are “piloted”. They are opened by a coil which applies
electromagnetic pull on the stem, but rely on a light spring and a small
amount of pressure drop across the valve to close it. At sizing conditions this
pressure drop should be 1 psi or greater. If a solenoid valve is too large, there
may not be enough ΔP across it to initially close it, particularly at low load
conditions.

If a solenoid valve is too small, there will be more pressure drop across it
than is desired. This causes loss of subcooling in a liquid line. If there is very
little subcooling, liquid can flash across the valve. A general guideline is that
the pressure drop should be no greater than 1°F of saturation (bubble point)
drop for the system refrigerant. For all the refrigerants in our discussion, 1°F
of saturation is 3.0 – 3.5 psi at 100°F.

For TEVs and distributors, we used 1 ton (12,000 Btu/hr) loads in our
comparisons. This is appropriate because there is typically one TEV per case
or unit cooler in a system. Often one solenoid valve will control flow on a
system branch supplying liquid to a few or several cases or unit coolers. For
solenoid comparisons, we will size for a two ton (24,000 Btu/hr) load at low
temperature conditions (-20°F evaporator, 100°F condensing) and a four ton
(48,000 Btu/hr) load at medium temperature conditions (20°F evaporator,

16
100°F condensing). A liquid line solenoid valve will be selected for R404A,
then that valve will be applied to R407A, R407F, R448A, and R449A to see
the effect on sizing. Selecting a solenoid valve for R404A that will have ΔP
greater than 1 psi, but less than 3.5 psi:

LOW TEMP R404A LL SOLENOID SELECTION: E6

(-20/100) R404A R407A R407F R448A R449A
Flow (lbm/min) 9.07 6.96 6.17 6.79 6.88
Sol ΔP (psi) 1.3 0.7 0.6 0.7 0.7
Sol % loading 66 48 42 48 49

MEDIUM TEMP R404A LL SOLENOID SELECTION: E8

(20/100) R404A R407A R407F R448A R449A
Flow (lbm/min) 16.2 12.8 11.6 12.5 12.6
Sol ΔP (psi) 2.6 1.5 1.2 1.4 1.5
Sol % loading 92 68 61 68 69

An E6 liquid line solenoid valve is selected for the R404A low temperature
system, with 1.3 psi ΔP. This is on the low end of the preferred range, so what
if the next smaller size E5 is selected instead? ΔP would be 4.2 psi instead,
with 117% loading. In this case, either would likely function without issue.
Usually the lower ΔP would be chosen for an actual system, so the E6 will be
used for comparison. As shown in the table above, that choice results in low
ΔP for all the alternate refrigerants. The situation is different for the medium
temperature system with a four ton load. An E8 solenoid valve is chosen
for R404A, with 2.6 psi ΔP. When applied with the alternate refrigerants,
ΔP’s from 1.2 to 1.5 psi are calculated. These are obviously lower, but still
acceptable.

When retrofitting an existing R404A system to R407A, R407F, R448A, or

R449A, liquid line solenoid valve sizing should be checked. If ΔP will be
lower than 1 psi with the alternate refrigerant, appropriately sized solenoid
valves should be selected to replace the existing ones.

3.4 LIQUID LINE SOLENOID VALVE CAPACITY WHEN

RETROFITTING FROM R22
Let’s perform the same comparison with R22 as a baseline, again using
a two ton (24,000 Btu/hr) load at low temperature conditions and a four
ton (48,000 Btu/hr) load at medium temperature conditions. A liquid line
solenoid valve will be selected for R22, then that valve will be applied to
R407A, R407F, R448A, and R449A to see the effect on sizing. 1°F saturation
drop for R22 at 100°F is 2.9 psi, so we select a solenoid valve for R22 that will
have ΔP greater than 1 psi but less than 3 psi.

LOW TEMP R22 LL SOLENOID SELECTION: E5

(-20/100) R22 R407A R407F R448A R449A
Flow (lbm/min) 6.05 6.96 6.17 6.79 6.88
Sol ΔP (psi) 1.6 2.2 1.8 2.2 2.3
Sol % loading 79 85 74 84 86

17
 MEDIUM TEMP R22 LL SOLENOID SELECTION: E6
(20/100) R22 R407A R407F R448A R449A
Flow (lbm/min) 11.4 12.8 11.6 12.5 12.6
Sol ΔP (psi) 1.8 2.4 2.0 2.4 2.5
Sol % loading 84 89 79 88 89

An E5 liquid line solenoid valve is selected for the R22 low temperature
system, with 1.6 psi ΔP. This would be common choice in an actual system.
As shown in the table above, that selection results in ΔP’s from 1.8 to 2.3 psi
for the alternate refrigerants. The E5 would function well, without excessive
ΔP for the alternate refrigerants. Results for the medium temperature system
mirror the low temperature system. An E6 solenoid valve is selected for R22,
with 1.8 psi ΔP. When applied with the alternate refrigerants, ΔPs of 2.0 to 2.5
psi are calculated, again well within an acceptable range.

When retrofitting an existing R22 system to R407A, R407F, R448A, or

R449A, existing liquid line solenoid valve sizing will in most situations be
acceptable, though it should be checked. If ΔP will be excessively high with
the alternate refrigerant, appropriately sized solenoid valves should be
selected to replace the existing ones.

SYSTEM RETROFITS TO R407A, R407F,

4 R448A, OR R449A: SUCTION SIDE

Section 2.2 notes the large differences in mass flow between
R404A and R22. This coupled with R22’s higher liquid density
means that on the liquid side of the system, a valve has to flow a much
greater volume of R404A than R22 to deliver a given amount of refrigeration
capacity. However, the density relationship between R22 and R404A reverses
for saturated vapor. R404A vapor is much denser than R22 vapor. Because of
this, the volume of vapor that a valve on the suction side of the system must
flow to deliver a given capacity is actually quite similar for both. Section
4 will examine how existing mechanical evaporator pressure regulators,
electric evaporator pressure regulators, and suction solenoid valves will
perform when a system is retrofitted from R22 or R404A to R407A, R407F,
R448A, or R449A.

4.1 EPR CAPACITY WHEN RETROFITTING FROM R22 OR R404A

An evaporator pressure regulator (EPR) valve is installed between the outlet
of the evaporator and the suction header of a supermarket refrigeration
system. Its function is to maintain the desired pressure in the evaporator while
suction pressure downstream of it is lower, so it is a “holdback” or open on
rise of inlet pressure device. Most evaporators in a system operate best when
set at a design pressure, and therefore operate at the proper temperature. The
evaporator will then deliver the proper temperature air to the case or walk-in
cooler, and not frost up too quickly. Systems typically have loads that require
different evaporator temperatures, but the lowest evaporator temperature
on a system dictates the common pressure at which the suction header must
be maintained. The EPR allows evaporators designed to operate at higher
temperatures to do so, and maintains this pressure within a small range to
buffer the evaporator from suction pressure fluctuations.

18
EPRs are sometimes used to control individual loads, but are more often
used to control suction pressure on a system branch or circuit with multiple
loads. For this reason, we will size for a two ton (24,000 Btu/hr) load at low
temperature conditions (-20°F evaporator, 100°F condensing) and a four ton
(48,000 Btu/hr) load at medium temperature conditions (20°F evaporator,
100°F condensing). We will assume the suction header is running at -25°F
saturated for the low temperature system, and at 17°F for the medium
temperature system. In an actual system, it would be typical to have 5°F
difference on a low temperature header and 3°F on a medium temperature
header. Sporlan offers EPRs with three different modes of operation: direct
acting, internally piloted, and externally piloted. Our comparison will
consider the externally piloted type, because it is the most common in this
application.

The suction vapor volumetric flow of R404A and R22 is about the same for
a given capacity under like conditions. Does that result in the selection of
the same EPR? When using the low temperature system parameters already
discussed, the Sporlan selection program gives the following results for
R404A and R22:

19
While percent loading varies some, the same SORIT-12 is chosen for both.
For medium temperature conditions:

A SORIT-15 is chosen for both, and loading is again within a few percent
difference. Let’s examine how these EPRs may function with the alternate
refrigerants, starting with the low temperature example from above.

LOW TEMP R404A/R22 EPR SELECTION: SORIT-12

(-25/-20/100) R404A R22 R407A R407F R448A R449A
ΔP Available (psi) 3.3 2.8 2.7 2.9 2.9 2.8
ΔP Required (psi) 3.0 1.8 2.6 2.1 2.4 2.4
EPR % Loading 95 83 97 86 92 93

Sizing for all the alternates is acceptable, with percent loading falling in
between that of R404A and R22 for most, and R407A being slightly higher.
For the medium temperature condition and load:

MEDIUM TEMP R404A/R22 EPR Selection: SORIT-15

(17/20/100) R404A R22 R407A R407F R448A R449A
ΔP Available (psi) 3.8 3.2 3.4 3.6 3.5 3.5
ΔP Required (psi) 1.4 1.0 1.2 1.0 1.2 1.2
EPR % Loading 62 57 61 55 59 60

20
Sizing for the alternates is acceptable at medium temperature conditions
as well, with little variation from that of R404A or R22. When retrofitting an
existing low or medium temperature R404A or R22 system to R407A, R407F,
R448A, or R449A, existing evaporator pressure regulator valve sizing will
likely be acceptable. It is good practice to check EPR sizing for confirmation.

While externally piloted EPRs were used in this example, there are many
internally piloted EPRs in use today, particularly on loop-piped systems.
The function of internally piloted EPRs can be more dependent on proper
sizing than externally piloted ones. Thus, they may be more sensitive to a
refrigerant retrofit. If a retrofit is to be performed on a system with internally
piloted EPRs, sizing must be checked with the alternate refrigerant to ensure
that sizing guidelines will not be exceeded. The manufacturer of the valve
will be able to provide guidance on how the existing valve may function with
an alternate refrigerant, and whether or not replacement is recommended.

4.2 EEPR CAPACITY WHEN RETROFITTING FROM R22 OR R404A

Electric evaporator pressure regulators are also widely used in supermarket
systems. EEPRs perform the same function as described in Section 4.1 for
an EPR, but are operated by an electric stepper motor, and controlled via
a pressure transducer and electronic controller. We also expect EEPRs for
R22 and R404A systems to be sized similarly. We will perform the same
comparison for our two ton low and four ton medium temperature systems
with EEPRs sized for R22 and R404A, then look at how the selected valves
might perform with the alternates.

LOW TEMP R404A/R22 EEPR SELECTION: CDS-7

(-25/-20/100) R404A R22 R407A R407F R448A R449A
ΔP Available (psi) 3.3 2.8 2.7 2.9 2.9 2.8
ΔP Required (psi) 1.6 1.0 1.4 1.1 1.3 1.3
EPR % Loading 72 63 74 65 70 71

MEDIUM TEMP R404A/R22 EEPR SELECTION: CDS-7

(17/20/100) R404A R22 R407A R407F R448A R449A
ΔP Available (psi) 3.8 3.2 3.4 3.6 3.5 3.5
ΔP Required (psi) 2.4 1.6 2.0 1.7 1.9 1.9
EPR % Loading 80 72 78 70 75 76

As with mechanical EPRs, EEPRs for R404A and R22 are acceptable for the
alternates. When retrofitting an existing low or medium temperature R404A
or R22 system to R407A, R407F, R448A, or R449A, existing electric evaporator
pressure regulator valve sizing will likely be acceptable. It is good practice to
check EEPR sizing for confirmation.

4.3 SUCTION SOLENOID VALVE CAPACITY WHEN RETROFITTING

FROM R22 OR R404A
The purpose of a suction solenoid valve is to provide electrically actuated
on/off refrigerant flow control. Most are piloted valves, as discussed
in Section 3.3. Suction solenoid valves are opened by the coil applying
electromagnetic pull on the stem, but require a small amount of pressure
drop across the valve to close it. At sizing conditions this pressure drop
should be 1 psi or greater. If a solenoid valve is too large, there may not be
enough ΔP across it to initially close it, particularly at low load conditions.

21
If a solenoid valve is too small, there will be more pressure drop across it
than is desired. In a suction line, pressure drop is a system inefficiency – it
causes the compressor to operate at a lower suction pressure than it would
otherwise. For the refrigerants in this exercise, 1.7 – 2.0 psi of drop is 3°F of
saturation at -20°F, and 2.1 – 2.6 psi of drop is 2°F saturation at 20°F. Because
of this, suction solenoid valves with no greater than 3°F saturation drop
at low temperature conditions, and no greater than 2°F saturation drop at
medium temperature conditions should be selected. Within this range, the
suction solenoid valve will function reliably without causing excessive loss of
system efficiency. These valves are then applied to the alternate refrigerants.

LOW TEMP R404A/R22 SUCTION SOLENOID SELECTION: E25

(-20/100) R404A R22 R407A R407F R448A R449A
Flow (lbm/min) 9.07 6.05 6.96 6.17 6.79 6.88
Sol ΔP (psi) 1.9 1.2 1.7 1.4 1.6 1.6
Sat Drop (°F) 2.7 2.2 2.5 2.0 2.3 2.3

MEDIUM TEMP R404A/R22 SUCTION SOLENOID SELECTION: E25

(20/100) R404A R22 R407A R407F R448A R449A
Flow (lbm/min) 16.2 11.4 12.8 11.5 12.5 12.6
Sol ΔP (psi) 2.7 1.9 2.4 2.0 2.2 2.3
Sat Drop (°F) 2.1 1.8 1.8 1.5 1.6 1.7

An E25 solenoid valve is chosen for both the two ton low temperature
load, and the four ton medium temperature load with R22 or R404A. When
applied to the alternate refrigerants, the goal of maintaining at least 1 psi
ΔP to ensure operation of the solenoid valve was met, while not excessively
reducing system efficiency. Saturation drop was acceptable for all
refrigerants at both medium and low temperature conditions.

When retrofitting an existing low or medium temperature R404A or R22

system to R407A, R407F, R448A, or R449A, existing suction solenoid valve
sizing will likely be acceptable. It is good practice to check suction solenoid
valve sizing for confirmation.

OTHER CONSIDERATIONS AND RETROFIT

5 SUMMARY
There are several other areas that must be considered when
performing a refrigerant retrofit. Among the most important of
these are filter changes, oil changes, and replacement of elastomeric seals.
This section will touch on these but not cover them exhaustively. There are
many good refrigerant retrofit guides that cover these topics in detail.

5.1 SYSTEM FILTRATION

Any system refrigerant retrofit should include a change of all system
filters, filter-driers or filter-drier cores, regardless of refrigerant. Clean
filters minimize pressure drop, reducing startup difficulty and maximizing
efficiency. It is best practice to check for significant pressure drop due to
contamination after a few hours of system operation, and change filters
again if significant contamination is found. Retrofits from R22/mineral oil

22
to HFC/POE oil are known to cause contamination deposits to dislodge and
move through the system, so filter changes are a very important step in any
retrofit of an R22 system. Filter manufacturers provide guidelines on filter
changes for system cleanup and for normal operation.

New filter-driers or cores will also help ensure the system is dry as possible,
benefiting system startup and operation. When the system is re-started
following a retrofit, liquid line moisture indicators should be reviewed after
24 hours of operation. If moisture is still indicated in the system, change the
filter-driers or cores again. This process should be repeated until the system
is indicated dry.

There will normally be little concern with the size of liquid or suction filters
when retrofitting from either R404A or R22 to R407A, R407F, R448A, or
R449A. Filters are usually selected with some amount of additional capacity.
It is good practice check the size of filter-driers, to confirm there will be both
adequate flow capacity and moisture removal capability in relation to the
size of the system. All Sporlan filter-driers, cores, and suction filters are fully
compatible with these alternate refrigerants.

5.2 OIL CHANGES

When retrofitting an R22 system to R407A, R407F, R448A, or R449A, the
existing mineral oil in the system must be changed to a synthetic POE oil.
Some R22 systems may have POE oil instead of mineral oil. Recommended
procedures for HCFC to HFC retrofits apply to the blends with HFO content
as well. Established guidelines call for reducing mineral oil content to
5% or less through successive oil changes. Specific oil type and viscosity
recommendations from the system or compressor manufacturer should be
followed.

An R404A system will already have POE oil instead of mineral oil. When
retrofitting to any of these alternates, a change of oil types is not necessary.
It is still good practice to change oil in conjunction with a retrofit, in order to
remove as many contaminants from the system as possible. The cleaner the
system is, the more efficiently and reliably it will operate. In this situation
of a POE-for-POE change, multiple oil changes (as may be necessary with
a mineral-to-POE change) are likely unnecessary, unless there is severe oil
contamination.

5.3 ELASTOMERIC SEALS

When retrofitting a system to R407A, R407F, R448A, or R449A, it is
recommended to change the elastomeric seals in the system to minimize
the possibility of external leaks. This is especially important with an R22
system retrofit. Elastomeric gaskets and O-rings that have been exposed
long-term to R22 and mineral oil have a greatly increased risk of leaking after
a deep vacuum being applied to the system, followed by exposure to HFC
refrigerants and POE oil.

When retrofitting an R404A system, the seals are not exposed to different
refrigerant and oil types after the retrofit. However, the existing seals will
likely have been in place for a number of years. Pulling a deep vacuum
on the system will still alter the set the seals have taken from being under
internal pressure, and cause small amounts of refrigerant the seals have
absorbed over time to begin to release from them. While the risk of leaks
may be less for an HFC-to-HFC (or HFO) and POE-to-POE retrofit, it is

23
nonetheless best practice to change the elastomeric seals. Minimizing
external refrigerant leaks benefits both the system’s owner and the
environment.

5.4 SUMMARY FOR RETROFITS TO R407A, R407F, R448A, OR

R449A
In a retrofit from R404A, existing liquid side components may be too large,
and TEV thermostatic elements are not compatible. Replacement of at
least TEVs may be necessary. R404A suction side components may be okay.
In a retrofit from R22, both liquid and suction side components are likely
okay. Though most critical for an R22 system, any retrofit should include
measures to ensure a clean and dry system, and minimize external leaks,
including changing of filters, refrigeration oil, and elastomeric seals. Figure 6
summarizes Sections 2.1 – 5.3.

Figure 6 - Guidelines for Retrofits to R407A, R407F, R448A, or R449A

FROM R404A FROM R22

COMPONENT
Existing is . . . Check Sizing? Recommendation Existing is . . . Check Sizing? Recommendation

Replace only if Replace only if

Distributor Likely Okay Required* Likely Okay Required*
sizing requires sizing requires

TEV Likely Replace only if

Required Replace TEV Likely Okay Required
(Sizing) Oversized sizing requires

TEV
Replace with R22
(Thermostatic Not Compatible N/A Compatible N/A Don't replace
element**
Element)

Liquid Line Possibly Replace only if Replace only if

Required Likely Okay Good Practice
Solenoid Valve Oversized sizing requires sizing requires

Externally Replace only if Replace only if

Likely Okay Good Practice Likely Okay Good Practice
Piloted EPR sizing requires sizing requires

Internally Replace only if Replace only if

Likely Okay Required Likely Okay Required
Piloted EPR sizing requires sizing requires

Replace only if Replace only if

Electric EPR Likely Okay Good Practice Likely Okay Good Practice
sizing requires sizing requires

Suction Replace only if Replace only if

Likely Okay Good Practice Likely Okay Good Practice
Solenoid Valve sizing requires sizing requires

Likely OK size, Likely OK

Filter-Drier but Good Practice Change filter-drier size, but Good Practice Change filter-drier
contaminated contaminated
Likely OK size, Likely OK
Change Change
Suction Filter but Good Practice size, but Good Practice
suction filter suction filter
contaminated contaminated

Refrigeration Change to Change until <5%

Compatible N/A Not Compatible N/A
Oil de-contaminate MO content

Elastomeric Best practice to

Leaks Possible N/A Leaks Likely N/A Change seals
Seals change

* In order to accurately check TEV sizing.

** Only if TEV sizing is OK, best practice is to replace the TEV.

24
© 2017 Parker Hannifin Corporation Form 5-492 / 22017

Parker Hannifin Corporation

Sporlan Division
206 Lange Drive • Washington, MO 63090 USA
phone 636 239 1111
fax 636 239 9130
[Link]