PDA

View Full Version : evacuation



reggie
17-12-2002, 07:15 AM
Ive just started a rebuilding of a recip chiller. The heat exchanger has burst its banks and flooded the entire system. This was probably due to poor commissioning which we'll investigate later.
However today we drained all the water out of the condenser coil with the help of nitrogen and then got the vaccuum pump ready.
Here is where my arguement starts.
I was talking to the old man who im working with about the vaccuum process. I was willing to take the torr gauge down to 4 mbars even if it takes all week but...
He claims if we take the vaccuum too deep then any moisture in the system will freeze and wont come out???
I always thought the whole process was based on the fact that if we lower the pressure in the system then the boiling point of water lowers and thus turns into a vapour state and is removed by the pump. Hence at atmospheric pressure i bar water has to be 100c before vapourising and at 2 mbar approx it can boil at -6c. He thinks that by decreasing the vaccuum to a deep level that we will lower the water temperature and freeze it??
He is adamant that hes right. I didnt want to argue with the guy too much as hes a lot older than me so i must be respectful but whose right??
Am i barking up the wrong tree?

Argus
17-12-2002, 11:57 AM
By and large he's right. If the vacuum pressure gets too high and the temperature of water in a system gets too low, for instance on a roof in the UK in the winter, it can freeze, then you will end up removing the water through sublimation instead of evaporation, which takes for ever. You definitely need to keep the vapour above freezing.

Ultimately you are working with the humidity within the system, therefore you need to increase the temperature of the whole chiller to maximise evaporation and to control the flow rate to keep the vapour temperature above freezing. In addition you will need a good two stage pump with a gas ballast system. The danger with dragging too much neat water too fast from a system is that it will condense in the oil of a vacuum pump and ruin an expensive bit of kit. At one time we used to use a freeze pot to condense the water on a cylinder that contained a mix of dry ice and R11, but that's illegal now. That way you could knock out most of the water before it got to the pump. Periodically you can save some time by flushing with nitrogen to carry out much of the vapour. I used to do this while we cleaned the vac pumps and re-oiled them.

Many, many years ago, Carrier had a bulletin on dehydration, that showed how to make some of the kit you would need, like a hygrometer in a glass tube, but it must be long out of print.

Some tips:
? Take all the guts out of all the valves and components.
? When you reassemble to put it on the vac pump, do a pressure test. There's no point evacuating a system full of holes.
? Keep the connecting pipes as short and as big a diameter as possible.Take out any shraeder valves.
? Put solenoid and NRVs in line on each vac pump.
? Control the vac with a couple or three medium pumps rather than one big one.
? If it is outdoors, cover the whole thing with a sheet, insulate with rock wool slabs, then sheet again - try installing some form of low grade heat inside to get the temperature up, 25 - 30 deg is sufficient. That way you will increase the evaporation and flow rate.
? Try to find a way to measure the humidity inside, by measuring the humidity of the exhaust gases - a high vacuum inside does not indicate a dry system.
? When you start up the system when finished, have plenty of driers handy for both liquid and suction lines. Change them after a few days.
? Run the comp for a few hours and then change the oil. Keep doing this until it checks clean and acid free.

Hope this helps, good luck
________
marijuana vaporizer (http://weedvaporizer.info/)

Andy
17-12-2002, 09:38 PM
Hi, Reggie, Hi, Argus:)
Argus your advice is sound and some of the best I have see to date.
Reggie, chin up mate, I am going through the exact same thing ata the moment with a 350 kW flooded R22 chiller which happen to burst a tube (might have something to done with the water return plate being fitted back the frount grrr:mad: ) Anyway whilst waiting for the tube plug I used 5 large nitrogen cylinders blowing through the vrious system sections and with a little help from an other engineer we stripped, cleaned and oiled the top end of the compressor and the cylinders. The crankshaft was then flushed through along with the shaft seal.
When I got the plug I pressure tested the whole system, sectioning it off and leaving the test over the weekend. I had to kit a number of the main shut of valves and tighten up the odd flange.
They I started evacuation, with the gas balast open, breaking the vac every so often with nitrogen. Gradually I closed in the balast with each time the vac was broken. At this point I have to confess I did not fit a torr gauge, ultimate vaccumn is not what I was looking for. After observing that no more steam was comming from the pump and that the pump was not displacing any air (feel for at blow at the pump outlet) I broke the vac.
I will no return in the morning to change the oil and filters
+ driers, on inspection during running the liquid line had no indication of wet refrigerant after the cores and the oil was not discoloured, although I did notice some very nice ice floating on the liquid refrigerant of the chiller (system runs as low as -4 deg c evap for a 2 deg c water temp, backed up with hot gas bypass)
At present I am well pleased considering that I drained maybe 80 gallon of water out before I came to the compressor.
Tomorrow I will pump out the chiller and decant the gas before blowing the ice out with nitrogen. In a way I am lucky in the fact this is a flooded chiller so we can let the moisture make it's way back to the chiller and remove it there.
My main points would be.


Remove the main moisture by using copious amounts of OFN
Pressure test the system and repair all leaks
Vaccumn out the system with a pump utilizing the gas balast function
Tripple evacuate often and don't pull a deep vac
Change the oil in your vac pump often, synthetic fridge oil will do
Break the vac when your are happy that no more moisture will be removed
Strip and clean your pump after use, filling with new vac pump oil
Change the oil, filters and driers often in the next 100 hours


Hope this helps, it may not all be right, but it's the best way I've got at the moment, I'am still perfecting it.
Regards. Andy:D

Jai
17-12-2002, 10:44 PM
Hi Andy, could you tell me what Type and Brand of sealant(if any) you used in plugging the burst tube.

reggie
18-12-2002, 07:19 AM
In response to the above feedback i worked in a chiller rental factory many years ago. This firm specialised in medium tonnage applications such as ice rinks etc. On numerous occasions we had to rebuild most of the system due to a burst heat exchanger. This involved removing all compressors with a fork lift truck rebuilding the components and pipework and replacing the shell and tube evaporators.
The condenser coil was put on vac for a week along with heaters blowing onto the surface area of the condenser coil itself and the vac pump oil drained every night (looked like milk) and changed. This process was carried out every day, whilst rebuilding the system until the torr gauge decreased to 4mbars. Thats a deep enough vaccuum isnt it? Why did we evacuate so low when there was a danger of freezing moisture??? Whatsmore are we saying the torr gauge although satisfied is giving a false reading??

Andy
18-12-2002, 09:46 PM
Hi,:)
nice to hear from Marc again.
Jai, the plug I used (I would have to take a look to see who manufactured it) is in principle operation the same as a sleeve anchor used to fix steelwork to mansonary.
To fit you ream the copper/cunifer tube to within a tolerance, then you tap in the plug until it is fully home, after which you torque the plug to 40 foot pound and strength test as required.
Marc I like your idea with the light bulbs, I seem to remember an old spark I worked with telling me about using this same method to warm motors that had been sitting about before firing them up.
What voltage would you apply, 240v a/c?
Anyway in my case I am lucky the motor is external on a direct drive.
I was working on this plant today, changed the driers and decanted the chiller barrel (flooded R22) I was then able to blow a nice handfull of ice out of a gutted liquid feed shut off valve on the bottom of the chiller. Actualy this one has worked out well with the moisture mostly in the chiller barrel. I think I will leave the plant a day or two and repeat this performance as it is working well.
Reggie's chiller maybe small or a plate design in which case frozen water would block it up.
As for the vacing out of the condenser the heat was the reason low vacs could be pulled without moisture freezing, the kW input being higher that the refrigeration effect from the water being vaccumned off.
Regards. Andy

Argus
18-12-2002, 10:41 PM
Andy, Reggie, this is turning into an interesting thread on the techniques of dehydration

It seems that there are a lot of people who have evolved their own techniques of dehydration over the years by experience, as I have, though I have been ?off the tools? for a long time now.

The main problem, as I recall, was in convincing the Pratts that be that it could not be done by tomorrow, and would take a long time and cost a lot.

One thing that occurs to me was that we could count on losing about 10% of the tubes in a large evap without too much penalty on the duty as long as they were not grouped closely together and were spread over the whole evaporator. The plugs were bought in from Carrier as a size made for their tubes. They were then sealed with Loctite, though I cannot remember what type. You used to whack a wedge in the end after anointing it all with Loctite.

One thing that I remember that may be useful to Reggie and can be overlooked.

As you have a shell & tube evaporator bundle I would suggest that you do a strength test on the remaining good tubes, Eddy Current or whatever.

Also check that the remaining tubes are still round. This is important because when they freeze, the compression is on the outside of the tube which produces a crushing action. You only need one or two tubes to actually split in order the flood it, but you could have a large proportion of the tubes flattened, but still holding pressure. At the very least they may be weakened. This happened to me once and I could not explain the capacity losses in the bundle.

Taking the bell-end off for a visual inspection is not the first thing you check!

A little trick I had to gauge the roundness of the tube was to get some single electrical cable, longer than the bundle itself, a single core 2.5 mm will do. Form a knot in it at one end with a diameter enough to clear the inside of the tubes and then draw it through each tube in turn. If there is no clearance in the tube for the knot to go through, suspect that tube as being deformed. It sounds long winded, but it is a check that can pay dividends.

Good luck
________
Ford VH platform specifications (http://www.ford-wiki.com/wiki/Ford_VH_platform)

Jai
18-12-2002, 11:40 PM
Thanks for the info on tube plugging guys. You guys are very helpful.

Andy
19-12-2002, 08:46 PM
Hi,:)
Marc thankyou for the info on the bulbs, quite simple really, hopefullly the opertunity will not arise too quickly for me to try that out.
As for the alt.hvac thing I will take a look. Could some of that info be posted in the form of an articile on the home page of refrigeration engineer, it would be very useful.
Regards. Andy.

david.findlay
20-12-2002, 06:05 AM
I have always found that preperation is the key when dehydrating large systems.
Purging with OFN & draining from all points possible is critical.
Using a large vac pump or pulling down to quickly is a recipie for disaster.
Depending on the ambient you can easily obtain a low torr reading with a system full of ice.
The only answer is heat & frequently breaking the system with OFN.
To cut down on the amount of oil used on really wet systems I've always used a "Cold Trap" fitted between system & Vac inlet.
The oil condition in the vac pump can be a good indication of the state of the system.
When you have achieved a good reading, say 2 Mbar a rise test whilst adding heat wherever possible can be infomative.

Andy
20-12-2002, 09:01 PM
Hi, I think we are all singing from the same hymn sheet, in some much as we all agree large vaccumn pumps are not the way to go and that heat is essential to boil off the moisture. Also OFN is important for flushing the system and for pressure testing.

I once went to a site and a fellow engineer commented to me that he was having problems otaining a good vaccumn in the system he was working on, maybe the pump was deffective. I then asked had he pressure tested the system, he said he had, I said what about we do a vaccumn test, so we did, with the vaccumn seen to rise quite quickly it was apparent the pressure test was not carried out properly. To cut a long story short the system had to be retested, then vaccumned out, removing the extra moisture added during his inital attempts at a vaccumn.
The moral would be to pressure test properly and the vaccumn is assured after that.
Regards. Andy.

reggie
22-12-2002, 08:33 AM
Andy very good point regarding the pressure testing of systems/evacuation. Ive taken a few short cuts myself in the past and sods law payed for it. Whenever ive been in a hurry and ommitted a proper pressure test you can bet your life i have trouble with the torr gauge coming down. Then you get that sinking feeling of not knowing if its a leak or moisture in the system. Thus your back to square one with egg on your face and deservedley so.
Theres a moral to this story but i just cant think what?


One question i would have if your reading this Prof Sporlan can we calculate the amount of time a vaccuum pump must be on by its cfm rating and the total size of the system condenser/evaporator and its associated pipework run?
Obviously its never going to be an exact science due to the variables of moisture or traces of ***** in a particular system but is their an approxiamate estimation we can use?

Andy
22-12-2002, 12:30 PM
Hi, Reggie:)
If your were thinking of working out how long the vac pump should be on ambient temperature would be one of the main factors.
If your are working in the winter a job takes quite a bit longer to acheive a decent vaccumn.
Regards. Andy

reggie
29-12-2002, 12:50 AM
Marc i can understand what your saying about the rise test but check this one out. Many years ago i was working with a large firm in England. I wont mention any names but it began with D and ended with O. We were vaccing systems out before recharge/commissioning procedures could take place. Their engineer was pulling down to around 6mbar and then leaving the circuit with pump switched off. He claimed that a 10 minute test and a creep back of around 2 mbars would indicate we were ready to charge?? Was he right or should there be no creep back at all??
Add to that ive encountered situations where the contracts engineer will sit and monitor the gauge with you and if any creep back occurrs we start again before he will sign off your sheet.

CuGe
14-08-2005, 06:33 AM
Hi Andy,

I used that type of plug when they first came out, the rep came down and fitted it but it was not successful. I prefer the tapered brass plugs, there are quite a few companys that make them for the specific size of tube which ensures the correct angle of taper. If you ask the manufacturer of the vessel, they will be able to tell you the length and the angle of taper. That, loctite 275 and a hammer will see you right!!

CuGe
14-08-2005, 06:35 AM
As you say, you must use that ballast valve and not vac too fast, it's a long drawn out process.

Peter_1
14-08-2005, 12:52 PM
This threat started even before I was a member.
We once had a chiller with soem leaking pipes.
We measured the inner diameter of the copper.
We then went to someone with a lathe (waw, a new English word for me) and let him make a brass plug with a steel cone in it.
We inserted the brass plug in the copper tube with Loctite and the steel cone was jammed in with a hammer also with Loctite, expanding the brass plug. I think I have somewhere some spare plugs.
It'sn ow running more than 10 years.

Regarding the making up of ice inside the unit: I think you can't vacuum fast enough to let freeze the water, given that the job is done at room temperature conditions. Surrounding heat will heat up the water inside the chiller faster then the pump can remove.

I saw in the Copeland factory that they connect a DC power supply to the windings while vacuuming the scrolls.

We did it in the past with a welding transformer and connected the Kriwan as as safety device in line with the transfomer.

Something besides bacuse I mention now the Kriwan: if you ever need a liquid level switch to control for example a pump in a well, you can use a Kriwan for this.:cool: