PDA

View Full Version : how to put compressors in parallel



Lc_shi
15-02-2006, 09:18 AM
What should be considerd by connecting two or three compressors in parallel? install a oil separator in discharge line and return the oil to suction side at times. Is it correct? welcome your input. thanks in advance!


best regards
LC:confused: :)

US Iceman
15-02-2006, 01:58 PM
LC,

Oil recovery is important as you suggested. The oil could be recovered by an oil separator and the oil levels equalized between the compressors.

I have seen this done two ways.

In one application the the oil level was equalized by a common line between two crankcases. In this system no oil separator was used.

Peter_1 also posted an installation bulletin about this sometime ago. This was another method of piping multiple compressors in parallel without the oil separators.

The more common method is to use an oil separator and the oil level regulators to maintain a constant oil level. You also need an oil reservoir for this system. The oil is returned from the separator to the reservoir. The reservoir then supplies oil to the individual oil level regulators.

Best Regards,

Rob S
15-02-2006, 03:46 PM
The other issue is proper discharge and suction piping.
Copeland has some great service bulletins on this.

Basicly all the discharge lines will flow in one direction. To prevent head locking.
All the suction lines will be Tee'd and Tree'd together or use of a header.

Carlo Hansen
15-02-2006, 05:22 PM
Hi Ic_Shi

Do not forget the one way valve in the high pressure line
from the compressors to the condenser.

Best regards

Carlo Hansen

Lc_shi
16-02-2006, 01:54 AM
thanks for all your input.
I'll check with compressor suppliers.

Lc_shi
16-02-2006, 02:25 AM
I've got a sketch.pls have a review and make comment.

thanks

rgds
LC

Argus
16-02-2006, 12:45 PM
Generally it is common to parallel 2 or more compressors in air conditioning chiller applications fitted to single evaporators and condensers - these are usually semi hermetic reciprocating compressors and the systems are not fitted with oil separators as a rule.

The method is simple and in my experience trouble-free.


All compressors must be level.

Discharge and suction pipes are connected together at the compressors with equal pipe and fittings.

Oil lines are interconnected between each sump.

The sump space above the oil levels are interconnected. (Carrier, for example, in their large packaged chillers, use a single 28mm pipe between the sumps connected to the spare sight glass flange; as this is at the optimum oil level, it maintains the levels of the oil in all the compressors and equalises the sump pressures at the same time in one pipe).

________
Toyota JZ engine history (http://www.toyota-wiki.com/wiki/Toyota_JZ_engine)

Servicefrigo
16-02-2006, 06:43 PM
I think is not the same things in low applications were is apsolutly necesary to have an oil separator
Regards

Argus
16-02-2006, 07:07 PM
I think is not the same in things in low applications were is apsolutly necesary to have an oil separator
Regards

Never said it was low temperature - in fact the oposite.

.
________
Yamaha FZ1 history (http://www.yamaha-tech.com/wiki/Yamaha_FZ1)

Peter_1
16-02-2006, 07:08 PM
Lichuan,
Your sketch is +/- right just for 1 thing: you may not use an NRV between oil reservoir and suction but yoy have to use a NRD which keeps the pressure of the reservoir +/- 1,5 bar higher then the suction pressure.
Otherwise, there's no possibility (no pressure differential) that the oil can flow to the sumps.

Servicefrigo
16-02-2006, 07:43 PM
Sorry the message wos't for you Argus.
Regards

Peter_1
16-02-2006, 07:47 PM
I think is not the same things in low applications were is apsolutly necesary to have an oil separator
Regards

If the compressors are connected together like argus specified - crankcases connected together - then I should never install an oil separator, even in a freezer application.
Give me some, or even better...one good reason why you should need this with 2 compressors in parallel in the case Argus explained.

I never installed one on a freezer.
But we discussed this already many times, porely piping is the most important reason why you should need an oil separator.

chillin out
16-02-2006, 09:31 PM
Pipework looks ok to me, Do you know about all the electrical componants?

S/V in oil lines, Might need oil sep heater.

What size of system is this?

If its only small you can just fit oil sep feeding straight into common oil equalizer. No need for oil reg devices. (this way must have oil sep heater).


Chillin:) :)

Peter_1
16-02-2006, 09:56 PM
This is a prallel setup with 3 compressors where you don't need an oil separator.

chillin out
16-02-2006, 11:23 PM
Peter, I would hate to change that middle comp!:eek:

What would happen if 1 comp was leaking across the head? would it fail the whole system? Theres no place to isolate an indivdual comp, is there?

Chillin:) :)

Peter_1
17-02-2006, 07:53 AM
There excists blind flanges to isolate a compressor but in this case, there are only 2 suction valves for 3 compressors. How's that possible ;-)

But why should you need to isolate one? These compressors are available of the shelve. If one fails - which is very rare with Bitzer - you get a new one, remove the end plates and you then remove the failed compressor.

The only s**t is that you have to remove all the oil first and no matter how well you have done this, it's allways a greasy mess when you replace one.

If it leaks.. well...that's a problem which can occur with every pack.

Josip
17-02-2006, 11:41 AM
:) Hello,

Piter,
http://users.skynet.be/vuurwerk2004/bitzer/Bitzer_trio.jpg

Bitzer-Trio is a joke. :) :)

http://users.skynet.be/vuurwerk2004/bitzer/de_echte.JPG

There are only 2 compressors with common suction line, thus mean the right and middle one. Check where are motors. The far left one doesn't exist.

Josip
17-02-2006, 12:56 PM
I've got a sketch.pls have a review and make comment.

thanks

rgds
LC
Hi, LC

Sketch is more or less ok. Speaking about there should be NRV on discharge pipes from each compressor. Also all other needed stop valves not show.

Oil separator to be or not to be installed? :confused:

For that purpose you can use only oil reservoir, but

definitely it is better (not a must) to install an oil separator to obtain better COP at condenser and keep oil in the shortest circuit: compressor - oil separator- oil reservoir - compressor, where oil must be ;) No need to run oil through condenser, receiver, evaporator back to compressor becuse each drop of oil in refrigerant line reduce our COP and cause some other problems :)

NRV at oil reservoir is ok (it must be there) becuse it is closed all the time when compressors are working. There we have enough high pressure to push oil back to crank case when come to low oil level. NRV is there to equilize the suction and discharge pressure at compressor stop. You can push all oil from oil reservoir into compressor crank case due to pressure equilizing and we do not want and do not need that. For that reason that line is as you can se "vapour" line.

Hope this is of little help;)

Peter_1
19-02-2006, 09:51 AM
:) Hello,Piter,
http://users.skynet.be/vuurwerk2004/bitzer/Bitzer_trio.jpg

Bitzer-Trio is a joke. :) :)

http://users.skynet.be/vuurwerk2004/bitzer/de_echte.JPG

There are only 2 compressors with common suction line, thus mean the right and middle one. Check where are motors. The far left one doesn't exist.

Josip, a colleague of my send this picture just the moment this thread was on RE.
Amazing you found that this link was even from someone in Belgium.
Anyway, the flanges exists and the visual trick is very well succeeded.

Speaking of the NRV, I think you're missing something...it has definitely be an NRD instead of an NRV.
If you install an NRV, there is indeed enough pressure to allow oilflow to the compressors.
But...for many floats, this HP will be far too high.

And second as soon HP in the oil reservoir reaches HP of the system, flow towards the oil reservoir will stop.

NRD is there to assure a constant differential pressure of +/- 1.4 bar between oil reservoir pressure and crankcase pressure.


NRV is there to equilize the suction and discharge pressure at compressor stop. You can push all oil from oil reservoir into compressor crank case due to pressure equilizing and we do not want and do not need that. For that reason that line is as you can se "vapour" line.:confused: :confused:

Sorry, but this information is in my opinion not correct. I think you have to read trough some service bulletins of AC&R, Henry, Parker or Sporlan.

This is one http://www.sporlan.com/110-10.pdf

http://www.henrytech.com/techtips_pdfs/HT-TT6.pdf

Josip
19-02-2006, 07:07 PM
Dear Peter,

You may or you may not agree with me. Hope you will agree about many different solutions for the same problem.


Josip,

NRD is there to assure a constant differential pressure of +/- 1.4 bar between oil reservoir pressure and crankcase pressure.


This example is shown on Sporlan's high pressure oil return sketch Figure 1 (BTW, issued 20 years after my first contact with much better design of STAL - Sweden) and that one (in my opinion) is not good, but I will try to explain that later on.


Speaking of the NRV, I think you're missing something...it has definitely be an NRD instead of an NRV.
Check it again, please.

NRV is there to equalize the suction and discharge pressure at compressor stop. You can push all oil from oil reservoir into compressor crank case due to pressure equalizing and we do not want and do not need that.For that reason that line is as you can se "vapour" line
I said to equalize suction and discharge pressure at compressor stop. Equalise of pressures will stop oil flow from oil reservoir preventing crank case to be flooded with oil in case if we have small leak at oil level regulators-floats due to pressure difference. Oil reservoir can be installaed lower than compressors but high pressure will still push oil into crank case if we not equalise it. Schematic drawing is not real one and sometimes what seems to be up (under ceiling) can be really down (at floor).


:confused: :confused:

Sorry, but this information is in my opinion completely wrong. I think you have to read trough some service bulletins of AC&R, Henry, Parker or Sporlan.

This is one http://www.sporlan.com/110-10.pdf
http://www.henrytech.com/techtips_pdfs/HT-TT6.pdf

Any different information can be wrong (at first sight), if I know only one way how to solve the problem, but then if I do not understand I must ask for explanation. There must be some other solution.....Anyhow, thanks for links:)



If you install an NRV, there is indeed enough pressure to allow oilflow to the compressors.
But...for many floats, this HP will be far too high.

Why? You can use crank case float with the orifice and needle. This float can be installed inside or outside of compressor's crank case. We have to take care about installation height to obtain proper oil level in crank case. Loosing oil from crank case float goes down opening orifice and leave oil into crank case until proper level. No need to say that the float is under suction pressure. Orifice and needle in this case is metering device (expansion valve) helping to evaporate remain refrigerant from oil. BTW for Danfoss HP float valve MWP (maximum working pressure) is 28 bar?


And second as soon HP in the oil reservoir reaches HP of the system, flow towards the oil reservoir will stop.
I can agree with you, in case when we have two vessels oil separator and oil reservoir, connected as shown on Figure 1. Coming to equal pressure, oil flow will stop.

Speaking about Figure 2 (one vessel as oil separator and oil reservoir), pressure relief valve we definately do not need.

Here I want to come back why solution with Oil differential check valve (Figure 1) or with Pressure Differential Valve (Figure 2) is not good.
Using any of valves we have a constant by-pass flow from high pressure side to low pressure side. Any amount of by-bypass flow is lost of work done by compressor.

We can utilize HP gases (no need to reduce that pressure) to return oil back to compressor's crank case.

Coming to final design....there are many ways to do it, but definately not all of them good.

liverpool Eng
19-02-2006, 11:23 PM
The oil floats may take the high pressure but not the diff.
for aplications above a 2 bar diff i would use a electronic version like oiltrax that has its own solinoid and pulses in the oil.
This stops the foaming and passing that give problems with the mech versions.
otherwise the NRD at 1.4 bar is correct for most aplications.
And if we are going to use a high pressure oil syestem then i would recomend the combind oil sep and res avaleble from the likes of Tempright.

Peter_1
19-02-2006, 11:38 PM
If it's a NRV and the line from that NRD is going straight to the suction as Lichuan sdrawed it, then you need a SV in that line to isolate it while the compressor is running.

Otherwise, like you said, you will have a constant bypass flow from oil reservoir at HP to the LP part of the sytem. (loss of compressor capacity, danger for strongly superheated suction gasses...)

If you monitor a sight glass just after an oil separator, you can notice an almost continuous flow.

I never saw this setup like I never saw a Danfoss oil-float.

Most of this equipment can widhtstand a high pressure for pressure testing purposes but this isn't the working pressure.


We have to take care about installation height to obtain proper oil level in crank case
I never saw another setup then those direct bolted to the compressor crankcase, eventually with an adapter between it. So oil level is allways right.

All the HP float valves I know are equiped with a SV mounted direct on it and these can be fed with oil under high pressure. But there isn't a needle in it anymore but an electronic sight glass.
Can you therefore provide me some more info of a float valve which can operate at an oilfeed pressure equal at that of the discharge pressure?

Lc_shi
20-02-2006, 04:15 AM
Many thanks for for all your constructive ideas:)

I conclude that the follow things need to add to my sketch:
# each discharge pipe install a NRV for preventing interact of compressors;
#NRD to keep a fit higher pressure than crank case pressure to make the oil flow back to the case;
#install an SV before oil level regulator for there only single compressor running situation?(not too sure):confused:

regards
LC

Peter_1
20-02-2006, 07:45 AM
...
# each discharge pipe install a NRV for preventing interact of compressors;
#NRD to keep a fit higher pressure than crank case pressure to make the oil flow back to the case;
#install an SV before oil level regulator for there only single compressor running situation?(not too sure)

LC,
I noticed that other have different views of how we allways do it here according to the specs of the materials we use but what concerns you last post now.

Placing a NRV in each discharge...we don't do it. Many compressors have a build-in NRV and even then...the HP valves in the compressors acts also as som sort of NRV.
In a standard set-up with 1 compressor connected to 1 evaporator and 1 condensor, you don't mount an NRV neither.
After standstill, there isn't also no equalising flowback from the condensor through the compressor. The valves acts then the same way as a NRV.
And if there's a flow, then the valves or the compressor needs to be replaced.

NRD, if you ask me.. I should install one if the supplier tells you to do it and I never saw it otherwise, unless with the electrically pulsed (expensive) oil floats.

You only need an SV before each oil flat in case the needle in the float leaks. If it leaks then you need an SV or you have to replce the float.
Install anyhow a special oil mesh filter in the oilfeed line.

Regarding the system JOSIP mentioned, I still don't see how the oil can flow to the oil floats without an SV in the line to the suction.
As soon the oil separator 'pumps's some oil to the oil reservoir, the pressure equalises immediately to the suction pressure via the equalising line. What is then the meaning of an NRV in that line? There can never be a flow backwards because suction pressure is allways lower then the reservoir pressure?:confused:

Peter_1
20-02-2006, 02:34 PM
Josip,

What is very surprising: how have you found that link of the Bitzer joke that fast?
It was made only some days before I posted it by a colleague of my who lives +/- 30 km from here.

My wholesaler is very interested how this could go so fast unless Josip = Gi... N.... himself :cool: :D ;)

liverpool Eng
20-02-2006, 07:28 PM
If it's a NRV and the line from that NRD is going straight to the suction as Lichuan sdrawed it, then you need a SV in that line to isolate it while the compressor is running.

Otherwise, like you said, you will have a constant bypass flow from oil reservoir at HP to the LP part of the sytem. (loss of compressor capacity, danger for strongly superheated suction gasses.If you monitor a sight glass just after an oil separator, you can notice an almost continuous flow.

Most oil sep i have worked with have there own float and only pass oil not discharge gas unless broke.

would this not be the case with a single comp with its own oil sep conected to the cranckcase ??
The SV in the line is there to prevent passing when offline and on a delay to pervent liquid lifting the float in a cold sep.

Josip
20-02-2006, 09:10 PM
LC,
I noticed that other have different views of how we allways do it here according to the specs of the materials we use but what concerns you last post now.

NRD, if you ask me.. I should install one if the supplier tells you to do it and I never saw it otherwise, unless with the electrically pulsed (expensive) oil floats.


Following some design we must install what is in design, still sometimes we can choose maual type or electric and of course brand :)

I have to explain something. I was and still today I am connected to large industrial ref/AC/process plants.
Therefor all solutions we use are sometimes little strange to refrigeration people working in some other fields


Placing a NRV in each discharge...we don't do it. Many compressors have a build-in NRV and even then...the HP valves in the compressors acts also as som sort of NRV.
In a standard set-up with 1 compressor connected to 1 evaporator and 1 condensor, you don't mount an NRV neither.

This is OK


The valves acts then the same way as a NRV.
And if there's a flow, then the valves or the compressor needs to be replaced.

What in case of scrolls or screws?


You only need an SV before each oil flat in case the needle in the float leaks. If it leaks then you need an SV or you have to replce the float.
Install anyhow a special oil mesh filter in the oilfeed line.

SV is welcome and filter too. Using SV gives more safety and some other way of control.


Regarding the system JOSIP mentioned, I still don't see how the oil can flow to the oil floats without an SV in the line to the suction.

Maybe I do not understand this, but I will try to explain.
We have 2-3 or more parallel compressors with comon oil separator/reservoir (in one vessel). Thus all oil coming from compressors is collected in that one oil tank under discharge pressure. From that tank we have separate pipes to each compressor's crank case. There we can have SV (ON with compressor) in front of each float. Float valve is constructed to be in connection with oil level inside of crank case becuse that level is one we have to control. On other side i.e. in front of orifice and needle we have high presure oil waiting to go back to crank case.


As soon the oil separator 'pumps's some oil to the oil reservoir, the pressure equalises immediately to the suction pressure via the equalising line. What is then the meaning of an NRV in that line? There can never be a flow backwards because suction pressure is allways lower then the reservoir pressure?

If we have two vessels we can use Pressure Reducing Valve connected to suction side to reduce pressure in vessels, but only for about 1,5 or 4 bar to obtain oil flow. OK high presure more faoaming but bigger pressure reduce = bigger lost. For me this design is not good but...Who is going to ask me;)

With one common vessel we have NRV, but I see it is maybe better to call it by-pass valve. When compressors stop discharge pressure imediately falls and spring open that NRV/by-pass valve to equalize pressures. There we can have also SV with timer.

About Bitzer compressors I found that by accident. First I saw the "wrong" one but I was in hurry. There was something wrong to me but I was thinking it must be maybe for some expriment or similar. Switch off my laptop and went out.Later on I remeber it again and looking for link I found answer and link to "right" picture. Luck!;)

Peter_1
20-02-2006, 09:34 PM
Josip, you're completely right about the check valve for scrolls and screws. I was only thinking on pistons.

This is my colleague's website - in Dutch of course - and I almost bet you were him because this was so fast that it couldn't be true.

www.nemegeer.com

Andy
20-02-2006, 09:38 PM
Hi NRD/NRV is a relief valve, a bit like a check valve, but with a spring pressure to defeat before it opens.

Kind Regards. Andy:)

Gerrit
27-12-2008, 10:14 AM
...cut
This is my colleague's website - in Dutch of course - and I almost bet you were him because this was so fast that it couldn't be true.
...cut



Hi Peter,
I just read your message, I'm the only real Gerrit from belgium.
The biggest and the greatest.
I'm the one who daily works with trio_Bitzers
I make them, I repair them and I sell them
Anyone interested?

:D :o ;)

Peter_1
27-12-2008, 03:12 PM
The one and only Gerrit, manufacturer and inventor of the tripple tandem.
Does everybody knows this new sort of compressor?

Probeer maar een beetje je weg hier te vinden want het aanbod is hier heel groot.

Gerrit did his F-Gas exams some weeks ago when I was chairman of the jury.
He's indeed the greatest refrigeration technician of Belgium, speaking then in terms of length or cm's (or inches for those live on the other side of the water :) )

zaheer
05-01-2009, 08:31 PM
What should be considerd by connecting two or three compressors in parallel? install a oil separator in discharge line and return the oil to suction side at times. Is it correct? welcome your input. thanks in advance!


best regards
LC:confused: :)
Yh.its sure. thi same system in daikin vrv11,vrv111.you can see.

Tom C
19-12-2009, 01:25 AM
We manufacture 2 and 3 compressor parallel systems for the foodservice industry.
On a current system (medium temp only) we have 3 equal scroll, 3 hp compressors. We have an oil separator, reservoir, and 3 level regulators (S-9010). We, as in the past, did not install a pressure valve between the vent line and the suction line. Up until now, we have not had any problems, except with the 9010's, which we are going to scrap in lieu of the OP-02.
The difference here is that this is in a low ambient condition on a rooftop in NYC (instead of an indoor wc system, which is normal for us). The installation tech is telling me that he can't get flow to the level regulators. Although, when the feed line connected to the level regulator is loosened, oil will flow.
We have lost (burnt windings) 1 compressor, so far.

The obvious, would be to install a pressure valve in the vent line to maintain the oil reservoir pressure slightly above the cc pressure. However, on the 2 compressor parallel, we use an equalization line to maintain levels. We have never had a problem with this, in any ambient condition, even though the cc pressure in both compressors is the same. On all other 3 compressor parallel systems, oil return has just not been an issue. Only this one.

The installation tech said that he had to adjust all of the level regulators. When the system left the shop, they were all working and set correctly. Of course, 3k miles on a flat bed tends to screw up any setting.

So, install the pressure valve and eat the compressor,
or, as I suspect, the oil lever regulators are out, and maybe stripped?

Comments would be appreciated.
Thanks.

Peter_1
19-12-2009, 08:30 AM
Insulating th oil separator or heating it slightly with a drain heater will perhaps help. Temperature is perhaps that low and load on pack so small that vessel cools down so deep that you loose pressure.