Saturday, August 25, 2018

Is the New Refrigerant R-1234yf Better or Worse?


Since 2012, a new refrigerant has been taking over the automotive air conditioning industry. The whole reason for the switch was a decision of the European Parliament back on May 17, 2006. They decided to prohibit refrigerants that had a 100 year global warming potential greater than 150. I'll forego discussing global warming in this post.

DuPont and Honeywell developed R-1234yf to meet the global warming potential (GWP) requirements. But this new refrigerant was worse than R-134a in several very important categories. It's more flammable, more toxic, and more expensive. The main reason why R-1234yf has such a low GWP number is because of the short atmospheric lifetime of 11 days. It quickly decomposes into trifluoroacetic acid (TFA).

If the goal was to find a refrigerant safer for the environment, they failed. It's no safer than R-134a. Unfortunately, the target metric was GWP, and that's about all they got.

Check out what happened when engineers at Mercedes-Benz tested it.
This video shows both R-134a and R-1234yf leaking onto a hot turbo. The R-1234yf catches fire, while the R-134a does not.

R-1234yf alone is hardly flammable at all. But it is when it's mixed with refrigerant oil, as is the case when used in air conditioners.

The fire isn't what worries me. It's the deadly fluorine compounds released as combustion by-products. These gasses are so reactive, they etch glass. The IDLH for HF is 30 ppm. That means that 30 ppm is immediately dangerous to life or health. This is an accident waiting to happen. Someone is going to get a very serious injury or die from an accident where this refrigerant is burned.

To compensate for the fact that R-1234yf is flammable, auto makers re-designed the evaporators to be more sturdy to prevent leaks into the passenger compartment.

Since it seems we're now ok with refrigerants that are toxic to the environment and release deadly gasses when burned, I propose we switch to a different refrigerant that is much safer for the environment, has virtually no toxic combustion by-products, and runs at about the same pressures as existing automotive A/C systems. It's only worse than R-134a in one category. It's highly flammable. We've seen that it is economically feasible to build better evaporators for R-1234yf, so why not do the same for R-600a?

R-600a has already been legal in the United States for use in stand-alone refrigeration units since July 31, 2013 (link).


Here are some other refrigerants used in or since the early 20th century.


Sulfur Dioxide is toxic. The IDLH is 100 ppm. It's corrosive, which means it will eventually leak out of the refrigeration unit. Who wants their refrigerator to smell like rotten eggs?

Ammonia is also toxic. The IDLH is 300 ppm. But it's cheap and an efficient refrigerant, which is why it's still used in some large-scale commercial systems today.

Chloromethane (aka methyl chloride) IDLH is 2000 ppm. It's carcinogenic. With a flash point of -20C, chloromethane is extremely flammable. It is no longer used as a refrigerant.

Methylene Chloride (aka dichloromethane) IDLH is 2300 ppm. No flash point, but can be flammable above 100C.

Methyl Formate (C2H4O2) IDLH is 4500 ppm, meaning it's toxic. It has zero GWP and zero ozone depletion. But it's highly flammable, with a flash point of -19C.

Propane is a simple asphyxiant (too much propane means not enough oxygen) so other than that, no symptoms from inhalation.

Carbon Dioxide has a GWP of 1 (which might as well be zero). It's safe, non-toxic, and non-flammable. CO2 refrigeration was patented in 1867 by Thaddeus Lowe (British patent 952), and used to cool an office building in 1906. The main disadvantage is the high pressures required.

One of the major reasons we used R-12 in automobiles was because it wasn't flammable. (Never mind the 50 liters of highly flammable gasoline in the fuel tank.) And R-12 runs at much more reasonable pressures than carbon dioxide.  The reason we stopped using R-12 was because of the ozone depleting potential. The chlorine in the ozone layer would act as a catalyst to reduce O3 to O2.

We converted over to R-134a was because it was better for the environment and I do agree that it is a better alternative than R-12, but the move to R-1234yf was supposed to be because it was even better for the environment. The only reason it has a lower GWP number is because it decomposes faster. It decomposes quickly (11 days) in air into trifluoroacetic acid (TFA), which is harmful if inhaled, and toxic for aquatic organisms, even at low concentrations. TFA is not biodegradable in water. It's worth noting that R-134a also decomposes into TFA, but at a slower rate than R-1234yf.

In this table, I compare several refrigerants to R-134a. A green background means it's better than R-134a in that category, and red is worse.


* The maximum possible AEL number is 1000, so that isn't a fair comparison.

Comparing the refrigerants, it appears that R-134a is the best non-flammable option. Otherwise, R-600a is the best choice. I didn't include carbon dioxide in this table since it isn't compatible with typical automotive compressors. But the technology is available to use it.

Conclusion

Let's stick with R-134a until we switch over to carbon dioxide. It's not too hard to do. If they could build a CO2 system for an office building in 1906, we can build automotive CO2 A/C units today.

Saturday, August 18, 2018

Compressor Clutch Failure Analysis


I got this GPD compressor from RockAuto.com. The clutch lasted only a few minutes before it failed open circuit. The wires were jiggling around while it was running, so I thought something was loose. So I'll do a little failure analysis to try and find the reason behind the failure.

First, I removed the center bolt holding the clutch face on. Then I removed the clutch face by just pulling it off.

Next, I removed the snap ring and pulled the pulley off. 

Here's what the inside of the pulley looked like. There was a bunch of goo from the melted electromagnet. 

Here's the melted electromagnet, still on the compressor. Do you notice anything missing?

See that groove at the base? That's where a snap ring is supposed to go. The top groove holds the pulley on, while the bottom groove holds the electromagnet on, well, if it had a snap ring installed. It seems somebody at GPD forgot to install it.

How does a loose electromagnet cause it to overheat? I think there was enough room for it to rub against the pulley and create heat from friction. There was a little noise coming from the compressor while it was running.

How could they have detected this at the factory? A test run using a thermal imaging camera, taking note of the temperature during operation would have been helpful. Noise measurements may have detected it. Following assembly procedure would have helped. Making sure the parts weren't loose before sending it out the door would also have helped. I can only assume their factory didn't have any of those controls in place.

What effect does one employee's carelessness have? You can lose a customer for every mistake you make. This wasn't the same exact type of compressor that was on the car, but it fit. So I can't use the old electromagnet on this one. I didn't buy an expansion valve, so I guess if I return this I won't get any credit. Oh, well, it made for a good blog post.

Not only did GPD fail quality, they failed customer service too. Here's my attempt to buy a clutch from them:

John Andrews providentengineering@gmail.com

Aug 31, 2018, 12:53 PM (23 hours ago)
I need a clutch for a compressor: GPD part number 6511491.

Can you sell me the clutch?
The one I got from rockauto didn't have the snap ring installed, so the electromagnet rubbed on the pulley, causing the clutch to overheat and fail within minutes.

Thanks,
John Andrews

Albert Brown abrown@globalpartsdist.com

Aug 31, 2018, 1:52 PM (22 hours ago)
Good Day Mr. John,

GPD does not offer the clutch for the 6511491. The clutch is only sold with a new compressor assembly.


Kind Regards,


cid:image001.jpg@01CF1B6C.8D8DC490                                 
Albert Brown
Inside Sales
Global Parts Distributors, llc
3279 Avondale Mill Road
Macon, Georgia 31216
Ph: 1 800 722 5089 ext 6329
Direct: 478 785 6329
Fax:  478 781 6006


From: John Andrews <providentengineering@gmail.com>
Sent: Friday, August 31, 2018 2:53 PM
To: CATALOG <CATALOG@globalpartsdist.com>
Subject: Need AC clutch


Disclaimer
The information contained in this communication from the sender is confidential. It is intended solely for use by the recipient and others authorized to receive it. If you are not the recipient, you are hereby notified that any disclosure, copying, distribution or taking action in relation of the contents of this information is strictly prohibited and may be unlawful.

This email has been scanned for viruses and malware, and may have been automatically archived by Mimecast Ltd, an innovator in Software as a Service (SaaS) for business. Providing a safer and more useful place for your human generated data. Specializing in; Security, archiving and compliance. To find out more Click Here.

John Andrews providentengineering@gmail.com

AttachmentsAug 31, 2018, 1:54 PM (22 hours ago)
Thank you for your response. 
Can you let me know the part number and manufacturer of the clutch?


[No response from GPD]

What Else Can Go Wrong? The Sonata's A/C

Vehicle:    2005 Hyundai Sonata V6
Problem:  A/C not working

It was going to be a simple diagnostic. Maybe replace an o-ring and do an evac and recharge. So I tested for leaks and found a small leak on the liquid line. Here's where the trouble began. 

I looked on several online auto parts stores for this hose. Nobody had it. I called the dealer (I didn't have the VIN handy) and asked them if they had the part. They had to order it in for about $470. I asked them what the part number was and they gave it to me. 

Cool. Now I can search online. As luck would have it, I found one on ebay for $160. Fast forward to opening the box and discovering it was completely wrong. As it turns out, there was a model change mid-year in 2005. Thanks Wikipedia!

Ok, back to the search engines. Since late 1998 to early 2005 were the same body style, I figured that the 2002 liquid line (part number 977683D001) would be correct. After all, it looked right . . . at first. 

I missed the slight difference on one connector. When I pulled the old one off, I saw that they were different, as you can see here. Here's what was on the car:
2005 Hyundai Sonata Liquid Refrigerant Line Connector
And here's the one I got from ebay:
2005 Hyundai Sonata Liquid Refrigerant Line Connector, part 977683D001
Then I thought I had a brilliant idea. I'd cut off the other end of the old one and put a couple o-rings on it to adapt the ends. So here's the end I cut off that connected to the expansion valve:


This'll go between the two parts with a couple o-rings and seal it right up . . . maybe.

It fits in this end.

With a little machining, it'll fit in the other end too.

Just trimming some thickness off the sides.

Now it fits.

I found one o-ring that looked good, and one that was slightly small.

It didn't seal well. It was also not a good match. Also, the stud wasn't quite long enough for the thicker connector. So I came up with another idea.

I ground off the crimp connectors and pulled the entire metal ends out of the hoses and put the old connector on the new hose. The hoses used some sort of sealant between the hose and pipe, so I tried epoxy. It should hold up over time.

Three hose clamps, epoxy, and a new o-ring should keep the refrigerant inside.

We're not done with things going wrong yet. Houston, we have a problem. We're losing pressure. Some mechanics have those fancy refrigerant gas detectors. I use a low-tech, yet effective solution: soapy water. My soapy water wasn't showing any leaks on any of the hoses or hose connections. Yet I was hearing a hissing noise.

Some mechanics have fancy stethoscopes to find noises. I used a 3 foot length of 3/4" heater hose to find the hissing noise on the compressor. Oh, that should be easy. Just replace the o-ring on the end of the hose connector.

Just a little verification with the soapy water, and . . . not so fast. No bubbles.

I finally found the source of the leak.


At first I thought it was a crack in the compressor, but it was actually where the parts of the compressor body join together. I missed that leak the first time I checked for leaks.

New Compressor Installation

After a week, we got a compressor in the mail and I installed it. Installation was pretty straightforward. I had to remove the small shroud under the compressor and remove the oil filter and then it came out fairly easily. After installing the new one, I checked for leaks and it passed. I evacuated the system and left the vacuum pump on for an hour. Then I was using my new refrigerant scale to charge it and I had almost a full charge in, 0.62 kg, when the clutch shut off.

New Compressor Un-Installation

Wait, what? After some troubleshooting, I determined the clutch electromagnet had an open circuit. I had to recover the refrigerant and remove the new compressor and put the old one back in. I think GPD didn't install the new clutch correctly because it would wiggle back and forth. Plus, there was that burning smell.

Filter Drier

I'm not sure why it took me so long to find it. But the filter drier is built into the condenser on the driver's side. You remove the plug on the bottom with a 10mm Allen wrench. The filter is right above the plug. Here's what they look like. I used mass airflow cleaner to clean out the filter screen. Then I applied PAG oil to the threads and o-rings before re-installing it.


Here's what the old and new desiccant (moisture absorbent) bags look like. After seeing this, it's no wonder why they recommend you always change the filter drier (or desiccant bag or accumulator) when you replace the compressor. Not changing it would be like replacing an engine and re-using the old oil filter.


Finished?

After evacuating the system and recharging it, the thermometer says I did an adequate job. By the way, the new compressor was a Denso brand. It came with the proper manifolds so I didn't have to swap them from the old unit.


So I disconnected my hoses from the charging ports and found one more small problem I need to take care of . . . tomorrow, because there's a wind and rain storm coming in. The high side charging port valve is leaking slightly. Somehow I didn't find that when I did leak checks the first two times.

And my valve core tool wouldn't hold onto the valve core to pull it up, so I couldn't finish it before the rainstorm hit. I didn't know that it was the large size core. My tool only works for the smallest size valve cores. You can see in this picture that the valve core on top just doesn't fit through that small hole.


Next Saturday, I'll have to change the valve core the old way - recover, replace, evacuate, and recharge.