ammonia flodded air cooler problem

[marsent]

We are in process of commissioning ammonia flooded air cooler for milk cold store where temp to be maintained is 0 deg c . evaporation temp is -10 deg c condensing temp 40 deg c . The air cooler is of refrigeration capacity 35 kw with 4 nos of fans and 5/8" SS coil WITH ALUMINIUM Fins of 12 ft long,16 pipe high and 6 rows deep. There are total 6 circuits of coil each is 12 feet long and 16 pipe high The cooler is connected with 300 dia vertical accumulator liquid connection from accumulator to coil is 20 mm pipe and suction from coil to accumulator is 65 mm pipe. The level controller is set at 80 % of coil height. Now problem we are facing is cold room temp is not droping below 18 deg c. Only bottom 2 to 3 coils are getting frosting on them but accumulator is full of frosting and level contoller is cutting off liquid line solenoid valve. Can anybody help me to trouble shoot this problem.

[Magoo]

Hi Marsent.
Run system with fans stopped and see if coil fills with liquid. Start fans and see if feed rate can keep coil full.

[NH3LVR]

Send us a sketch of the system. It sounds to me as if you are not getting liquid into the coil from the accumulator. A common reason for this is a design issue. If you have a solenoid valve to isolate the coild during defrost in the liquid leg it will not open by itself. It must be a hot gas operated valve. I have seen these valves stick on occasion. If you manually open the valve and it operates you have found the problem.

[Sandro Baptista]

Marsent,

I think our friends point it to important things so you can check out.

About the connecting pipes you have refer it 20 mm and 65 mm. Do you meant 20 mm2 and 65 mm2 internal cross section?
What's the capacity we are playing with?

BR

[RANGER1]

I've seen a few very old units similar, but this practice probably stopped 25-30 yrs ago.

Its not thermosyphon just relatively inactive ammonia sitting in evaporator.

Vessel is usually mounted inside room next to top of evaporator.

Vessel is very small so is liquid feed out of it.

Would expect trouble as its cheap & nasty set up.

[HVACRsaurus]

Does the suction (wet return) vent into the vapor space of the accumulator? Is there a sight glass or level indicator on the accumulator?

[RANGER1]

Oil would'nt help!

[Sandro Baptista]

Oil would'nt help!

I suppose the periodically hot gas defrost should clean out the evaporator from oil.

Marsent, have you establish periodically defrosts? How the oil separators from the compressors behave?

[marsent]

Dear Sirs ,
Thanks for your replies ,
I am enclosing photograph of installation , Coil gets flooded with liquid when fans are off , Liquid connection from accumulator to coil 20 mm ID pipe and vapour connection between coil and accumulator is 65 mm ID pipe , vapour from coil gets discharged in to vapour portion of accumulator , there is no hot gas defrost system with this installation , As per manufacturer of coil ammonia side pressure drop across coil is 0.005 bar and level controller maintains liquid level at about 550 mm bottom liquid in connection of coil.
With best regards

[NH3LVR]

I thought it was a misunderstanding when you said a 20mm pipe for the liquid. The liquid line is WAY too small. It looks to me like this coil was designed for a pump recirculation system.

[Sandro Baptista]

What's the capacity we are talking about it? This point is very important since it related with the mass flow of ammonia circulating. 0,005 bar of pressure it is not very significant however the pressure drop is dependent of the mass flow...what's mass flow are we talking about?

Try to up the suction acummulator about 200 mm and see what happens.

[Sandro Baptista]

I'm sorry, I just see carefully the first post. It is 35 kW at 0ºC air temperature with DT=10K.

ID=20 mm just gives about 0,2 m/s with recirculation factor=2 and based on it's small lenght I think it would creat almost 0 bar of pressure drop.

However it is possible you have to increase the level difference between evaporator and acummulator so you can guarantee the mass flow needed.

[Josip]

Hi, Marsent

We are in process of commissioning ammonia flooded air cooler for milk cold store where temp to be maintained is 0 deg c . evaporation temp is -10 deg c condensing temp 40 deg c . The air cooler is of refrigeration capacity 35 kw with 4 nos of fans and 5/8" SS coil WITH ALUMINIUM Fins of 12 ft long,16 pipe high and 6 rows deep. There are total 6 circuits of coil each is 12 feet long and 16 pipe high The cooler is connected with 300 dia vertical accumulator liquid connection from accumulator to coil is 20 mm pipe and suction from coil to accumulator is 65 mm pipe. The level controller is set at 80 % of coil height. Now problem we are facing is cold room temp is not droping below 18 deg c. Only bottom 2 to 3 coils are getting frosting on them but accumulator is full of frosting and level contoller is cutting off liquid line solenoid valve. Can anybody help me to trouble shoot this problem.

Can you explain how you get a liquid ammonia into suction vessel .... is there another evaporator, what is capacity of compressor you using and how is that compressor connected to this evaporator? Condenser type, receiver if is installed, have you ammonia pumps installed?

Best regards, Josip

[marsent]

Dear Sirs ,
Thanks for your replies ,
This installation is replacement of old air cooling units which were running since 20 years . The accumulator gets liquid feeding through 15 mm NB liquid line through solenoid valve ( which is controlled by electronic level controller) and hand expansion valve arrangement and main liquid header from HP receiver is of 32mm NB size . Dry suction from accumulator is connected to main suction header of 100 mm NB by 40 mm NB pipe . And there are 4 AIR COOLING UNITS connected parallel to this suction and liquid header, and all units are facing similar problem.
With best regards

[Magoo]

Hi Marsent.
Is your system a rapid liquid regeneration type system. Regenerating liquid from the suction accumulator.

[RANGER1]

Marsent,
Is there a possibility that liquid header on bottom of unit has orifices in each pipe?

Can you ask manufacturer of air coolers?

As NH3LVR mensions it looks like a pump recirculation type unit.

[NH3LVR]

As I said my bet is that it is a pump recirculation unit. I would most likely have orfices in it then. Is this manufactured in your country? Were they new or used units. This type of flooded system is very reliable and simple, but I have never seen one designed for use with an accumulator with such a small liquid inlet.

[marsent]

Dear Sirs ,
Thanks for your replies,
It is not rapid liquid regeneration system , Its simple gravity flooded system , Manufacturer of coil claims there are no orifices in coil. totally there are 6 circuits of 5/8" od SS tube and length of each circuit is approx 192 feet ( 56.7 meters ) and all are fed through single 3/4 " liquid pipe at bottom.
With best regards

[Sandro Baptista]

Marsent,

Are you sure the head pressure loss 0,005 bar measured for the the operating conditions Tair=0ºC and Tevap=-10ºC?

As I said before I would increase the increase the high of the suction accumulator to guarantee a potential force of R717 circulation. It seems you have superheating for almost the evaporating circuit.

[RANGER1]

marsent,
If you turn off fans, out of interest how long does it take before other return bends frost up?

Can you & have you drained oil from bottom of vertical accumulator as it has to be eliminated.

Can you measure air on/off air cooler as well as temp readings on return bends to see what variances there are as it goes up coil block.

[Sandro Baptista]

Don't believe the problem be the oil. the pipe of the level controller is all frost and also the inlet header of the evaporator.

[RANGER1]

Don't believe the problem be the oil. the pipe of the level controller is all frost and also the inlet header of the evaporator.

Sandro,
Me eiither but I'm running out of logical ideas.

[Magoo]

90 + metres per circuit plus elbows = high pressure drop. Is the liquid HP supply or pump circulation.
Your picture is not clear, but appears the liquid feed goes through/ into the drop leg on the accumulator. Something is not right, if level control activates and accumulator is frosting. Suggest there is a scavenge recirc orifice, and with high pressure loss in coil circuits the liquid is bypassing the evaporator.

[marsent]

Dear Sirs ,
Thanks for your replies , I have forwarded your valuable suggestions / observations to manufacturer of coil awaiting their reply. Mean time for my academic interest can somebody tell if there is any thumb rule about maximum permissible length of pipe in each circuit of air cooling unit coil for gravity flooded and pump circulation application.
With best regards

[Magoo]

There are no rules of thumb, gravity feed coils need surge drum at least 2 metres above suction header of coil. The spec density of NH3 is low,average 0.8 kg. Pump circulation is totally different coil configuration. HP liquid is another configeration. Rapid liquid regeneration maximizes coil performance, another coil configeration. Add counter flow versus airflow
So rule of thumb should be banned from your thoughts.
By chance have you connected the liquid feed to the oil drain connection of accumulator, dumb question that has to be asked.

[marsent]

Dear Sir ,
Thanks for your reply liquid feed is not connected to oil drain , oil drain valve is at bottom most point of drop leg of accumulator and liquid feed is connected through shell of accumulator horizontally and then turned downwards up start of drop leg of accumulator. What I want to ask is whether is it normal to use 60 meter long single coil or it is to be divided in two or three circuits for gravity flooded arrangement ?
With best regards

[Josip]

Hi, marsent

Dear Sir ,
Thanks for your reply liquid feed is not connected to oil drain , oil drain valve is at bottom most point of drop leg of accumulator and liquid feed is connected through shell of accumulator horizontally and then turned downwards up start of drop leg of accumulator. What I want to ask is whether is it normal to use 60 meter long single coil or it is to be divided in two or three circuits for gravity flooded arrangement ?
With best regards

It is OK but there are some other rules to follow too ....

Let's start from beginning ...

you replaced old evaporators (suppose they were working fine) with new, but with different coil construction, am I right? Now you have a problem with feeding of evaporators with liquid ammonia ...

I cannot imagine complete system, maybe you can make a scheme of piping, scan it and upload as pdf file ...

I have a feeling that you did something (not like must be for gravity feeding) with the headers and with the pipes ....

to run the plant normally, maybe, the only way out is to install pump for feeding of evaporators ... BUT, all of this is guessing ... we are engineers and we need accurate drawings to be able to help .... of course I can be wrong too, but without complete info it is not possible to give diagnose and prescription for cure ...

Best regards, Josip

[Magoo]

As per Josip,
we need a sketch to look at.
Seems that the feed is regenerating liquid from accumulator, the pressure drop in each circuit is the problem. 90 metres plus return bends is calculated in pa per metre, add the liquid flashing off creates another pressure internally. The accumulator needs to be metres above suction header to overcome losses.
Again, is the liquid pump feed or high pressure sub cooled liquid.

[mad fridgie]

I would say that your vapour return is below the vessel liquid level. Basically you are getting a "vapour lock".
Drop the float level.

[marsent]

Dear Sirs ,
Thanks for your replies ,
There is no vapour lock as float level is about 150 mm below vapour return pipe from coil to accumulator we also tried controlling liquid level manually at various levels but the results were more or less same. I am attaching rough hand made sketch of system it may be helpful ( i am sorry I can make only hand made sketch as I am traveling now ). I am also waiting for manufacturers reply.
With best regards

[RANGER1]

marsent,
Suspect you had at least 3/4" tubes on your old air coolers which may make all the difference
with pressure drops for this style of design especially when room is hot.

Have you kept the room running to see if room temps & level in units improves?

[Magoo]

Feed the lower header direct and bypass drop leg, or get supplier to reconfigure the circuiting in coil to twin pipe feed per circuit, which will effectively halve the existing circuit PD.

[mad fridgie]

Doing a few very rough numbers, based upon 5.5Kw per circuit (5/8" OD) you have a peak pressure drop of 25Kpa, you do not have enough head to achieve this.
I would ask the manufacture what is the design pressure drop. This will give you your required head for the flooded vessel.

[Magoo]

Hi MF.
The real issue is the coil PD, added is the design liquid injection into the drop leg of accummulator [ without an inductor orifice] so no flow through coil. With the SD of NH3 at .8 ish equals no flow. Fortunately the liquid feed is HP.By twin feeding each circuit would halve the PD through coil. The primary goal is to fill the coil with liquid, which will not happen currently. The aggitated flow rates for NH3 is a complex calculation. hence the suggestion to twin feed each circuit, that will require a major rebuild of each coil, from sketch there is no accumulator before compressors, so the sucton accums.., at each evap.. is critical.
Big bore tubes and multiplex circuits are required, add inductors. Suggest the whole lot gets riped out and start again.

[anand]

Hi Marsent
As per pic and drawing, this seams to be nature circulation / gravity feed system. We have faced similar problems earlier.
Try to lift up the accumulator.
Change the liquid entry to float chamber to accumulator bottom instead of stand pipe, make sure that return connection from coil is connected to accumulator above float level. Many times if liquid entry to float chamber is near to the evaporator liquid entry, liquid short circuit happens through float chamber.
If this all falis, contact the manufacturer of coil toredisgn / change the coil

[mad fridgie]

Hi MF.
The real issue is the coil PD, added is the design liquid injection into the drop leg of accummulator [ without an inductor orifice] so no flow through coil. With the SD of NH3 at .8 ish equals no flow. Fortunately the liquid feed is HP.By twin feeding each circuit would halve the PD through coil. The primary goal is to fill the coil with liquid, which will not happen currently. The aggitated flow rates for NH3 is a complex calculation. hence the suggestion to twin feed each circuit, that will require a major rebuild of each coil, from sketch there is no accumulator before compressors, so the sucton accums.., at each evap.. is critical.
Big bore tubes and multiplex circuits are required, add inductors. Suggest the whole lot gets riped out and start again.

If flow and load were equal to each circuit then total coil PD would be around 25kpa.
I do not believe your liquid feed is into the drop leg (straight into the veseel) very hard to see at the back.
because the refrigerant is at saturation plus the liquid head (sub cooling) with load the coil will never be full of liquid, by volume must be full(ish) of vapour. Hence the pressure drop. (which is more complicated than my suggestion indicates). The circuits are to long for the liquid head available.
For example using the same calc method if the original coil was 1" steel same length and configuration PD was around 4Kpa

[Josip]

Hi, marsent

thanks for photo and for schematic diagram ....

We are in process of commissioning ammonia flooded air cooler for milk cold store where temp to be maintained is 0 deg c . evaporation temp is -10 deg c condensing temp 40 deg c . The air cooler is of refrigeration capacity 35 kw with 4 nos of fans and 5/8" SS coil WITH ALUMINIUM Fins of 12 ft long,16 pipe high and 6 rows deep. There are total 6 circuits of coil each is 12 feet long and 16 pipe high The cooler is connected with 300 dia vertical accumulator liquid connection from accumulator to coil is 20 mm pipe and suction from coil to accumulator is 65 mm pipe. The level controller is set at 80 % of coil height. Now problem we are facing is cold room temp is not droping below 18 deg c. Only bottom 2 to 3 coils are getting frosting on them but accumulator is full of frosting and level contoller is cutting off liquid line solenoid valve. Can anybody help me to trouble shoot this problem.

.... your coil is SS with pipes of 5/8" thus internal diameter is about 18mm what is giving cross section around 250mm2 (square) ... we have 6 pipes what is all together about 1500mm2 (square) ... by gravity you are trying to feed those pipes through pipe/header with ID of 20mm what is giving cross section of 314mm2 .... to make equal cross section of 6 coil pipes you need feeding pipe of 2" (2330mm2) for gravity feeding .... by this feeding pipe/header that is not possible by gravity .... with pump or TXV is another story ....

Dear Sirs ,
Thanks for your replies ,
I am enclosing photograph of installation , Coil gets flooded with liquid when fans are off , Liquid connection from accumulator to coil 20 mm ID pipe and vapour connection between coil and accumulator is 65 mm ID pipe , vapour from coil gets discharged in to vapour portion of accumulator , there is no hot gas defrost system with this installation , As per manufacturer of coil ammonia side pressure drop across coil is 0.005 bar and level controller maintains liquid level at about 550 mm bottom liquid in connection of coil.
With best regards

.... from your photo it is visible (not exact level) that you can rise level within accumulator vessel up to level switch, but you cannot discharge that liquid into coil through too small pipe/header with ID of 20mm ... also is visible that your level switch is too much bellow the last 16th row of pipes .... only top last one or two rows can be dry otherwise you are not utilize complete evaporator ....

what to do .... not easy to suggest, but you must try something because this is not working .... maybe to try with one evaporator ....

.... change feeding pipe to bigger of appropriate size even 2 1/2" and maybe rise as much as possible accumulator vessel i.e. maximum level to be around 14th row of pipes within evaporator coil ....


Best regards, Josip

[Magoo]

Hi MF,
earlier on Marsent said the liquid feed entered the accumulator and then bend down to inject into the drop leg of accumulator, the drop leg then feeds the bottom header. Being HP liquid after a HAX vav it would be better to feed the bottom header direct, and give half a chance to flood evap coils. I would be inclined to lower the float switch level percentage as well for safety margin and stop flood back. With bottom feed gravity head, the surge drum/ accumulator has to be two metres plus above the suction header, as well twin pipe feed. [ two bottom rows at once ]