Interesting Problem

[NH3LVR]

Compressor is a NH3 Frick RWBII 76H, Variable VI, nineteen years old. Total hours are unclear, but most likely 75,000 Plus.
150 HP, Liquid Injection Oil Cooling. Economizer is feed from a TX Valve (currently off).
The original installation has been modified to work on Blast Freezers in nearby Building. It originally was connected to a Spiral Freezer. The Spiral Freeze is no longer in use. The Suction was piped to the nearby building and the HP Receiver is piped to an LPR in the next building. System seems to work OK from the standpoint of two systems hooked together. There is a total of 500HP in the next building connected to the same suction line which runs @ 14” HG or so
I installed a New Frick Panel five years ago on this machine. The old one was becoming erratic. The Install went smoothly. At the time I recall a bit of trouble doing the Slide Valve Calibration, but nothing of major importance. One of my co-workers work on this machine years before that and seems to recall a great deal of wear in the slide valve.
The Customer reported a lack of ability to get any work from the machine.
In the past year and one half the Maintenance people have replaced the Suction Check Valve (it stuck closed) , as well as both the Capacity and Slide Stop Solenoid Valves. In addition the Program Chips have been replaced as well as the Liquid Injection Solenoid Valve that diverts liquid to one of two ports depending on the VI. The customer is competent and has a Serviceperson of their own that travels to their different plants.
The operator demonstrated that at 100% Slide Valve and 5.0 VI the Amperage Draw was 50%. As he dropped the VI in Manual Mode the Amp draw increased, as was to be expected.
Starting with basics I calibrated the Slide Stop and Slide Valve. SV Pot rotation (In 5.0 170 Degrees, and 2.2 300 Degrees). VI Pot travel was normal. I also checked the CT Factor and Measured the Amp Draw to confirm .
Starting up I noticed these conditions.
(100% SV, 5.0 VI) 3.0 LBS Suction-150LB Discharge-Motor Current-50%-Suction Temp +70F, Suction Check banging
Changing the VI to 4.4 yielded this.
(100% SV, 4.4 VI) 6”HG Suction-150LB Discharge-Motor Current-70%-Suction Temp -5F, Suction Check clicking
Changing the VI to 2.2 yielded this.
(100% SV, 2.2 VI) 12.5HG Suction-150LB Discharge-Motor Current-95%-Suction Temp -40F
It was apparent that the lower I took the VI the more gas was being pumped. The lack of sound from the Suction Check and drop in Suction Temp confirmed that.
Calibration is not an issue. I worked the SV and Vi Solenoid Valves manually and got the same results. Everything moves in the proper direction at the proper time.
I contacted Frick. They are a very competent bunch and very helpful. (I commend them.)
However they were unable to suggest a fault (Internal or External) that could cause this.
According to their figures, at 2.2 VI it should be drawing 200 HP.
My only option at this point is an internal inspection.
Has anyone else every seen anything like this?

[Andy]

Hi NH3LVR

In a word NO


Possibly a seal ring away in the VI Slide, but that's a guess

I'd have the slide ripped down and the bearing tolerance checked (end float)

Kind Regards Andy

[US Iceman]

Here is some data from the Frick program:

SST = Saturated Suction Temp.
SCT = Saturated Condensing Temp.

-40F SST & 84F SCT w/ Vi @ 2.2
Fully loaded (SV @ 100%)
Capacity = 33.5 Tons
Power = 194.9 BHP (pretty close to the Frick numbers)

-40F SST & 84F SCT w/ Vi @ 2.8
Fully loaded (SV @ 100%)
Capacity = 33.5 Tons
Power = 151.4 BHP

-40F SST & 84F SCT w/ Vi @ 3.0
Fully loaded (SV @ 100%)
Capacity = 33.5 Tons
Power = 142.9 BHP

-40F SST & 84F SCT w/ Vi @ 5.0
Fully loaded (SV @ 100%)
Capacity = 33.5 Tons
Power = 108.0 BHP

As you can see from this, the power input starts to decrease quickly as the Vi is adjusted to where it needs to be.

The low Vi at those operating conditions is causing the compressor to over-compress the gas, that's why your motor amps went up at Vi = 2.2.

The Vi does not help the compressor to pump more gas, it controls the pressure of the discharge gas within the compressor housing.

The adjustments you made were going in the wrong direction. You need a low Vi (2.2 or so) to start up. As the suction pressure decreases, the Vi needs to increase (5.0).

Another way to look at it is:

• High compression ratio = high Vi
• Low compression ratio = low Vi

The drop in the suction temperature is just due to the gas flow cooling off the sensor. It looks like the machine was pulling down while you were making the Vi adjustments. Is that right?

My first impression of the suction check valves rattling or clicking is they do not have much gas flow through them to hold the check open.

With the Vi set at 5.0 when the compressor is operating at approximately 9 in. hg and 150 psig discharge, the power should be at about 100 to 110 BHP at full load.

If the compressor is in good condition, the amp draw should be around 125 amps (460-3-60 power supply). Give or take a bit.

It the amps are a lot lower, I suspect the compressor is just about junk. The rotor tips are probably smooth and leaking a lot of gas, so the compressor is not doing much work.

You have a single stage compressor that has been running at a high compression ratio with liquid injection oil cooling. I would say 75,000 hours has been pretty good.

I suggest a rebuilt compressor...

[NH3LVR]

That is kind of the point. The customer had discovered that the machine only worked when he lowered the VI. As you note the only thing that should change when the VI lowers is that the Horsepower requirement goes up.
There is really no pulldown as the machine is on a common header with another Engine Room. They maintain a low suction at all times.
The machine will not pump at all at 5.0 VI.
In addition the Horsepower requirement is not being seen at 2.2. Only about 75% of the 200HP we expect.
I agree that the compressor is almost certainly worn out. The mystery is why lowering the VI improves the performance. The fine people at Frick have never seen this before.
If the Customer had not noted that the compressor worked at the lower VI I might well have missed it.

[US Iceman]

The mystery is why lowering the VI improves the performance.

Well this is certainly a puzzling event I have to give you that.

This is only a guess on my part, but I think is pretty close to what you are seeing...

When the compressor is running at a low Vi, but with a high compression ratio the discharge pressure in the discharge chamber (the port area) will be a lot higher than it needs to be.

As we know the rotors seal with the rotor speed, the rotor tips, and the oil providing a liquid seal for the rotor tips. In effect this is a labyrinth seal.

If high pressure is present on the discharge port area, there is more potential to "push" gas back across the rotors. In other words, the gas is bypassing the rotors, or more blow back is occurring.

This is the same as a reduction in volumetric efficiency. Essentially, less useful volume flow.

My guess also continues to think the increased blow back would be similar to a recip compressor with a broken discharge valve. The discharge temperature would increase after the maximum capacity of the liquid injection valve is exceeded. Now, this is where it gets a little more eccentric.

The discharge temperature may not in fact be increasing, since my guess is the compressor is not pumping very much gas. Less mass flow means you would have less oil cooling load.

More than likely, the liquid injection valve has sufficient capacity to provide enough cooling to compensate for the lower mass flow and high energy input (motor amps/HP).

If the rotor tips are leaking bad enough, this could I think be similar to broken suction valves on a recip compressor.

Some gas is probably being pumped, since you said the suction temperature comes down. However, I don't think very much gas is being pumped.

I believe what the customer is noticing as "pumping" with the higher current draw and low Vi, is in effect the over compression that is occurring.

You would only see the 200 HP value, IF, the compressor was in reasonably good condition. The lack of pumping capacity would lower the estimated and anticipated value of 200 HP.

The machine will not pump at all at 5.0 VI

Are you thinking this because of the low amp draw at Vi=5.0? That might also explain the rattling suction check valve. Basically, no flow or very little flow.

Even if a brand new compressor was fully unloaded, it still pumps about 15-20% of it's rated volume. If the compressor is fully loaded in the condition I think it is in, the amount of volume flow would indeed be very small.

Once you loose the sealing on the rotor tips, the volume flow is absolutely terrible. You see the same thing when the rotor speed slows down below about 50% of full speed.

Is that helpful? It's about the best answer I can come up with.

[NH3LVR]

I think that is as good an answer as I am likely to find.
Also I think it is most likely the correct one.
I am however at the limits of my knowledge on screw compressor design.
With the VI set too low don’t we have in fact under compression?
I did conclude the rattling check valve and high suction at the compressor (compared to the Suction line) was due to the lack of gas flow.
You are probably correct about the low gas flow through the machine even at low VI. I have no good way to quantify that without shutting down part of the other Engine Room, which the Customer would be reluctant to do as an experiment. Or I could startup the Old Spiral Freezer, but that would entail a long pump out to put it back in mothballs.
The plant will be in continuous operation until October, even then it is a five hour roundtrip to test anything.
Fortunately this is a Frick which measures the Suction Pressure at a port on the Compressor, rather than a FES which uses a port installed upstream of the Suction Check Valve. That might have made things more difficult to observe.
Unfortunately I may never see the inside of this machine. We will most likely get a exchange from Frick.

I have been told before by knowledgeable people (As well as the manuals) that a screw produces about 15% of its rated flow when unloaded. I wonder what the effect of liquid injection cooling is on that.
Earlier this year I started up a used 200HP FES Booster in parallel with a 250HP Frick. Everything was going well as I ran it unloaded, checking Oil Pressure etc, until someone asked why the Suction Temp was at100F+. I found I had to increase the Capacity slightly to bring down the Suction Temp. My theory was that the minimum capacity was taken up by the Liquid injection. A simple adjustment to minimum Slide Valve was the cure.
I appreciate your input.

[Andy]

I think that is as good an answer as I am likely to find.
Also I think it is most likely the correct one.
I am however at the limits of my knowledge on screw compressor design.
With the VI set too low don’t we have in fact under compression?
I did conclude the rattling check valve and high suction at the compressor (compared to the Suction line) was due to the lack of gas flow.
You are probably correct about the low gas flow through the machine even at low VI. I have no good way to quantify that without shutting down part of the other Engine Room, which the Customer would be reluctant to do as an experiment. Or I could startup the Old Spiral Freezer, but that would entail a long pump out to put it back in mothballs.
The plant will be in continuous operation until October, even then it is a five hour roundtrip to test anything.
Fortunately this is a Frick which measures the Suction Pressure at a port on the Compressor, rather than a FES which uses a port installed upstream of the Suction Check Valve. That might have made things more difficult to observe.
Unfortunately I may never see the inside of this machine. We will most likely get a exchange from Frick.

I have been told before by knowledgeable people (As well as the manuals) that a screw produces about 15% of its rated flow when unloaded. I wonder what the effect of liquid injection cooling is on that.
Earlier this year I started up a used 200HP FES Booster in parallel with a 250HP Frick. Everything was going well as I ran it unloaded, checking Oil Pressure etc, until someone asked why the Suction Temp was at100F+. I found I had to increase the Capacity slightly to bring down the Suction Temp. My theory was that the minimum capacity was taken up by the Liquid injection. A simple adjustment to minimum Slide Valve was the cure.
I appreciate your input.

Hi NH3LVR

Say the slide was not actually loading, but the piston and rod were moving up the cylinder. You coud get a similar effect when the volume slide moves over the rotor. But then you would need the rotor to be missing before the volume section

I would pull the compressor down to check before sending away, just incase it is a slide problem.

Kind Regards Andy

[US Iceman]

I have been told before by knowledgeable people (As well as the manuals) that a screw produces about 15% of its rated flow when unloaded. I wonder what the effect of liquid injection cooling is on that.

The numbers I mentioned for this (15-20%) are an approximate range. Some will say 10%, others will say 15%. If I'm concerned about no-load operation I tend to use the higher numbers to be safe, just in case it might matter,

You were probably seeing the 100F suction temp, since the sensor was on the compressor side of the suction check. On the FES machine, it may not have been noticeable.

I think that was another condition of very little suction flow. I could see this happening with the slide fully unloaded at 0%. Depending on the operating conditions, the suction check could have been only partially open.

Unfortunately I may never see the inside of this machine. We will most likely get a exchange from Frick.

Can you request a copy of the tear down report with some pictures? It might make for some good reading.

With the VI set too low don’t we have in fact under compression?

There is a good explanation for this, as soon as I can remember it. It's been too long since I was involved in a really good technical discussion of screws. I need to think this one through a bit.

In either case, over-compression or under-compression cause the motor amps to be higher than desired.

With over-compression the gas is compressed to a greater pressure (in the discharge port area), than the pressure in the discharge line.

With under-compression the discharge line pressure is greater than the discharge port pressure. In this case, the compressor is working against the discharge pressure in the line.

[US Iceman]

With over-compression the gas is compressed to a greater pressure (in the discharge port area), than the pressure in the discharge line.

With under-compression the discharge line pressure is greater than the discharge port pressure. In this case, the compressor is working against the discharge pressure in the line.

OK, I think I have it now. Here is the relationship for Vi and over/under compression.

Pdisc = Psuct X Vi^k

Pdisc = discharge pressure (psia)
Psuct = suction pressure (psia)
Vi = Volume ratio setting with variable Vi, or fixed Vi of compressor
k = specific heat ratio of the refrigerant (NH3 approximately equal to 1.37)

If you open the attached PDF I quickly made up you will see various Vi's listed as a set of curves on the chart. The solid red horizontal line represents a constant discharge pressure of approximately 180 psig (12.41 bar-gauge) (about 195 psia, 13.44 bar-abs.) for a condensing temperature of 95F (35C).

If you look at the lower left hand area of the chart at 20 psia (1.38 bar abs.), you will see the light blue line for the Vi=5.0 intersects with the red line.

This would be an ideal match for that operating condition. If the suction pressure increases with a Vi=5.0, the discharge pressure from the compressor will be greater than the red line. This is over compression.

If the suction pressure is 10 psia (0.69 bar abs.) and you still use a Vi=5.0, then the discharge pressure in the compressor will be about 100 psia (6.89 bar abs.). This is under compression.

The discharge pressure is below the condensing pressure. The compressor still has to get the pressure up to about 195 psia (13.44 bar abs.) to get the gas out of the compressor.

Either over or under compression requires additional energy. That's why Vi selection is so important for screw compressors.

On recip's it doesn't matter. As they are variable Vi by design. In fact they were the first variable Vi compressors. A long time before this was "invented" for screws.

Because of physical limitations in the screw compressor housing a certain amount of restrictions decrease the available range of Vi that can be built.

For some applications (like single stage, high compression ratio) a Vi of 5.0 or 5.8 is still not high enough for best efficiency. It's just something we just have to live with.

The compressor manufacturers can only provide a certain amount range because of construction details in the compressors.

It took me a while to remember this, but does that clear up the questions?

[NH3LVR]

Thank you for taking the time to work that out and creat the PDF. You did send me back to school With the Carat Symbol (^).
I did talk to a friend who recently retired as the Service Manager for a Frick Factor. He belives the Slide Valve is worn badly and bypassing. Hopefully we will get a chance to look inside before shiping it off to Frick.
Again thanks for the help.
NH3LVR