PDA

View Full Version : Screw compressor theory interesting for some?







RANGER1
21-08-2015, 11:40 PM
http://ethesis.nitrkl.ac.in/110/1/seshiah.pdf

A question at work about oil injection in twin screws, where & why.
Why is oil injected part way through the compression process instead of in suction to maintain discharge temperature within it's limits.

Example Stal screw compressor S70 series
- oil pumps all the same size throughout series although varying capacities, S71,73 & 75.
-oil pump supplies say 120l/min, preference to main bearing, shaft seal & balance piston.
- oil pressure to bearings controlled by pressure regulator to 3-4 bar above discharge Temperature.
Any excess oil above 3-4 bar above discharge is directed into oil injection port to control discharge temperature.
- If main bearings worn more oil flow is directed to bearings, leaving less oil for oil injection to maintain discharge temperature.
- Any oil flow to bearings has a return gallery to lower pressure area of rotors, as well as some spills directly into rotors at suction & discharge end of compressor.

OK still same amount of oil entering compressor, but not in places manufacturer designed it to.

Why does the discharge temperature become to high if not directed into oil injection port (field observation).

From articles i have read can only come up with the following

- oil injection should be injected into the rotors at an area where compression takes place & where oil injection & compressed gas at same temps.
This also seals leakage paths at points where compression takes place.
- where oil is injected very turbulent flow for maximum mixing & heat transfer.
- if oil flows into suction it may not mix with compressing gas as well as centrifugal force throwing it out.


Anyone have any other or further theories or things to add on the matter?
Any observations of their own!

Hope it's not to boring to talk about, but found it to be interesting for me at least

Tycho
23-08-2015, 09:40 PM
quickly browsed through the paper you posted (gonna read through it when I have more time for leisure reading :))

I think your question was answered pretty well on page 0, where it said
In an oil injected twin-screw compressor, the lubricating oil is deliberately injected into the gas stream to absorb the heat of compression. This enables much higher pressure ratio in a single stage without intercooling and it provides protection against corrosive gases

almost the same as having LRI, the oil injection in the middle of the compression provides not only a medium for the compressor to dissipate it's heat too, it also provides lubrication and ensuring a seal between the rotors towards the end of the compression.

the compressor sucks in a set amount of gas, compressing this gas creates heat, now, with LRI, you introduce more gas towards the end of the compression, giving the heat more gas to dissipate between.

on systems without LRI, you introduce oil, doing much the same thing, dissipating heat, but you can't introduce that oil on the suction side of the compressor, because then the oil would take up valuable space needed to compress the gas. too much oil in the suction and you have an oil pump and not a compressor :)

Hence, the cooled down oil is introduced in the middle/end of the compression cycle, to help with compression and cooling down.

On systems with oil coolers, we adjust the discharge temperature with the oil injection, on systems with LRI we adjust the temperature with the TEAT valve, but we still need the oil injection, so we adjust that very very slightly, until we have no foaming in the oil separator.
Because too much oil injection on a system with LRI will cause foaming in the oil and excess oil carryover :)

Segei
23-08-2015, 10:47 PM
As far as I understand this study was done for air compressors. Injected oil for these compressors is pure oil. In refrigeration injected oil have refrigerant as well. Initially this refrigerant will evaporate and it should be compressed to discharge pressure. Injection at late stage of compression require less job to compress evaporated refrigerant.

Tycho
24-08-2015, 01:04 AM
As far as I understand this study was done for air compressors. Injected oil for these compressors is pure oil. In refrigeration injected oil have refrigerant as well. Initially this refrigerant will evaporate and it should be compressed to discharge pressure. Injection at late stage of compression require less job to compress evaporated refrigerant.

I don't think the amount of refrigerant in the oil has much to say once it's introduced into the compression, unless it's a system with a very wet suction, where the oil is foaming like crazy.

RANGER1
24-08-2015, 05:53 AM
Thanks for input.
Difficult to find any hard data, this was the best I could find.
I guess it's in house information for compressor designers & manufacturers.
If anyone has better information, would be interesting.

Tycho
27-08-2015, 09:09 PM
Since you didn't take my word for it, I spent some time and I found this for you:

http://www.petroassist.com/Tech%20Know/ScrewCompressors.pdf
basis in this one says:
All common refrigeration screws use oil injection into the compression area for
lubrication, sealing of leakage paths, and cooling. Injected oil quantities are
approximately 10-20 gal/min per 100 HP. The use of such large oil quantities transfers
most of the heat of compression to the oil and allows discharge temperatures to be very
low even at high compression ratios. Running single stage at 20:1 compression ratio
on ammonia would yield 650 F discharge temperatures in non-flooded types of
compressors. With oil flooding, the screw discharge temperature does not exceed 200F. Running screws at 20:1 or even higher, single stage, though not energy efficient
compared with two stage systems, will not harm the compressor. Many such systems
are running today

This:
http://turbolab.tamu.edu/proc/turboproc/T28/Vol28015.pdf
States the same as above

and this:
http://docs.lib.purdue.edu/cgi/viewcontent.cgi?article=1046&context=icec
States:

13858


and this as a bonus if you want to get into screw compressors with LRI and oil injection :)

http://docs.lib.purdue.edu/cgi/viewcontent.cgi?article=1809&context=icec

in another bonus note, most screw compressors loose their compression at around 20 Hz.
So when running with a VFD, don't let the compressor "coast" when it stops (for those that don't know "coast" it means that the VFD let's the motor go and just let's the compressor run free until it spools down), use the VFD to retard the compressor down to 5 Hz before it let's go.
This gives slow acting check valves in the suction time to close, instead of having the compressor run free for too long and have oil foaming the wrong way through the compressor while it is in the range where it's not compressing anything and past the suction check valve and out into the system :)

Tycho
27-08-2015, 09:12 PM
the post above could also be named "and that's how I spent my last three evenings" :)

RANGER1
27-08-2015, 09:36 PM
the post above could also be named "and that's how I spent my last three evenings" :)


Tycho,
Thanks for your efforts as I did similar but only 2 nights.
You must have settled into married life very quickly:D

Stupid question what is LRI, liquid injection?

RANGER1
27-08-2015, 09:44 PM
Interestingly oil injection in your paper suggests 1% oil injection, but Mycom & Howden 50% of total oil flow.
The older S70 Stals probably 10% but all pumps for there range S71, 73, 75 have same oil flow of 120 l/m.
Mass flow rate does not indicate higher oil flows compared to other machines.
Mycom or Howden might require 240l/min with full lube pump for bearings & oil injection.
Big difference?

Tycho
27-08-2015, 09:58 PM
LRI is liquid refrigerant injection

it's introduced into the compression a little bit further towards the discharge than the oil injection is, or it could be introduced directly into the discharge pipe.
I'm not a fan of the discharge pipe connection, because then you are only cooling the gas leaving the compressor, instead of cooling the gas as it compresses and while you may read a normal discharge temperature, you may have higher temperatures in the compressor, causing "burnt oil" before it reaches the discharge port.

the principle is that you are adding a volume of gas/molecules towards the end of the compression for the heat to dissipate through, so instead of needing an oil cooler to cool the oil, you remove the oil cooler entirely.

you take liquid from the HP receiver, and use a TEAT valve where the bulb is placed on the discharge pipe close to the compressor, adjust the superheat of the valve until you have the desired discharge temperature (with LRI I usually set the temp to 55-60C, while with an oil cooler we would be running with 70-75C).
Using a TEAT with the range +35-+65, my experience is that you need a condensing pressure of +19c to start cooling with a discharge pressure of +30.

it's not really liquid injection, but a mist that is introduced towards the end of the compression, I find it more reliable for a steady discharge and oil temperature on a system then one with an oil cooler and a three-way oil mixing valve.

:)

Tycho
27-08-2015, 10:05 PM
Interestingly oil injection in your paper suggests 1% oil injection, but Mycom & Howden 50% of total oil flow.
The older S70 Stals probably 10% but all pumps for there range S71, 73, 75 have same oil flow of 120 l/m.
Mass flow rate does not indicate higher oil flows compared to other machines.
Mycom or Howden might require 240l/min with full lube pump for bearings & oil injection.
Big difference?

I don't know the mass flow of oil or the percentage, I want to know how much I should open the reg valve, I have asked here, and I have asked howden and I have had the guys in the office ask the compressor package manufacturers, without getting any answers other than adjust it till you have the desired discharge temp.

on the XRV range with oil coolers, it's pretty simple, start with the valve 1/2 turn open, the open until it stabilizes on the desired discharge temperature (usually end up with 70-75C)

on wrv units with oil pumps, especially those from SES, it's very tricky, because the difference between a stable discharge temp and maintaining oil pressure is opening the door to the machinery room and looking at the valve.

RANGER1
27-08-2015, 10:20 PM
LRI is liquid refrigerant injection

it's introduced into the compression a little bit further towards the discharge than the oil injection is, or it could be introduced directly into the discharge pipe.
I'm not a fan of the discharge pipe connection, because then you are only cooling the gas leaving the compressor, instead of cooling the gas as it compresses and while you may read a normal discharge temperature, you may have higher temperatures in the compressor, causing "burnt oil" before it reaches the discharge port.

the principle is that you are adding a volume of gas/molecules towards the end of the compression for the heat to dissipate through, so instead of needing an oil cooler to cool the oil, you remove the oil cooler entirely.

you take liquid from the HP receiver, and use a TEAT valve where the bulb is placed on the discharge pipe close to the compressor, adjust the superheat of the valve until you have the desired discharge temperature (with LRI I usually set the temp to 55-60C, while with an oil cooler we would be running with 70-75C).
Using a TEAT with the range +35-+65, my experience is that you need a condensing pressure of +19c to start cooling with a discharge pressure of +30.

it's not really liquid injection, but a mist that is introduced towards the end of the compression, I find it more reliable for a steady discharge and oil temperature on a system then one with an oil cooler and a three-way oil mixing valve.

:)

Thanks we use it but rarely, only on special application like a mine site where water is scarce.
It derates compressor capacity & uses more power, but used if necessary.
Big in USA I believe, as you have to do what opposition use, or you don't get the job (no oil cooler or cooling water required etc).

Magoo
28-08-2015, 03:39 AM
Never been a fan of liquid injection oil cooling, goes from great to disaster in a nano second.

Glenn Moore
28-08-2015, 11:28 AM
Liquid injection has always been troublesome due to various reasons. In the UK the Danfoss TEAT liquid injection valve was used on most systems either as a direct injection valve or as a pilot valve with a PMFL or H main valve. Broken capillarys, blocked orifices, diaphragm damage caused by pressure pulses coming from economiser ports etc. plus the TEAT is a very old style valve with a 20 deg c P band which was difficult to set up accurately this caused oil foaming , oil carry over etc
Danfoss were persuaded to manufacture a control called an EKC 319A electronic liquid injection controller which drives the AKV or AKVA pulsed expansion valve via signals from a PT 1000 sensor. These are fitted on to Hallscrew packs and I have retro fitted these to Stal screws and others. The temperature is spot on as you can see via the controller the temperature it is controlling at. The one control replaces all the temperature and capacity variants of the TEAT valves and gives superior control with little or no oil foaming and very little oil carry over plus it has the benefit of a low and high temperature alarm built in making it a safer liquid injection.
Tycho get one and try it you'll never fit a TEAt again

RANGER1
28-08-2015, 11:46 AM
Liquid injection has always been troublesome due to various reasons. In the UK the Danfoss TEAT liquid injection valve was used on most systems either as a direct injection valve or as a pilot valve with a PMFL or H main valve. Broken capillarys, blocked orifices, diaphragm damage caused by pressure pulses coming from economiser ports etc. plus the TEAT is a very old style valve with a 20 deg c P band which was difficult to set up accurately this caused oil foaming , oil carry over etc
Danfoss were persuaded to manufacture a control called an EKC 319A electronic liquid injection controller which drives the AKV or AKVA pulsed expansion valve via signals from a PT 1000 sensor. These are fitted on to Hallscrew packs and I have retro fitted these to Stal screws and others. The temperature is spot on as you can see via the controller the temperature it is controlling at. The one control replaces all the temperature and capacity variants of the TEAT valves and gives superior control with little or no oil foaming and very little oil carry over plus it has the benefit of a low and high temperature alarm built in making it a safer liquid injection.
Tycho get one and try it you'll never fit a TEAt again

never used Akva on it, most
recent would be ICM, as engineers worried about the pulsing.
From your description seems to work ok!
as for TEAT never had any issues, just change orifice every year or two, on ammonia that is.

Tycho
28-08-2015, 04:55 PM
Liquid injection has always been troublesome due to various reasons. In the UK the Danfoss TEAT liquid injection valve was used on most systems either as a direct injection valve or as a pilot valve with a PMFL or H main valve. Broken capillarys, blocked orifices, diaphragm damage caused by pressure pulses coming from economiser ports etc. plus the TEAT is a very old style valve with a 20 deg c P band which was difficult to set up accurately this caused oil foaming , oil carry over etc
Danfoss were persuaded to manufacture a control called an EKC 319A electronic liquid injection controller which drives the AKV or AKVA pulsed expansion valve via signals from a PT 1000 sensor. These are fitted on to Hallscrew packs and I have retro fitted these to Stal screws and others. The temperature is spot on as you can see via the controller the temperature it is controlling at. The one control replaces all the temperature and capacity variants of the TEAT valves and gives superior control with little or no oil foaming and very little oil carry over plus it has the benefit of a low and high temperature alarm built in making it a safer liquid injection.
Tycho get one and try it you'll never fit a TEAt again

Good to know :)

We have made our own adaptive controller (same principle as the EKC319A) that we use on different things.
Funny you mentioned this, because one of the other engineers actually brought this up last week while we were on a job in Denmark.

The biggest issue I have had with the TEAT is the 20c p band you mention, and that the thermo element fails, we have a bypass with a AKVA and a reg valve for emergency use but the bypass isn't large enough to handle the full compressor load.

I'm not worried about pulsing, with a properly sized nozzle and a fast regulator I can easily imagine an AKVA being better and more stable than a TEAT valve.

Tycho
28-08-2015, 04:57 PM
Thanks we use it but rarely, only on special application like a mine site where water is scarce.
It derates compressor capacity & uses more power, but used if necessary.
Big in USA I believe, as you have to do what opposition use, or you don't get the job (no oil cooler or cooling water required etc).

I think the reason we use it is because we have had more problems with the danfoss ORV oil mixing valve than we've had with the TEAT valves.
it also saves cost in not needing an oil cooler or a separate oil cooling pump in applications where the condenser pump can't handle both.

Segei
29-08-2015, 02:04 PM
Thanks we use it but rarely, only on special application like a mine site where water is scarce.
It derates compressor capacity & uses more power, but used if necessary.
Big in USA I believe, as you have to do what opposition use, or you don't get the job (no oil cooler or cooling water required etc).
As far as know in North America majority plants use 2 types of oil cooling. Liquid injection and thermosyphone. Liquid injection is not efficient and can damage compressor, but it is cheap. Usually, it used for small plants with 1-3 compressors. Typically, large plants have thermosyphon cooling. Amot oil mixing valves work good to keep required oil temperature.

RANGER1
29-08-2015, 10:54 PM
As far as know in North America majority plants use 2 types of oil cooling. Liquid injection and thermosyphone. Liquid injection is not efficient and can damage compressor, but it is cheap. Usually, it used for small plants with 1-3 compressors. Typically, large plants have thermosyphon cooling. Amot oil mixing valves work good to keep required oil temperature.

Segei,
Is there any reason water cooled is not used?
Maybe it's price driven as well!

Segei
30-08-2015, 12:30 AM
Segei,
Is there any reason water cooled is not used?
Maybe it's price driven as well!
I didn't see any water oil coolers. Water pipes can rust and oil cooler should be cleaned periodically. Thermosyphon cooling is maintenance free, but it requires initial investments.

Josip
30-08-2015, 06:08 PM
Hi all,

thanks for links and inputs ... very interesting things ...

Few decades ago (speaking about NH3 plants) we used LRI both for oil cooling and for discharge cooling on the first stage of the plant (we had a lot of small problems to adjust it) and in general water cooled oil coolers on the second stage and sometimes also for the first stage ...

later on change complete design philosophy (of course on the new plants) to well designed thermosyphon system with Amot valve and just forget all problems ... as Segei wrote it is higher initial investment, but later on almost maintenance free ... OK sometimes need to drain oil ;) .... in another hand with water cooled oil coolers is normal to have more maintenance and more problems ... due to water pipes corrosion, mud, scale and of course high water price ...

Best regards, Josip :)

RANGER1
30-08-2015, 10:50 PM
Hi all,

thanks for links and inputs ... very interesting things ...

Few decades ago (speaking about NH3 plants) we used LRI both for oil cooling and for discharge cooling on the first stage of the plant (we had a lot of small problems to adjust it) and in general water cooled oil coolers on the second stage and sometimes also for the first stage ...

later on change complete design philosophy (of course on the new plants) to well designed thermosyphon system with Amot valve and just forget all problems ... as Segei wrote it is higher initial investment, but later on almost maintenance free ... OK sometimes need to drain oil ;) .... in another hand with water cooled oil coolers is normal to have more maintenance and more problems ... due to water pipes corrosion, mud, scale and of course high water price ...

Best regards, Josip :)

Just to add, on some older plants that have grown over many years, it can be difficult to install
Thermosyphon, as sort of has to be purpose built to a certain degree.
Water cooled closed circuit systems have been installed, which are basically maintenance free as well.
I think it would be fair to say amount of water used is the same on all these systems, as heat still has to be rejected in evaporative cooling tower.
As for why the S70 Stals use less oil injection compared to similar capacity compressors I don't know.
Some have extra line upstream of pump supplying differential pressure lube to oil injection, but most do not.
Any theories Josip?

Segei
31-08-2015, 01:32 AM
To work properly, a thermosyphon should be installed at certain minimal height above the oil coolers. Some plants don't have this height. I saw one plant that have oil coolers cooled by ammonia similar to thermosyphon but ammonia was moved by the pump not by gravity like thermosyphon.
Some compressors don't have oil pump at all. Oil supplied by the pressure difference.

sterl
03-09-2015, 09:53 PM
My best recall is that relative few people this side of the pond had trouble with the DAE- OROA arrangement for liquid injection but a whole lot of people had trouble with the solenoid that accompanied those pieces....And the machine heated up in a few seconds if the liquid delivery was disrupted, there is no thermal accumulator in the LI arrangements. I have more than once heard the package folks rationalizing big oil inventories in their horizontal separators based on the vagaries of liquid injection....
If you look up a couple of articles by Joe Pillis, you will find some reasoning behind oil sump temperatures exceeding discharge by a specific amount, particularly as it applies to halocarbons.

Tycho
04-09-2015, 06:49 PM
I didn't see any water oil coolers. Water pipes can rust and oil cooler should be cleaned periodically. Thermosyphon cooling is maintenance free, but it requires initial investments.

Thermosyphon, while good, is not quite maintenance free ;)

I've been on quite a few systems where the thermosyphon cooler has become the oil collector on various NH3 systems.

Site engineer is very puzzled because he keeps filling oil on the compressor unit, but gets nothing from the LP oil collector when he tries to drain it.

Worse one I was on, I drained close to 400 liters of oil from the thermosyphon cooler, but again that says something about the efficiency, since it was almost 3/4 full of oil before they started experiencing problems with the oil temperature :)

Tycho
04-09-2015, 07:00 PM
I see many of you mention the Amot valve, we have used this before, but the last 15 years our compressor packages has been delivered with the danfoss ORV valve.

I don't remember servicing many Amot valves, but the danfoss ORV has been a repeat offender on many systems.

The biggest problem with the ORV has been an internal leak past the teflon strip that seals between the valve body and the part of the valve that slides back and forth.

I have noticed that when installing a new teflon strip, there is already a gap of 2-4mm where the ends should meet, and while it's soft and pliable when installing, when we get a problem with high OT and removing the old strip, it's hard as rock and I have to break it into pieces to get it out.


Anyone else been having these problems with the danfoss ORV?

RANGER1
04-09-2015, 08:03 PM
Thermosyphon, while good, is not quite maintenance free ;)

I've been on quite a few systems where the thermosyphon cooler has become the oil collector on various NH3 systems.

Site engineer is very puzzled because he keeps filling oil on the compressor unit, but gets nothing from the LP oil collector when he tries to drain it.

Worse one I was on, I drained close to 400 liters of oil from the thermosyphon cooler, but again that says something about the efficiency, since it was almost 3/4 full of oil before they started experiencing problems with the oil temperature :)

I could honestly say we never get oil in thermosyphon oil coolers on our systems.
Mae have opposition companies that do.
Without studying in detail, I believe the liquid stub in bottom of thermosyphon vessel probably sticks up at least 10mm & vessel is sloped away slightly with liquid line to plant taking any oil with it.
Also good oil separation is also a must, so as not to overload system with oil.
Some, but not all screw packs, we may only top up 3 months to a year, or after changing oil filters.

Segei
05-09-2015, 01:41 AM
Thermosyphon, while good, is not quite maintenance free ;)

I've been on quite a few systems where the thermosyphon cooler has become the oil collector on various NH3 systems.

Site engineer is very puzzled because he keeps filling oil on the compressor unit, but gets nothing from the LP oil collector when he tries to drain it.

Worse one I was on, I drained close to 400 liters of oil from the thermosyphon cooler, but again that says something about the efficiency, since it was almost 3/4 full of oil before they started experiencing problems with the oil temperature :)
i saw many plants which don't drain oil from oil coolers. They are maintenance free. And I know one plant with thermosyphon which keep 150 psig (10 bars) condensing pressure all year around. At lower condensing pressure oil will overheat. This is matter of right design and installation.

Segei
05-09-2015, 01:45 AM
I see many of you mention the Amot valve, we have used this before, but the last 15 years our compressor packages has been delivered with the danfoss ORV valve.

I don't remember servicing many Amot valves, but the danfoss ORV has been a repeat offender on many systems.

The biggest problem with the ORV has been an internal leak past the teflon strip that seals between the valve body and the part of the valve that slides back and forth.

I have noticed that when installing a new teflon strip, there is already a gap of 2-4mm where the ends should meet, and while it's soft and pliable when installing, when we get a problem with high OT and removing the old strip, it's hard as rock and I have to break it into pieces to get it out.


Anyone else been having these problems with the danfoss ORV?
Didn't work with danfoss ORV. Amot are reliable valves. However, sometimes thermo element gradually lose charge and valve was not able to keep required temperature.

piewie
05-09-2015, 05:18 AM
Rule of thumb. For every 1 correctly designed,installed, commissioned & maintained thermo-syphon plant. There are 4 or 5 that do no work. Not only in Europe but in Australia and Africa as well. Good design is imperative.

RANGER1
05-09-2015, 11:10 AM
Rule of thumb. For every 1 correctly designed,installed, commissioned & maintained thermo-syphon plant. There are 4 or 5 that do no work. Not only in Europe but in Australia and Africa as well. Good design is imperative.

Piewie,
What do you mean when you say they don't work.
Mot at all or not reliable?
What if they don't work, someone else repairs!
we have a few that lay up now & then, but 99% ok.
I think Stal were the gurus in the 70's & it progressed from there.