ammonia compressor efficiency to calculate mass flow 1

[themroc]

I need to evaluate the performance of a heat exchanger (evaporator)

The exchanger evaporates ammonia
In order to do the calculation I need to know the ammonia mass flow. Which is not meassured directly.

The only information I have got is about the compressor which runs after the evaporator and has the following conditions meassured:

the ammonia vapour is compressed with a compressor from 82mbar (abs)(Saturation pressure) to 13.1 bar (abs)Discharge temperature (84.9°C)(superheated)

The Motor Ampere is meassured with 257 Ampere
The Voltage is unknown, I pressume it is 380V

My question:
Can I calculate the mass flow by
m * (enthalpy outlet - enthalpy inlet) = P compressor = 257 * 380. ???

I pressume I need to take into account the efficency of the compressor. I do not have any information about the type of compressor, Can anyone advice?
What kond of efficiency can one use??

 

[25362]

If the client arbitrarily changed the compressor's conditions from a suction pressure of 1.1 down to 0.8 bar, with the same discharge pressure 13.1 bar, as fixed by the condenser, while the suction temperature dropped from 248 K to 240 K, the ratio HP/(kg/r) is supposed to increase according to the equation I mentioned above.

Namely, for the same BHP, less kg/h ammonia vapor would circulate at the lower pressure, the small increase in latent heat would probably not compensate for the diminished flow rate.

One assumes there are no inward leaks of air when working under vacuum.

 

[dcasto]

26362, The temperature was given at the inlet of the compressor as 4 degrees of superheat and the temperature upstream of the reducing valve is the saturation temperature.

As for the temperature rise and specific volume change going through the compressor, that doesn't matter only the specific volume at the suction dictates the compressor capacity.

The enthaphy entering the cooler is soley dependant on the condensor outlet conditions since the enthalphy across the reducing valve is zero.

I looked at the PH diagram and found that at 250 psia and 100 F, ammonia hac a heat capacity of about .8 BTU/lb-DegF


themroc,
The change from 1.1 to .8 bars will tranlste to a 30% decrease in mass througput and a 1 % lower duty in the chiller

 

[25362]

To dcasto, I do not disagree with your points, except for the isenthalpic expansion meaning that the change in enthalpy is ~zero, not that the enthalpy is zero.

Please note that themroc's question was what happens with the ratio HP/(kg/h) of the compressor when the suction conditions drop from 1.1 bar and 248 K down to 0.8 bar and 240 K, with a constant discharge pressure of 13.1 bar as fixed by the condenser.

I have a small disagreement with Montemayor concerning the presence of droplets in a superheated vapor. Not about the theory or the definition of superheat, but with the facts.

This is a short-lived (let's say, transient) non-equilibrium situation that may happen in refrigerants when the superheated vapor flows in parallel with a droplet-containing saturated stream, and in particular when the superheat [Δ]T (as driving force) is small. Besides, the density of the droplets being almost 800 times greater than that of the vapor, they may tend to collect and coalesce.

In gas quenching, it has been noted that when the flows of mass and heat are opposite, as for gas cooling with humidification, the overall heat transfer coefficient would be 0.3 to 0.8 times what we'd expect from heat transfer alone.

On the other hand, when both, mass and heat, flow in the same direction, as for gas cooling and dehumidification, the OHTC may be 1.5 to 3 times higher than what could be expected from heat transfer alone.

Therefore, if the residence time is short, there is a chance that while bulk temperatures indicate a degree of superheat there still may be drops of liquid refrigerant reaching the compressor.

Comments are appreciated.

 

[dcasto]

25362, Sorry if I wasn't clear, but yeah, the delta enthalpy is 0.

In more detail (give us all a break here 'cause we do not have every P&ID, manufacturers data book, etc..) The HP/(kg/hr) will increase by 10%, but the compressors ability to move gas will decrease by 30%. Net effect is a 30% less than design in kg/hr to the HX available heat transfer, less some amount for a change in the heat transfer coefficient for being at a reduced rate (less than 5% here). There will be a corresponding 10% increase per kg/hr for a net of the compressor running at 77% of its original design amps.

As for not getting all the liquid vaporized, this can happen a flooded horizontal super heat controlled exchanger. I'm in the process now of getting rid of one and replacing it with a BKU type, not because I'm convinced it is allowing droplets past, but because it can and if that happens not only do I loose capacity, I may loose the compressor.

 

[25362]

To dcasto, as I see it, when estimating the mass flow rate you considered just the vapor densities while keeping the compressor to be working with the same volumetric efficiency in both cases.
But what if it were partly unloaded at 1.1 bar and fully loaded at 0.8 bar ?

Regarding your own plans, installing a KO drum on top of the flooded exchanger with suitable droplet-catching devices may be cheaper than replacing it altogether. Have you considered that ?

As an aside, an oil-flooded screw type of compressor could handle some liquid droplets w/o damage.

 

[Montemayor]

25362:

I seriously doubt there are any droplets in a flow stream that has undergone superheating and overcome a 6.7 psi pressure drop. The flow is obviously turbulent and well-mixed. The primary reason I doubt the prescence of any droplets is that it is mere hearsay up to now.

I've undergone this situation where liquid entrainment (without a resultant downstream liquid level) is suspected but only in theory many times in the past. The end result is that no one has accepted my challenge to catch just a few of these micron-sized "droplets" and saved them in a paper bag for my inspection. In fact, upon further inquiry, everyone admits they have never seen a refrigerant droplet in a refrigeration compressor suction line existing at -25 ^oF. I say this not to reinforce my stubborness and hard-headedness but just to point out that all of the reports of droplets existing in these lines are theoretical - at least those that can't produce a subsequent accumulation of liquid refrigerant.

Please don't interpret my comments as contradictory to taking - and employing - sound engineering judgment and good vapor disengagement practices when feeding a saturated ("wet") refrigerant to a compressor. On the contrary. I believe you will find all the systems I've designed and built/installed to be very conservative in that specific section of a refrigeration cycle. I merely wanted to point out that while the entrainment of droplets may be occurring, it is quite difficult to sustain a saturated liquid drop in a superheated stream that is subject to further heat-up as it approaches the suction valves of a compressor. And it is quite impossible to prove the prescense or size of the drops within the system under normal conditions - at least in my opinion. Basically, I believe themroc's client is trying to confuse the issue and point to a defective evaporator design without offering any proof. I still maintain that themroc's way to prove otherwise is to do a heat and material balance around the evaporator.

Most of the major mechanical refrigeration cycles I've operated in the past have had a small hot gas recycle back to the suction drum as a capacity control + superheating effect to ensure that any liquid accumulating in the lines leading to the compressor suction is "swept" with superheated gas to evaporate it.

You have done an excellent analysis of this refrigeration system once we were given the factual data.

 

[25362]

Montemayor, my thanks to you. I fully agree in that we are indeed speaking of seconds or fractions of a second for the droplets to vaporize.

To anyone interested in reading about hot gas quenching with liquid sprays with examples, I suggest the book by
Donald R. Woods: Process Design and Engineering Practice Prentice Hall, ISBN 0-13-805755-9

 

[themroc]

Thanks all for your comments.
I see a little bit clearer. In my opinion the droplet issue is also very suspect, according to our client the effect of droplets is clearly noticable due to noise in the compressor. Since I have got no experience at all I belived him in the beginnings. But your comments seams to indicate that this is quiet unlikely. Due to the pressure and temperature measurements I also thought the superheat is sufficient in order to evaporate the remaining droplets.
I will try to find out the actual piping length between evaporator and compressor inlet. This will give an indication for the residence time in the piping.
In this context:
Does anyone have an experience whether the noise level is an indication of droplets pressent in the suction stream?

Annother commend / question reading through all the answers:
The client claims the evaporator does not peform well. I can base my balance only on the shell side single phase heating fluid because here I have the mass flow and temperature difference. Taking this into account He is right and the duty is much lower than expected.
THe problem is that he operates the evaporator at much lower pressure than in the design calculations.
AS I mentioned about 0.82 bar instead of 1.3 bar.
Obviously he does it in order to achieve a higher driving temperature difference in the evaporator.
But I think and I understand that is also your opinion that the penalty for the compressor in succing at a much lower pressure is worse than the gain in driving temperature.
In addition simulations indicate that at the lower ammonia pressure the heat transfer coefficient in the tubes of the evaporator are also decreasing.

How is actually exactly the pressure level of the cycle maintained.
My understanding is that the compressor determines the succion pressor that means the evaporator exit pressure. So in case the evaporator produces more ammonia vapour mass flow than the compressor can handle at this pressure the pressure will rise automatically, making it easier for the compressor to handle the flow Is this correct?
Would that also mean in case the evaporator produces less vapour mass flow that the succion pressor is reduced?

On the other hand what determines the evaporator inlet pressure? Is this the exit pressure plus pressure drop in the evaporator. Or is the inlet pressure of the evaporator a set value determined by the throtteling valve and all the other pressures (evaporator outlet, compressor inlet) are established from this point?

Sorry for the additional Questions?
but thanks in advance

 

[Montemayor]

themroc:

From what I see, the issue isn't dead; it has changed complexion or the scope has amplified. Because of the additional information and explanations, we can now contribute more positive help. But we need concrete data on:

1. What size, type, model, etc. of ammonia compressor(s) are being used?
2. Is this a 1-stage or 2-stage cycle?
3. What's your capacity control devices on the compressor?
4. How is the evaporator controlled? Is it on level control? i.e., is it flooded with refrigerant? It probably is not since I find it hard to believe that a horizontal BEM type with the refrigerant on the tube side would be. And this is the difficult part: for proper industrial control, you should really have a flooded BKU type of evaporator with the refrigerant on the shell side - just as dcasto implies. I believe dcasto is implying that you have a "DX" type of evaporator and if that is true, then all bets are off on trying to have positive, industrial type of control on the quality of your suction vapor going into the compressor. With a varying evaporator load and other conditions I don't see how you could successfully design a DX unit for these condtions. That's why industrial-grade evaporators are almost always of the flooded type. These you can instrument and control with reliability and dependability. A DX unit has a scope of design that mandates a steady, predictable evaporator load with ZERO load excursions.

We'll wait for your response with the data to form postiive comments and suggestions.

 

[themroc]

Thanks Montemayor,
I will try to find out.

But do you have some experience with droplets in the compressor. Do the produce a loud noise?

 

[Montemayor]

Unfortunately I have had too many experiences with liquid particles entering reciprocating compressors. In all these cases I have been called to identify the situation and remedy it. Liquid particles present a dangerous hazard to a reciprocating compressor when they are in the form of a liquid slug large enough to cancel all the clearances normally available in the cylinder of a recip. The results can be from mild physical shock to a devastating rupture of the cylinder head cover plates. It all depends on the size of the invasive liquid slug.

One such recip incident that occurred in a foreign field installation involves a case where wet natural gas was allowed to accumulate liquid in the interstage suction separators without a sight level gage to visually detect the liquid level and to calibrate a defective dP cell used to control that liquid level. The result was that liquid filled the separator(s) and entered the cylinders. Fortunately the suction valve design was so weak physically that the suction valves were chattered and this saved the compressor cylinder and other pressure components. In other cases, I have had to extract cast iron pistons that had pieces of valve components embedded into the face of the piston. A positive displacement machine is designed to positively displace a given volume - and it will try to do exactly that with every bit of horsepower made available to it.

You can't generalize, as you are doing, just stating that "droplets" are getting through to the cylinder. You don't know that for sure and you can't prove it until damage results. If you hear a loud noise it may be that liquid is getting into the cylinder, but that is just one source of trouble. It could be other factors - depending on the type of compressor, the suction conditions, the mechanical state of the machine, etc., etc. Up to now, we still have not been told WHAT TYPE of compressor is being used. If it is a flooded screw, it could take a lot more so-called "droplets" than a recip. Therefore, it is only guess work without real basic data.

 

[25362]

I figure the possible entrance of droplets and the resulting devastating consequences, may have been one reason for designing the cylinders of reciprocating compressors horizontal, with the inlet and outlet piping on a vertical line, that is to allow quick draining.