Vacuum Decay After a Condenser Replacement
After a low-speed front-end impact, a 2019 hatchback with an R1234yf system (label charge 620 g) has had its condenser, receiver-drier and two O-rings at the condenser connections replaced. The system is reassembled and evacuated at 22 °C and 65 % relative humidity. After 25 minutes the micron gauge reads 480 microns. The technician closes the valve, switches off the pump and records: 6 minutes - 1,500 microns; 15 minutes - 1,650 microns; 30 minutes - 1,700 microns. He says the system 'holds vacuum' and wants to charge it before lunch so the car is ready for the 15:00 collection. The workshop has dry nitrogen with a two-stage regulator, a digital micron gauge at the manifold, an electronic leak detector and UV dye. The customer has been quoted for the condenser repair only and has already been told the car will be ready.
1. Judge the vacuum decay result. Is the system leaking, and what does the shape of the curve tell you?
Read the worked answer
It does not hold vacuum to the required standard, but the curve does not look like a leak.
Read the shape rather than the single number. From 480 to 1,500 microns in six minutes is a fast rise; from 1,500 to 1,650 over the next nine; from 1,650 to 1,700 over the next fifteen. The curve is flattening towards a plateau. A true leak into a system under vacuum draws air in continuously and the pressure keeps climbing towards atmospheric - hundreds of thousands of microns - with no plateau. A rise that levels off in the low thousands is dominated by moisture evaporating and outgassing from desiccant, oil films and surfaces that are still wet, and by residual refrigerant or solvent slowly releasing from the oil.
Also, a 480 micron finishing point is not proof of a deep vacuum: it can be reached at the gauge while the far end of the system is still wet, which is exactly why the hold test exists.
Conclusion: probe for a leak before charging, but do not condemn joints on the vacuum test alone. Charging now risks trapping water, which will show up weeks later as a frozen expansion valve, acid formation and compressor damage - and the workshop will own that repair, not the customer.
2. Set out how you would separate a genuine leak from residual moisture, and what you would do if the leak turned out to be at one of the new condenser joints.
Read the worked answer
Separating the two
- Dry nitrogen pressure test. Break the vacuum with dry nitrogen to a pressure within component limits - typically around 10 bar for the high side - and watch a calibrated gauge for 20–30 minutes. A leak shows as a continuous, steady pressure fall; moisture causes no pressure change at all because it is not a gas source once the system is pressurised above the vapour pressure of water.
- Localise it. Electronic leak detector at joints, Schrader ports and the evaporator drain, then soap solution as confirmation, then UV dye only as a last resort because it adds a contaminant.
- Only if nitrogen holds go back to evacuation with nitrogen breaks and re-run the micron hold test, extending the time if the ambient is humid.
If a condenser joint is leaking Recover, then remove the joint, discard the O-ring (never reuse it), inspect the seat and the flare or bead for damage or swarf, fit the correct material O-ring for R1234yf and the specified oil, lubricate it, and torque to specification with a calibrated torque wrench - over-torque distorts the seat as often as under-torque leaks. Re-run the nitrogen test, then the full evacuation. Do not use leak sealant: it voids any component warranty, contaminates the recovery machine and can block the expansion valve or the drier.
3. Once the system is sound, describe the final evacuation, charging and performance check, and how you would handle the 15:00 collection.
Read the worked answer
Evacuation Triple evacuation with dry nitrogen breaks, each pull to below 500 microns with the pump isolated and held, the final hold at least 10 minutes with no rise above about 500 microns. Use the micron gauge at the manifold, not at the pump, and change the vacuum pump oil before starting if it is milky.
Charging Charge by weight to the label 620 g using a calibrated scale with the cylinder on the scale, liquid charging through the high side with the system static where the procedure allows, then top up as vapour with the system running. Add the specified quantity of the correct oil for the condenser and drier change. Record the actual weight delivered, not the intended weight.
Performance check Ambient and humidity recorded; engine at 1500 rpm; blower high; recirculation on; log vent temperature and both pressures at two-minute intervals until stable; compare high side against ambient plus a sensible approach, check subcooling and superheat, confirm the condenser fan and compressor control behave, then road test and re-check for leaks with the electronic detector and a visual inspection for oil staining at the new joints. Record everything on the job card.
The 15:00 collection Do not charge a system that has failed a vacuum test to meet a deadline. Speak to the customer before 15:00, explain that the vacuum test showed moisture that must be removed and that releasing the car now would risk a compressor failure, and give a realistic revised time. A diagnostic decision that protects the customer's system is the professional one even when it costs a difficult conversation.