View Single Post
Old 07-30-2012, 05:54 PM
NewBMWownr NewBMWownr is offline
Officially Welcomed to the 'Fest
Location: Chicago
Join Date: Apr 2006
Posts: 111
Mein Auto: E90 335xi (ED 11/07)
My '08 works perfectly. Then again, I only have 33K miles. I think there are several points to be made on this topic. First, aside from mechanical failure (i.e. accident), evaporators should always be suspected for refrigerant leaks. While often well designed, the evaporator is the site of brutal conditions: liquid refrigerant under pressure is being forced through an orifice tube or expansion valve where it sublimates into a gas absorbing heat. The hurricane-like conditions inside the evaporator (liquid being sprayed at high pressure and speed) is matched by the thermal conditions the evaporator is being exposed to (heat and then cold) which causes expansion and contraction of the evaporator. All of this adds up to stress: thermal stress and mechanical stress and stress equals cracks and leaks over time.

Diagnosing an evaporator leak solely using dye is an exercise in futility in my opinion. The evaporator is buried in the dash with makes finding the tell-tale dye stains from a leak next to impossible regardless of whether the mechanic is using a visible light or UV (ultraviolet) dye. The better solution for diagnosis is to use a combination of UV dye and a refrigerant gas detector (aka 'sniffer'). The refrigerant gas detector has a small diameter flexible tube that houses the detector head and, in many cases, the plumbing for a mechanical vacuum pump. The mechanical vacuum pump draws leaking refrigerant in to the detector head, triggering an alert to the operator. See here for example of a refrigerant gas detector. Using a refrigerant gas detector is to find an evaporator leak can be particularly effective because the long flexible tube can be inserted into dash and defroster vents in proximity to the evaporator. The user need not visually see the leak or the leak detector tell tale to confirm a leak.

There are 2 SAE standards for leak detectors: SAE-J-2791 (newer) and SAE-J-1627 (older) The newer standard requires that detectors be able to detect leaks at a level of 4g/joint/year versus 14g/joint/year. Many techs may have a leak detector certified to the older standard and may therefore miss very small leaks.

Finally, the e90 has a relatively small amount of refrigerant: 590g (1.3 lbs) and a tight tolerance band (+/- 50g .02lbs). A small leak and a small loss would likely be noticed. Also, given the tight tolerance band, recharging the system really isn't a DIY activity. I've done a lot of air conditioning work on a lot of cars but I won't touch the e90.

There are leak sealers that are designed to seal small leaks in both hoses as well as metal components of the air conditioning system. These are not the additives that can be found in some forms of R-134 but rather stand alone 4 ounce cans of product that use a small amount of r-134 as a charge. They are injected into the system in the same manner as refrigerant. I must note that the Bavarian Gods would probably take a dim view of using said products in one of their master werks and also some shops will refuse to recharge systems with leak sealers citing risk to the shop's refrigerant recovery machines. Use at your own risk.

It's also worth noting that there's yet another refrigerant on the horizon: R-1234yf which will replace R-134. R-1234yf was developed to meet EU mandates for a lower global warming potential refrigerant than R-134 (while R-134 is chlorine free and therefore not an ozone depleting substance, it still has a global-warming potential that is 1300 times that of carbon dioxide). Expect to see R-1234yf very soon (GM has allegedly begun the transition).

More info for those as sadistic as me:
Reply With Quote