How to Charge a Winter Install When Superheat Cannot Be Trusted

Why this matters

A system set in January has to leave the job charged, and the two gauge methods most techs trust are both outside their charts on a 28 F day. The tech who charges by gauges anyway leaves a system that is wrong in a direction set by the weather, not by the refrigerant, and nobody finds out until the first hot week. The tech who weighs it in correctly has done the right thing but has no proof yet. This card is the procedure that gets both halves: a charge set by mass today, and a booked visit that proves it in cooling weather.

Why the gauges lie in the cold, and which way each method errs, is the subject of Refrigerant Charging Methods and Their Outdoor Temperature Limits. The weigh-and-verify method in normal weather is in Charge Verification by Weight and Superheat. This card is the winter procedure only.

Before the first step: who does this, and what it can hurt

Recovering, charging and opening a refrigerant circuit are work for an EPA Section 608 certified technician under 40 CFR Part 82, Subpart F, and the venting prohibition there applies at every hose connection in this procedure (only de minimis releases during good-faith recovery practice are excepted). Liquid refrigerant flashing off a hose fitting causes frostbite in seconds, and in the cold your hands are already numb enough not to feel it start, so gloves rated for refrigerant and eye protection go on before the first hose. On a system using an A2L (mildly flammable) refrigerant, use the tools, ventilation and ignition-source controls the manufacturer lists, and keep a torch or a portable heater out of the space where a hose is being connected. Nitrogen for the pressure test goes through a regulator, never straight from the cylinder.

The procedure

1. Measure the line set yourself, and write the number down. Tape the run from the outdoor service valves to the indoor coil along the actual path, count vertical rise separately, and note the liquid line's outside diameter. Why: the adder is computed from length, and the length on a quote is a guess made before anyone saw the route. What breaks if you skip it: the weigh-in, the one method that works in winter, is exactly as wrong as the tape you did not pull.

2. Read the installation instructions for three things before touching a valve. The line length the factory charge already covers, the added charge per foot for your liquid line size, and the maximum length and lift allowed without extra accessories. Some also say what to do on a line shorter than the included length; follow that rather than assuming no change. What breaks if you skip it: a remembered rate from another model, which lands a few ounces off on every install.

3. Pressure-test with dry nitrogen, and correct the reading for temperature. Pressurize to the test pressure the instructions give, record the gauge reading and the line temperature at the start, and let it stand for the time the instructions or your shop standard require. Why the correction matters in winter: nitrogen pressure in a sealed line changes with its absolute temperature, and a line set tested at 4 pm and read at 6 am may have cooled 10 to 15 F overnight. That drop alone pulls the gauge down several psi with no leak at all. Convert to absolute pressure (add 14.7 to psig) and absolute temperature (add 459.67 to F), scale the start pressure by the ratio of end to start temperature, and compare the result to what you read. What breaks if you skip it: you either chase a leak that does not exist, or you learn to ignore a drop of a few psi and miss one that does.

4. Evacuate to the standard, and give the cold its extra time. Pull to the micron target the instructions set (500 microns is a common target) and pass the decay test your instructions or shop standard specify. Two cold-weather traps. Vacuum pump oil thickens in the cold, so a pump that rode in an unheated truck box pulls slowly until it warms; run it in the cab or the house before connecting, and change oil that looks milky. And moisture in a cold line can freeze under vacuum: the gauge stalls at a respectable reading while ice sits in the line, then climbs during the decay test as the ice slowly turns to vapor. A decay that keeps rising and then levels off well above the target is the signature of water, not of a leak. Why it matters: moisture left in the circuit reacts with oil and refrigerant into acid. What breaks if you skip the extra time: a system that passed a rushed vacuum and fails a compressor years early.

5. Keep the refrigerant cylinder warm, and never heat it. A cylinder that rode at 28 F has low pressure and pushes liquid slowly, so let it sit in a heated space (the house, or the shop overnight) before the charge, never in the truck cab, where a leaking valve fills a closed cabin. Never put a torch, heat gun or heater on a cylinder; that is how cylinders exceed their rating. What breaks if you skip it: a weigh-in that stalls halfway and gets finished by guesswork.

6. Weigh in the adder, as liquid, on a zeroed scale. Most current refrigerants are blends and are charged as liquid so the mix stays correct. Where the instructions allow it, the adder can go into the evacuated line set and indoor coil through the liquid service port before the outdoor valves are opened, because liquid flows readily into a vacuum. If pressures equalize before the full amount is in, the rest goes in with the system running, metered into the suction side in small amounts so liquid does not reach the compressor, and only in a mode the unit is allowed to run in at that outdoor temperature. Then open both service valves fully and cap them. What breaks if you skip the scale: you are back to charging by feel at a temperature where no feel is right.

7. Record the charge as weighed, not as "charged to spec". The ticket carries measured length, liquid line size, included length, adder rate, adder, scale reading, outdoor temperature, method and the words "pending warm-weather verification". What breaks if you skip it: in July the next tech reads "charged" and starts adjusting.

8. Start the system and check operation, not charge. On a heat pump, run heating and confirm the outdoor coil frosts evenly, the indoor airflow is in range and the amps are normal. On a cooling-only system, do not run cooling below the minimum outdoor temperature the instructions list; the indoor coil can ice and return liquid to the compressor. Do not adjust the charge from anything you read in this step. When you open an electrical compartment to check amps, turn off the disconnect first, prove the circuit dead with a tester checked on a known live source before and after, then make the clamp connection and restore power with the panel closed and your hands clear.

9. Book the return visit before you leave, with its trigger written down. Put a check-back on the schedule for the first day forecast inside the charging chart's outdoor range, with the house at a normal cooling setpoint. Tell the customer what it is for, in one sentence, and that it is part of the install. Why: a visit with no date drifts into summer, and a visit with no trigger gets run on another cold day and repeats the problem. What breaks if you skip it: the weigh-in is never proved, and an error in step 1 or 2 rides undetected until it costs a compressor.

Worked example: a January heat pump

A crew sets a heat pump with an expansion valve on January 22. Outdoor air is 28 F at commissioning. The quote listed a 50 foot line set.

Step 1. The run had to go around the garage rather than through it. Measured with a tape along the actual path, it is 62 feet with 9 feet of vertical rise, 3/8 inch liquid line.

Step 2. Say the instructions read: factory charge covers 20 feet, add 0.5 ounce per foot of 3/8 inch liquid line beyond that, maximum length 80 feet, maximum lift 25 feet. The install is inside both limits.

Step 3, and it is the step that nearly sent the crew looking for a leak. The line was pressurized the previous afternoon. Say the instructions call for a standing test at 400 psig (an illustrative value; use the instructions' figure). Start: 400 psig with the line at 36 F. Next morning at 6 am the line reads 22 F and the gauge reads 389 psig. That is an 11 psi drop, and the apprentice called it a leak.

The correction, worked line by line:

TEMPERATURE CORRECTION, NITROGEN STANDING TEST
start   400 psig + 14.7      = 414.7 psia
start   36 F + 459.67        = 495.67 R
end     22 F + 459.67        = 481.67 R
ratio   481.67 / 495.67      = 0.9718
expect  414.7 x 0.9718       = 403.0 psia
expect  403.0 - 14.7         = 388.3 psig
read                           389 psig
gap     389 - 388.3          = 0.7 psi high
verdict inside gauge reading -> no leak shown

Check it the other way: 14 F of cooling over roughly 496 R absolute is about 2.8 percent, and 2.8 percent of 414.7 psia is about 11.7 psi, which matches the 11 psi the gauge actually dropped. Had the gauge read 372 psig, the gap would have been 388.3 minus 372, which is 16.3 psi below the expected value, and that is a leak to find before anything else happens.

Step 4. The pump rode in the cab and pulled to 480 microns. During the decay test the reading rose to 900 microns over the first 10 minutes, crept more slowly, and leveled off near 1,100 microns. The tech broke vacuum with nitrogen, swept the lines, and pulled again; the second decay held under the shop standard. Water, not a leak, which is what the creep-then-level pattern suggested.

Step 6. The adder: 62 feet measured minus 20 feet included leaves 42 feet; 42 feet times 0.5 ounce per foot is 21 ounces. Check from the other side: 21 ounces divided by 0.5 is 42 feet, plus 20 included, is 62. The 21 ounces went in as liquid into the evacuated lines, and the valves were opened.

What the quote's number would have done. On the quoted 50 feet, the adder is 30 feet times 0.5, which is 15 ounces, 6 ounces short. Say the nameplate factory charge is 180 ounces (illustrative). The correct total is 180 plus 21, which is 201 ounces; the short charge is 195 ounces, and 6 of 201 ounces is about 3.0 percent low. Small enough that no one notices in January. Large enough to show as low subcooling and a longer run time in July, on a system whose ticket says it was charged by weight.

Steps 7 and 9. The ticket reads: 62 ft, 3/8 in, 20 ft included, 0.5 oz/ft, adder 21 oz, 28 F, weigh-in, pending warm-weather verification. The check-back is booked for the first forecast day inside that unit's subcooling chart.

The return visit. On April 18 the forecast shows 71 F, which is inside the chart. Say the target subcooling is 10 F with a tolerance of plus or minus 3 F (illustrative; the instructions give the real one). The reading after the system stabilizes is 9 F: inside the band, weigh-in confirmed, ticket closed. Had it read 4 F, the tech would treat that as a finding to diagnose (airflow, a restriction, a leak since January), not as a cue to add refrigerant until the number looks right, because a measured, weighed charge that reads wrong is telling you something else is wrong.

Where the answer changes

A cooling-only system with a furnace. Steps 1 through 7 are unchanged. Step 8 shrinks to confirming the outdoor unit is wired and the disconnect is set, because there is no mode the instructions allow it to run in on a 28 F day without the manufacturer's low-ambient provisions. The customer has heat from the furnace; the check-back carries all of the verification.

The instructions publish a low-ambient charging procedure. Then that procedure is the method, step by step as written, and the warm-weather visit becomes a confirmation rather than the only proof. An improvised version (cardboard over the condenser) is not that procedure.

An existing system whose line set was lengthened. Without the original install record, recover the full charge into a weighed cylinder, evacuate, and weigh in the nameplate charge plus the adder for the new measured length.

References

• 40 CFR Part 82, Subpart F (EPA Section 608), technician certification, recovery and the venting prohibition
• Manufacturer installation instructions for line-set adders, length and lift limits, test pressure, evacuation target and charging charts
• See related: Charge Verification by Weight and Superheat
• See related: Refrigerant Charging Methods and Their Outdoor Temperature Limits
• See related: The Condenser Pad That Froze the First Night