Index Home About Blog
From: (George Goble)
Message-ID: <3b0ddh$>
Subject: Re: A/C replacement for BMW - no Freon
Date: 23 Nov 1994 21:49:37 GMT

In article <3avso7$> writes:

>       Getting your system recharged at places that do not properly
>       evacuate the system is almost unavoidable for most people.

 SO TRUE.... I estimate over 90% do not do good vacuums..

>       The vacuum pump used must be in good working order with regular
>       oil changes of the proper oil to even come close to moisture
>       free evecuation. A micron guage (an expanded scale vacuum guage)
>       must be used to verify that the required 29.90 in hg is pulled.
>       A perfect vacuum is 29.921. A guage set will be not accurate
>       enough. A bad or improperly maintained (probably most) pump
>       will not pull that deep at all.

This is very true... 29.0 inches of vacuum is 25,400 microns,
29.921 is 0 microns..  water boils out at about 1500 microns..
Good practice requires 500-700 microns.. with the vac pump
"blanked off" (valved off), with the micron gauge reading
the system vac.. for 5-10 mins.. If any water is left, vac will
float to 1500ish microns.. if leaking, it will go higher (worse vac).

>       A pump that is oversized may freeze the moisture before it 
>       can be removed by dropping the vacuum too fast.  

This is not completely true.. no such thing as an oversided pump..
a much larger pump will not evac any faster in many cases..
The water may freeze for awhile.. but it will "sublime" (vaporize
directly)... or melt as heat returns to the area.. It takes lots
of water to make chunks of ice.. and your system is probably
ruined if it had that much water..

>       The point is that some moisture being left in the system has
>       a high probability. 

Very true


>Greg Marciniak

From: (George Goble)
Newsgroups: sci.engr.heat-vent-ac
Message-ID: <4cm6pv$>
Subject: Re: Refrigeration oils - boiling pressures
Date: 6 Jan 1996 16:08:31 GMT

In article <4cm45g$> Lorne Lowry <> writes:

>Some time ago I posted a similar question but received only partial
>responses.  By partial, I mean partial.  Half posts etc.  The
>balance(s) of the posts are still swirling in the ether.
>Does anyone know the actual boiling pressures of refrigeration oils?
>I know Copeland has stated that mineral oils boil at approximately 250
>microns.  Is this figure accurate?  How about POE's?  Is there
>information available about the boiling pressures for these?
>Considering their affinity for moisture, specifiers are demanding, I
>believe, vacuums which are too low and may be detrimental to the
>integrity of the oil.

250 microns is about right for the mineral oils.. I am not sure about
PAG oils, but I have heard that PAG is much more volatile..
Deep vacuum will not screw up mineral oils, you can pump for
days, and only boil out a teaspoonfull or less, which contaminates
your vac pump oil, and this self limits the vac to 250 microns.

From: (Bruce Hamilton)
Newsgroups: sci.chem
Subject: Re: Drying silica gel with vacuum
Date: Thu, 23 Jul 1998 18:24:20 GMT

Uncle Al <> wrote:

>Richard Kinch wrote:
>> Uncle Al <> wrote:
>> Do we understand "microns" the same way?  The serviceman's "25 micron"
>> pump achieves a vacuum of 0.025 mm Hg (0.00049 psi) absolute pressure.
>> The boiling point of water becomes -51 deg C. Is that poor by
>> laboratory standards?
>25 microns would be a piss-poor lab vacuum.

Given that the most common laboratory vacuum pumps I've encountered
have been water venturi systems ( approx. 15 torr ), oil-free
mechanical vane pumps ( 100 torr ), and water ring pumps ( approx
1 torr ), then 25 microns ( 0.025 torr ) would be an excellent
laboratory vacuum system in almost every laboratory I've worked.

Why wouldn't you normally use micron pumps in a general chemical
laboartory?. Well, some of the most common vacuum operations are
rotary evaporation and distillation, and high vacuums mean
that the solvent will not condense using tap water in the
condenser cooling coils. For example, we use a water ring main
vacuum system on samples first, followed by a 2-stage high vacuum
system for removing final solvent residues.

Now, if we were talking about high vacuum systems - such as using
single and double stage oil-sealed rotary pumps, then we can start
talking about microns - but then that's only in the catalogue,
because any working single or double stage vacuum pump and system
will be pumping vapours and thus have vacuums in the 10s of

For example, common high vacuum pumps here are the Edwards E1
( 1 stage ) and E2 ( 2 stage ).
Their specified new performance ( micron )
                                    E1M5           E2M5
Without gas ballast                  20              1
With gas ballast                   1500             20
max. inlet H2O vapour pressure       25             10

Note that the pump size, ballast, and max. inlet vapour pressure
help define how much water a pump can remove, as wel as the actual
achievable vacuum. Once used on solvents ( including water ), these
systems often are operating above 5 micron.

>Any lab pump that didn't pull 5 microns would be in line
>for overhaul.  An adequate direct drive forepump with
>direct drive oil would go below 1 micron.

Not in my experience, our routine pump tests last month gave
between 4 and 30 microns for our high vacuum pumps - the actual
values often depend on the history of the pump oil and the care
during use to prevent condensation, not just the condition of
the pump. I've never encountered single stage pumps ( used to
be the most common, but now 2-stage are price competitive ) that
could pull 1 micron on real systems.

>> Here's the application:  Air conditioning systems contain a silica gel
>> canister in the refrigerant circuit to remove damaging moisture from the
>> refrigerant-oil mix.

I can't speak for the USA, but here many of the hermetic systems use
Molecular Sieve 4A driers, which have a much greater capacity ( 22% mass ),
and less volume is required. They are not able to be regenerated because
the required temperature to remove water ( 220C - 350C ) will not remove
the oil - and the heated oil degrades on the active surfaces and destroys
the ability to absorb water, and higher temperatures result in physical
breakdown of the sieve. Older systems used much larger silica gel driers.

>>  Whenever the system is opened, one is supposed to
>> replace this drier component with a fresh one.  Technicians commonly
>> dismiss that expense (in my experience they are not that familiar with
>> the replete scientific literature), and believe (as an article of faith,
>> not science) that pulling a 25 or 50 micron vacuum for an hour will
>> evaporate all the moisture in the circuit piping, heat exchangers,
>> residual oil, AND (most critically) in the silica gel.

Actually, many technicians here do replace the driers - and also realise
that the vacuum is mainly to remove non-condensable gases ( Nitrogen etc.),
and the long time is required because they are pulling gases through
narrow lines - such as the restrictor that separates the high and low
presssure sides. The non-condenable gases are a real killer if not
removed. If a good vacuum is achieved, and the driers are unlikely to be
already saturated with water ( catastrophic failure - such as a accident
breaking a line ), then the serviceman may take a chance - but driers are
relatively cheap compared to labour charges, and one technique for wet
systems is to use a cheap drier and replace it after operating the system
for a few hours.

>I wouldn't bet on the silica gel going dry, not even a little bit.  When
>I dry out a vac line (KOH bath, rinse, dry.  The main line doesn't fit
>in any drying oven) I use a heat gun from the furthest points moving
>toward the pump.  That's at least 100 C.

I've never, and would never, use this procedure, but if it works for you...
It would be extremely unwise to use on a used refrigeration system, as well,
because there is a layer of oil on the inner surface, and the moving heat
heat source results in two cold traps....

>I wouldn't bet on a pool of oil drying out, either.  Vapor pressure
>depends on mole fraction in otherwise inert solution.  1% water
>dissolved in oil has 1% its nominal vapor pressure at temp.  A little
>droplet of water under an inch of oil doesn't see 25 microns ambient
>pressure, it sees 25 microns plus the one inch head from the water.

Most "wet" systems don't havee lots of free water, as the water has
usually reacted with aluminium or copper, or diffused into motor
winding insulation -serious contamination id junk time. For minor
conamination the compressors have to be removed, drained of oil and
baked in a warm oven - hence the usual attempted to trap water using
a temporary drier.

>>  While any free water will evaporate,

Only if the high vacuum is applied long enough - most recommendations
assume a dry system, and the time is to remove non-condenable vapours,
rather than water - as the system is expected to be reasonably dry.

>> my question is whether that such a vacuum at room
>> temperature is sufficient to remove the water from silica
>> gel (doesn't have to be perfect, just enough to restore most of the
>> potential of the silica gel).  It is not practical to heat the silica
>> gel, at least not to anything like 110 deg C, although if it really made
>> a critical difference in this maneuver it would be worthwhile to know
>> as a resort when replacement driers are unavailable.

Actually, with the oil present, the vacuum alone will not regenerate
a silica gel drier to any major extent over reasonable timescales.

The other problem is that rapid water accumulation or removal tends
to break the silica gel crystals ( and also wet molecular sieve beads )
into fines that are very abrasive. The service manual recommendation
to replace driers is sensible and the best option for the client. It
was possible to partially regenerate used silica gel driers in special
drying rigs ( not the same rigs that are used to activate new silica
gel before use ), but I suspect that the cost is not worthwhile,
especially as a failure to remove water can really accelerate
corrosion and destruction of motor windings.

The design of hermetic refrigeration systems requires that water
should be removed as quickly as possible, but the narrow restrictions
necessary to provide both high and low pressure sides, the oil
throughout the system, and with the requirement to only have one
charging point, means that most of the guidelines and techniques for
pumping down large bore laboratory vacuum systems are inappropriate.

             Bruce Hamilton

Index Home About Blog