Oil harvesting (defrosts) >Marc O'Brien<
10073...@compuserve.com
On 11 May, 1996, T.Morgenthaler wrote...
> In <4n1qhd$uv$1...@mhadg.production.compuserve.com>
> 10073...@compuserve.com(Marc O'Brien)
<10073...@CompuServe.COM>
> writes:
> >
> >My posts are taking day to reach the newsgroup if at all !
> >
> >I'm repeating myself here, but have you set up your oil harvest
> >cycle correctly? More of but shorter etc.
> >Oil harvests are more commonly known as defrosts.
> >
> >--
> >Marc O'Brien
> >Ind.Ref.Mech.
> >http://ourworld.compuserve.com/homepages/marc_vicky
> >
> The unit doing this in particular is a McQuay with two model 3000
> compressors. There is no defosting cycle on these units, capacity of 60
> Tons. I recently changed out both compressors. They are about 5 years
> old, however this a 24 hour, 365 day/yr building so it could be argued
> that these have run for 15 years. Compressor #1 will, during pumpdown
> drop it's oil level to the first ring on the sight glass. At start up,
> it will come back within perhaps a minute to 3/4 sight glass, strong
> flow indicated with the splashing. Compressor #2 originally maintained
> a high level during all cycling. Now, it has dropped to slightly less
> than 1/2 the sight glass and has remained that way for the past few
> days. I am wondering if I should add some oil. #2 also has a tinny
> noise topside during loaded conditions. When the unloader is engaged,
> the noise disappears. Again, these are Copeland 30 T model 3000
> Compressors.
> Thanks for the comments.
> Ted
Hi again T.Morgenthaler
Although each brand and then model has it’s typical problems, I don’t like to
discuss system brand but rather over all component configuration. In the same
way that a surgeon wouldn’t delay operating on people named Marc or Trever.
I also can’t work out the exact situation here. I can see now that it isn’t a low or
med temp set up but rather an AC application. Originaly I understood the whole
system was new, now it looks like maybe only the compressors are new. Why did
the originals need to be replaced and what wear patterns were there?
Normally noise and vibration are associated with unloading and not loading due to
balance swings. Oil levels do fluctuate but can drop very low sometimes on
systems that don’t cycle often enough especially with sustained border line low
loads. Starting up with the TEV fully loaded after a shut down does help harvest
the oil.
The above is especially true for systems with system velocity reduction and
hence mass flow reduction type unloaders. If the system is way oversized due to
recent building load changes or what ever, then full load oil return cycles become
rare because the comps can easily settle in an unloaded state for hours on end.
Here you may want to consider double risers.
Are you shy on too low a superheat here maybe?
After setting full load superheat, check that fully unloaded superheat doesn’t
fluctuate bringing liquid to the comp sump and then causing excessive oil
circulation caused by foaming. Full load superheat ideally would be 6K
Also sometimes the oil level indicator is influenced by shaft spin direction.
Oil levels do fluctuate and drop most when the TEV and mass flow drop i.e.
unloading or low load periods.
To put a point across, you might in the short term think of timing the compressors
off every hour for ten or so minutes. Just during the low load periods i.e.
mornings. This would alow for periodic fully opened TEV operation harvesting the
oil. Experimentation is always best.
Never, never add oil for the sake of maintaining an ideal oil level, oil level is
circumstantial.
A normal recip compressor can pump all its oil out in about an hour. Then with a
non spiral oil sepperator of 60 % efficiency this is near 3 hrs or so. So you
can see that if there is no return at all the comp will fail soon.
Also check that you aren’t overcharged and hence reducing mass flow due to
liquid subcool beyond design.
Marc O'Brien
Ind.Ref.Mech.
http://ourworld.compuserve.com/homepages/marc_vicky
George Goble
In article <4n54c5$o...@dub-news-svc-2.compuserve.com> <10073...@compuserve.com> writes:
>A normal recip compressor can pump all its oil out in about an hour. Then with a
>non spiral oil sepperator of 60 % efficiency this is near 3 hrs or so. So you
>can see that if there is no return at all the comp will fail soon.
WOW.. which kind/brand of compressor can pump all it's oil out in an hour?
I made an oil miscibility test stand from a Copeland Med temp 2Ton (R12)
semi-hermetic compressor, with an oil sight glass. It holds about 64
Fl Oz oil. Evap is just a 50' coil of 5/8 tubing, with the loops
vertical (to trap oil).
Even when running R-134a in there with mineral oil (no oil return at normal
mass flows), it appeared it would take 2 or 3 days to pump out 1/2
to 3/4 of the oil. I have seen automotive systems with approx 10%
oil in a liquid line sample from a running system.
--ghg
http://worldserver.com/R-406A
http://ghg.ecn.purdue.edu
George Goble
In article <4n54c5$o...@dub-news-svc-2.compuserve.com> <10073...@compuserve.com> writes:
>A normal recip compressor can pump all its oil out in about an hour. Then with a
>non spiral oil sepperator of 60 % efficiency this is near 3 hrs or so. So you
Have you ever heard of the Temprite 9000 series of oil separators?
The ULT (ultra low temp folks) told me they are the best.. nearly
100% oil removal..I build 3 or 4 pieces of recovery equipment, using 9000 seps
(back when it was legal to do so).. the largest being a 3-1/2 ton
scroll.. Blowing some liquid out of the condenser onto a sheet of fine
typing paper, showed NO oil stain at all.
--ghg
Marc O'Brien
g...@cidmac.ecn.purdue.edu (George Goble) writes:
> In article <4n54c5$o...@dub-news-svc-2.compuserve.com> <10073...@compuserve.com> writes:
> >A normal recip compressor can pump all its oil out in about an hour. Then with a
> >non spiral oil sepperator of 60 % efficiency this is near 3 hrs or so. So you
>
> WOW.. which kind/brand of compressor can pump all it's oil out in an hour?
> I made an oil miscibility test stand from a Copeland Med temp 2Ton (R12)
> semi-hermetic compressor, with an oil sight glass. It holds about 64
> Fl Oz oil. Evap is just a 50' coil of 5/8 tubing, with the loops
> vertical (to trap oil).
>
> Even when running R-134a in there with mineral oil (no oil return at normal
> mass flows), it appeared it would take 2 or 3 days to pump out 1/2
> to 3/4 of the oil. I have seen automotive systems with approx 10%
> oil in a liquid line sample from a running system.
Ignoring miscibility, pour point or flock point can you give me
the velocity through the evaporator and suction line. Are you sure
the loops weren't acting as oil return traps
Also I think the compressor you're talking about has no unloaders.
Reading some of Copelands old notes:
..consider a compressor having an oil charge of 150 ounces
with the normal circulation rate being 2 ounce per minute
If a minimum of 30 ounces is required the compressor would run out of
oil in 60 minutes with zero oil return.
george you and I know that no system has zero oil return even your
experiment. Any math regarding system design is done per job.
But I'm only trying to emphasize a point.
You and I know what defrosts are for but approaches are changed
when expressing a point. People who learn refrigeration from me
appreciate it.
One of the things I love about the fridge game is that there are
no absolutes. The more you learn the more it seems so.
I hope I never ever work it all out. I just enjoy solving those
so mystery faults.
Marc