To begin, HFCs (134a and 404a / 507) MUST have the mineral oil flushed out and replaced with polyolester (POE). Most manufacturers recommend less than 5% residual mineral oil, and in some cases down to 1%. With the retrofit blends (401A, 401B, 402A, 402B, 408A, 409A, and similar blends) the answer isn't as clear. With one exception, compressor manufacturers recommend that some of the mineral oil be changed to either alkylbenzene (AB) or POE (usually 50% or more). On the other hand, some refrigerant manufacturers have made conflicting claims that "no oil change is needed." These are based on oil-refrigerant miscibility tests or measuring oil return in actual systems.

The main concern about lubricant choice is protecting the compressor, which means providing for oil return from the system. The two factors that will affect oil return are chemical mixing of refrigerant and oil and physical system design which promotes "mechanical" oil return. Smaller, warmer-evaporator systems will generally show better oil return than larger, colder-evaporator systems.

Because we generally can't change the system design on a retrofit, the best way to improve oil return is to change the less miscible mineral oil to the more miscible AB or POE. This is true even with R-502, which often has oil return problems. Generally, the more mineral oil changed, the better the oil return.

Given the properties of the refrigerant/oil-type mixture, and after reviewing the system design, the contractor or technician should be able to decide how much, if any, of the oil should be changed.
Back to top

It depends. Studies were done a few years ago to show how higher glide blends behave during leakage and they showed significant fractionation, which affected the properties of the blend. When the system was topped off, the properties came back close to original. The cycle was repeated to see how many times the system could leak before topping off became a problem (the recommendation was not more than five). These studies were done on containers at rest, which promotes the worst case of fractionation.

Another study was performed recently on a system running full time, then cycling normally (2/3 on, 1/3 off), which found that the blend did not fractionate when the refrigerant is moving around inside, and not much fractionation occurred when cycling. Low-glide blends didn't show much fractionation in any case.

What this means is that running systems found to be low on charge have probably not fractionated the blend much, and can be repaired and recharged directly. If the system has been off for a long period (more than a day) and found to have leaked (worst case is about half the charge), it's probably better to pull what's left and charge with fresh, unless very little is gone, or very little is left. Low-glide blends won't cause any fractionation-related problems.
Back to top

There are several reasons for bubbles in the sight glass. If one of the traditional refrigerants showed vapor in the sight glass it often meant there wasn't enough liquid refrigerant being fed to the valve, and more refrigerant was added to the system.

Blends could show flashing for the same reason, however, they can also flash when there is plenty of liquid in the receiver. Ironically, this liquid in the receiver could be causing the problem, particularly when the equipment is in a hot environment. Blends will come out of the condenser slightly subcooled - at a temperature below the saturated temperature of the blend at the existing high side pressure.

Yet when the blend sits in the receiver, it can "locally fractionate," or change composition slightly by shifting one of the components into the vapor space of the receiver. This will effectively produce a saturated liquid in the receiver, at the same pressure you had before, which flashes when it hits the expanded volume of the sight glass. In most cases these bubbles will collapse when the blend gets back into the tubing which feeds the valve, and the system will operate just fine.

Check other system parameters such as pressures, superheat and amperage to confirm whether you have the right charge. Don't rely solely on the sight glass.
Back to top

A single component refrigerant always had a "boiling point." Zeotropic blends change composition when they boil or condense, and therefore have a continuously changing boiling point. The most useful temperatures to know are where the boiling starts and ends. Bubble point and dewpoint are terms used in the chemical industry to define these two temperatures.

Bubble point is the temperature where the saturated liquid starts to boil off its first "bubble" of vapor. (Picture a pot of liquid with the first bubbles starting to appear.) It is also called the "liquid side temperature/ pressure relationship." Dewpoint is the temperature where saturated vapor first starts to condense, or the last drop of liquid evaporates. (Picture a room full of vapor with a few drops forming on the ceiling.) This is also called the "vapor side temperature/ pressure relationship."
Back to top

The two columns on the PT chart give the liquid and vapor pressures at the listed temperatures. Single component refrigerants and azeotropic blends have bubble points and dewpoints equal to each other, and we simply call this the boiling point. When there is only one column on the PT chart, lowglide blends would have very similar numbers in the two columns, and often the PT chart will only have one column as well for them.

How a two-column PT chart is used is straightforward. Most times you're interested in knowing the saturated temperature of the refrigerant at the system pressure, so you can compare it to a measurement you're making on the system (for example, to check a superheat or subcool setting). Simply keep track of the condition of the refrigerant where you're measuring, and cross-reference the same side of the PT chart.

Superheat measurements check the line temperature of superheated refrigerant vapor coming out of the evaporator versus the saturated vapor temperature, so you would use the vapor side of the PT chart.