Compressor Floodback

[DeanQuest]

Hello All,

We are using a condensing unit(scroll compressor) and brazed plate heat exchangers(3) to chill a process fluid. The condensing unit runs continuously, even in zero cooling load conditions. We're using a hot gas bypass valve to keep flow going to the scroll compressor.

The temperature at each of the brazed plate heat exchangers is controlled via a thermal expansion valve and evaporator pressure regulator.

At zero cooling load, the accumulator on the condensing unit becomes frosted from top to bottom. Also, the scroll compressor becomes frosted from top to bottom and even a few feet of the compressor discharge line. We're apparently filling the accumulator with liquid refrigerant and it is being drawn into and through the scroll compressor. Thank goodness its a scroll.

What I think may be happening: during zero cooling load, the EPR's close but liquid refrigerant continues to trickle through the TXV because TXV's are not leak tight(or are they?). The HX becomes liquid full and eventually forces the EPR to open enough to relieve pressure in the liquid full HX. The liquid passing the EPR's chills the return lines and accumulator and subcooled liquid eventually fills the accumulator.

Does any of this sound plausible? Anybody have any experience with this type of problem? This is the first refrigeration system I've ever had to troubleshoot. I'm not a professional, just a tech with a need to make it work.

Thanks to all who took the time to read the above,

DeanQuest

[Gary]

What is the compressor inlet superheat?

[Gary]

It is a mistake to assume that frost means liquid. It ain't necessarily so.

For example, let's say that we measure the pressure at the compressor inlet. We refer to our P/T chart and find that this translates to a saturation temperature of 0'F. We measure the temperature of the suction line at the compressor inlet and it reads 20'F. We have 20 - 0 = 20'F superheat, which means there are no liquid droplets entering the compressor... yet the line is cold enough to form frost.

Only by measuring superheat do we find out what we need to know.

[DeanQuest]

Superheat at compressor inlet is near saturation. 21F at about 60 psig.

[Gary]

60psi doesn't tell us anything if we don't know what refrigerant is in the system.

[goshen]

Thank goodness its a scroll

hi there your comp will not last long working with liquid will eventully dilute your oil and trash you bearings!!!
please elaborate on type of refrigerant and pressures, type of head pressure control
and temp band nature of fluid being cooled?
good luck

[DeanQuest]

Hi Evereyone,

Thanks for all the responses. The refrigerant is R404a. Head pressure control is via Sporlan OROA-5-180. Ambient temperature for air cooled condensing unt is 68F. We're trying to cool process fluid to approximately 50F.

[Gary]

Does the system have a liquid line solenoid?

Describe the hot gas bypass.

[DeanQuest]

Gary,

Yes, the system does have a liquid line solenoid valve for pumpdown purposes. But pumpdown is the only time it is used, and our system is setup to run continuously, 8 hours per day.

The hot gas bypass injects the hot gas on the downstream side of one of our thermal expnasion valves. We are using the HGBV to inject refrigerant vapor into one of the heat exchangers so that the condensing unit can run continuously even during low cooling loads. No short cycling. We have the HGBV adjusted to its maximum value of 80 psig.

[mad fridgie]

Do you have a liquid reciever on the plant (I will presume yes at this stage)
Have you purchased a standard condensing unit (normally used for cold rooms etc)
If so you tend to find that the reciever is way to big ( PHE chillers; these systems could "almost" be critical charge) You tend to have a lot of refrigerant which not required. Under low load all this refrigerant will migrate to the low side (as seen). I still like recievers, but i just install a small one. (normally less capacity than the accumulator) this basically means your are unable to flood the compressor. this in it self does not reduce the very low superheat, there fore install a small hot gas by-pass in the suction line, just to raise the temp a few degrees

[Gary]

Gary,

Yes, the system does have a liquid line solenoid valve for pumpdown purposes. But pumpdown is the only time it is used, and our system is setup to run continuously, 8 hours per day.

The hot gas bypass injects the hot gas on the downstream side of one of our thermal expnasion valves. We are using the HGBV to inject refrigerant vapor into one of the heat exchangers so that the condensing unit can run continuously even during low cooling loads. No short cycling. We have the HGBV adjusted to its maximum value of 80 psig.

Is the flooding coming from the HX with the bypass?

[mad fridgie]

I hope that these periods of low load are short, please do not tell me that you are running this for hours with no load. If so you need to install some sort of thermal buffer tank

[Sledge]

what got my attention was the frosting on the compressor discharge!!!

Definite indicator of liquid in the compressor. zero degree subcooling another definite indicator of floodback.

I am thinking that the question of whether the liquid receiver is oversized or not is not a concern, except that a smaller receiver would reduce the amount of refrigerant circulating, hence limiting the opportunity for filling the accumulator and compressor crank with liquid...not enough refrigerant to acheive it. Since you are running with an OROA headmaster control you need a big receiver to handle the refrigerant during the summer, and you need a big charge to maintain winter head pressure. The migration of refrigerant to a colder location is a sympton, not the disease.

A consideration to hide the problem may be to increase the suction accumulator size and heat it...maybe with hot gas...use it like a condensor, but this can be like using pretzel logic.

I would look at stopping refrigerant flow during low load.

I am not certain if the TXV should close off 100%. If it doesnt close off completely, it may be possible to achieve 100% shut off by using a different power head on the valve. Talk to Sporlan or Danfoss for this info.

Alternatively install liquid line solenoids to each evap and shut them off during low load. Obviously this will be involved, requiring temperature monitoring and a control system.

[tudorice]

Have you checked tev is properly functioning and that no-one has fiddled with the needle setting.

[mad fridgie]

what got my attention was the frosting on the compressor discharge!!!

Definite indicator of liquid in the compressor. zero degree subcooling another definite indicator of floodback.

I am thinking that the question of whether the liquid receiver is oversized or not is not a concern, except that a smaller receiver would reduce the amount of refrigerant circulating, hence limiting the opportunity for filling the accumulator and compressor crank with liquid...not enough refrigerant to acheive it. Since you are running with an OROA headmaster control you need a big receiver to handle the refrigerant during the summer, and you need a big charge to maintain winter head pressure. The migration of refrigerant to a colder location is a sympton, not the disease.

A consideration to hide the problem may be to increase the suction accumulator size and heat it...maybe with hot gas...use it like a condensor, but this can be like using pretzel logic.

I would look at stopping refrigerant flow during low load.

I am not certain if the TXV should close off 100%. If it doesnt close off completely, it may be possible to achieve 100% shut off by using a different power head on the valve. Talk to Sporlan or Danfoss for this info.

Alternatively install liquid line solenoids to each evap and shut them off during low load. Obviously this will be involved, requiring temperature monitoring and a control system.
You are right and wrong at the same time, If you do not have the large amount amount of refrigerant in the system then you reduce the likelyhood of floodback, but this is a sympton, handling the low load is the cause, from what is indicated, this is a new plant, more than liklely a fixed price contract.
What we do not know is how and when the low load are occuring and how critical the cooling medium is (short term verses long term averages)
If the short term is not to critical, turn of the plant and use an anti cycle time to stop short cycling, if critical, then you have to introduce false load (which he is trying,) or you introduce a thermal buffer tank, so that the unit, is able to cycle off.
So presuming that there not much in the contractors budget, hows this to be resolved. reduce the refrigerant volume (smaller reciever), I also suspect that the the Hot gas injection is to small.

[DeanQuest]

Mad Fridgie,

The receiver on this standard Copeland condensing unit is at least three times bigger than the accumulator. Can you please elaborate on how the refrigerant can migrate from the receiver to the accumulator during low loads? Thanks.

Also, we already have a hot gas bypass valve in the system and I think it is working.

[mad fridgie]

Mad Fridgie,

The receiver on this standard Copeland condensing unit is at least three times bigger than the accumulator. Can you please elaborate on how the refrigerant can migrate from the receiver to the accumulator during low loads? Thanks.

Also, we already have a hot gas bypass valve in the system and I think it is working.
The reciever is way to big, Plate heat exchangers need very little working refrigerant (compared to for example shell and tube or Air DX evap)
Your TXV will NOT shut down 100% (lucky if it shuts to 25%) Pressure will always travel from high to low, so migration will always occur High liquid pressure to low suction accum. pressure, if you have no loadto boil of the refrigerant, then liquid will flow to the accumulator.
How many evaps do you have. Is there a commom EPR or one for each.
You indicated only one hot gas injection (may be the right size for one evap not the whole system. It could be that only feeding one evap leg, will cause the other evaps to over feed with liquid. you should be able to run the system on the hot gas injection only! (not that I would recomend this, as you can end up with high suction superheat).
What is your load profile (how long is it at very low load) I presume that you do not want to spend fortune?

[mad fridgie]

Sorry my freind, can not read, you have 3 heat exchanges.
Is your reciever horizintal or vertical?

[mad fridgie]

So if reciever is a horizontal lift the end opposite to the liquid outlet. (so it looks tilted) this will give you a quicker liquid seal. (so less refrigerant required)
Install hot gas by-pass on all evaps (or common supply to all if the valve is arge enough to handle the total system).
If you want to do this cheaply use a solenoid and a presure switch.
I do not understand how you can pump down with an EPR unless you have a by-pass, by all means turn of liquid soleniod valve when comp is off to stop migration.

[Sledge]

Hey Mad Friggie, Can you elaborate...you said I am right and wrong at same time, not sure what you mean?

I am confused as to why you are only looking at the size of the evap's as the main indicator for sizing the receiver...it is my understanding that while evap size is a factor the primary issue for receiver sizing on a system with a headmaster setup is storage of refrigerant during warm seasons. The required refrigerant volume in such a system in the winter time can be massive. The receiver must be sized to handle this during the summer.

I understand (and noted it) that reducing refrigerant charge will reduce the floodback because there is not enough refrigerant present, but this is not a realistic option on a system that will operate in winter.

I dont see why the time period that the system will encounter low loads is a factor. Any repair/modification that is designed around set time periods for the low loads, will not be satisfactory in a production environment where demand can be all over the map. It should be possible to make the system function regardless of demand cycle.

I didnt know that a TX valve only closes to 25%. Does that include all TXV set-ups regardless of powerhead charge? If that is the case then liquid line solenoids for each evap is a must.

[lana]

Hi there,

If you have head pressure control valve like in the picture then you have liquid flood back in two cases :
1- Condenser is located above the compressor and there is no trap (number 7 in picture) or a check valve at the inlet of the condenser.
2- Discharge line is running up but there is no trap at the bottom (8).

From the evap side also you can have flood back. If you placed the Hot gas by-pass only in one evap the the others can flood the compressor at no load.

Normal TEV never closes shut.

Other thing which got my attention is this statement "Thank goodness its a scroll"
Whatever the compressor type is, flood back means broken compressor.

Cheers

[Gary]

I think the first thing to do is to find out which evap(s) is flooding.

Check the superheat at the outlet of each evap.

[mad fridgie]

Hey Mad Friggie, Can you elaborate...you said I am right and wrong at same time, not sure what you mean?

I am confused as to why you are only looking at the size of the evap's as the main indicator for sizing the receiver...it is my understanding that while evap size is a factor the primary issue for receiver sizing on a system with a headmaster setup is storage of refrigerant during warm seasons. The required refrigerant volume in such a system in the winter time can be massive. The receiver must be sized to handle this during the summer.

I understand (and noted it) that reducing refrigerant charge will reduce the floodback because there is not enough refrigerant present, but this is not a realistic option on a system that will operate in winter.

I dont see why the time period that the system will encounter low loads is a factor. Any repair/modification that is designed around set time periods for the low loads, will not be satisfactory in a production environment where demand can be all over the map. It should be possible to make the system function regardless of demand cycle.

I didnt know that a TX valve only closes to 25%. Does that include all TXV set-ups regardless of powerhead charge? If that is the case then liquid line solenoids for each evap is a must.
You are right, flood back is a symptom. Saying the size of the reciever has no relevence then you are wrong.
Correct you size the reciever to do the job required, What i said was the reciever was over sized, I did not say no reciever was required, I said i prefer a reciever, What I should of said "I prefer a reciever the of correct size"

Correct a system should be design to meet the load demand cycle, but not at the expense of the machinery (which is the present problem and is this not what we are talking about)
Simply, if you have no load you machine should be off, but practically if the process is critical, then using anti cycle delays on the compressor can cause problems to the process. So requirement for false load or thermal buffer. This is not set around a fixed time period, but about time periods that have little negative effect on the refrigeration machinery and the process.
without knowing the excact details of the whole system and knowing every TXV ever made, I can not say that some where in the world a TXV will not close less than 25%, (thought! Sporlan make a double port valve, so these do close less than 25% of rated load)
So you have closed your liquid solenoid valves, great no liquid flood back, so false load, hot gas injection Ok running like this for hours, what happens to the suction superheat, it getts hotter and hotter, so you then need to introduce suction line liquid injection.
All can be solved, but the best way is to understand the process and your clients requirments,
No point rushing in with a "Sledge hammer"

[DeanQuest]

Folks,

My sincere thanks to all of you for your efforts in trying to help me in this situation. Everyone's comments and advice have been very much appreciated. Truly kind of all of you to try and help someone in a time of need. This is obviously a swell site with capable and caring members. Thanks again to all of you.

I will be sorting out a few things over the next couple of hours. I think with the information provided in this thread I can go forward with a plan of action that will be successful in meeting our customer's needs.

I hope all of you find your kindness is rewarded or rewarding.

DeanQuest

[DeanQuest]

Hello All,

Fortified with the information received from all of you, we did some further component checking. One of the TXV's was rated for 9 tons. We have determined the load on the HX being fed by this TXV is more realistically a 5 ton max load but often is less than 1 ton. This system is a new design and apparently oversizing this TXV is one of our "bugs". The other two TXV's seem to be sized correctly.

What we are doing to resolve the compressor floodback problem:

1. We are replacing the 9 ton TXV with a 6 ton balanced TXV. This should give us better control at the frequent low load conditions.

2. We are redirecting the hot gas bypass to feed directly into the return line.

3. We are installing a desuperheating TXV downstream of the hot gas bypass.

With 2. and 3. above, we will be safely introducing the "buffer load" referred to earlier in this thread, except it will be operating all the time. In this re-arrangement, the hot gas bypass valve will now work directly on the refrigerant passing through all three HX's during low load conditions. The HGBV is adequately sized for this system.

We will be performing the above rework next week. If anyone senses we will be doing something wrong, sure would appreciate knowing.

Thanks again to everyone. I will report back to this thread with our immediate results.

DeanQuest

[DeanQuest]

Hello All,

Just a follow up to let everyone know why we were experiencing compressor floodback.

First of all, we did redirect the hot gas bypass valve flow to the suction line of the condensing unit and added a desuperheating valve as mentioned in my last post. However, the problem did not go away. With all the effort we went through, the refrigerant was still flooding back to the compressor. Back to the drawing board!

Finally, while reviewing pictures of the installed system, I noticed the 1/4" equalization line on one of the TXV's had been installed incorrectly. Instead of being connected to the refrigerant return line downstream of the evaporator, it had been connected to the fluid process line downstream of the evaporator. Hence, the TXV was running full open because the pressure in the process fluid line was much higher than the pressure in the refrigerant return line where it should have been connected. The TXV was operating as if the vapor being discharged from the evaporator was highly superheated.

We made the correction in about 2 hours and the system is now running as originally intended. Great feeling :D! However, I am humbled by the error I made by installing the eq line in the wrong place to begin with. And then not noticing it for months while we try just about everything else to remedy the problem! Just goes to show, again, that discovering root causes for problems must start with the simplest of possibilities first before focusing on the more complex. If I had done a thorough job in following this rule, I would have saved myself a great deal of time and money.

Good luck to you all and Merry Christmas. Thanks again for all the advice and knowledge shared.

DeanQuest

[mad fridgie]

Well done DeanQuest, firstly for finding and sorting the problem and mostly for your honesty.
I wonder how many times this happens, when we are unable to see the wood for the trees.