View Full Version : System Pump Down

09-05-2001, 11:28 PM
I had a thought today, no it's OK it didn't hurt at all, regarding pump down operation on larger systems.

My experience of them is in relation to shell and tube water chillers using semi-hermetics, I am not including the smaller refrigeration units in this.

Re-reading papers concerning compressor losses due to oil starvation or dilution I began pondering the effects of repeated pump downs after pressure increases due to temperature rise within the evaporator. I appreciate that leaking solenoid valves will make the problem worse but that is another matter.

With these repeated pump downs does, or will, the oil level be lowered to a dangerous degree prior to the next start proper, bearing in mind that most oil safety switches have a 2 minute time delay before activation.

Also, of course the repeated starts do not help the state of the electric motor or motor starting components.

On a large system and assuming that a continual pump down is required, as against a once-only pump down, would it make sense to install say a small hermetic compressor across the main compressor purely to maintain the required pump down pressure?

Is this type of protection already in use or can anyone see a use for it?

10-05-2001, 04:14 AM
Interesting thinking, Brian The first thing that comes to my mind.... the thing that surprised me about advice regarding pumpdown... is that you don't necessarily set the cut-out just above atmospheric pressure, as I was programmed to do for so many years. Setting the pumpdown higher should decrease short-cycling, improve oil return and improve performance over the lifetime of the unit.

Can you say more about your specific example that prodded you to think about an auxiliary compressor?


10-05-2001, 08:08 AM
No, there wasn't a specific problem that I had in mind Dan it was just monitoring some pump downs recently.

Admittedly I didn't have cause to investigate as I was involved on other things but the 'normal' shut down seemed to be to pump down as required. Shortly after this the compressor would do a rapid start/stop as the pressure rose again.

Occasionally this would repeat several times before the remaining vapour had been removed.

I suppose it is because I dislike heavy machinery being jump started when it shouldn't.

10-05-2001, 08:10 AM
Originally posted by Dan
...set the cut-out just above atmospheric pressure, as I was programmed to do for so many years. Setting the pumpdown higher should decrease short-cycling, ..etc

Yes I agree with that and was what made me think about it in the first place.

Would a small 'pump down' compressor overcome the tech with itchy fingers:)

11-05-2001, 05:20 AM
Originally posted by fridgetech
There would normally be a large liquid mass in the evap at the end of a cycle, more so if the low load operation was lengthy before the pump down, then too there'd be a large oil mass in the evap.

If the controls are configured for non interupt pump down then there can often be multiple pump down recycles before the suction settles below LP cut-in.

Often these days one would introduce an interupted pump down controls configuration and then at the next start up have a pump down start-up with a LP switch timed overide.

Let me be the gadfly here, but mostly during the first pumpdown you have relieved the system of liquid floodback worries. That is when the large flooding issue is resolved. After that it is repeated starting and stopping for no good end. Having the cut-in pressure as high as practible is a good idea, though.

I disagree with Marc regarding the high refrigerant mass in the evaporator at the end of a cycle. I would think that when temperature is achieved that you would have the lowest mass in the evaporator. The compressor is pumping away just like it does when loads are high and the TEV is feeding bunches of refrigerant, but now doing the same thing when the TEV is throttling and permitting less refrigerant into the evaporator, due to low load conditions.

Brian's notion of a dedicated pumpdown compressor has theoretical and practical merit, but I would only want to employ such a complication after failure, once I utilized time delays or lockout circuits that could perhaps accomplish the same thing with less complication.

Interesting stuff, though.


17-05-2001, 02:10 AM
Oil migration is more critical with certain refrigerants. Ammonia can oil separate very easily. Yoy mentioned chillers so you are no doubt talking about suction temperatures above 30 F, Refrigerants like R22 do not have oil separation so repeated pump downs should not be a problem with the oil. Even with systems where the chilled fluid is down around 20F and the suction is about 10F, Ihave witnessed no problems. When in doubt use a good oil separator in the discharge line.

25-06-2001, 04:33 PM
Aside from the unnecessary complication, the problem I would expect with adding a small pumpdown compressor is that it may find itself inundated with oil.

09-10-2001, 06:53 AM
I was waiting for someone to raise the question of oil level in the small "pumpdown" compressor. My guess though, is that it would lose oil. I don't believe that the refrigerant velocity would be adequate to return it. The suggestion was offered to increase the cutout pressure setpoint. We have problems in the Northern climates with systems that fail to restart due to cold ambient conditions with outdoor units in the wintertime. If the TEV shuts down completely when the compressor shuts off, there should not be a problem with flooding the evaporator and slugging on startup. However, many valves are worn and leak slightly permitting the evaporator to fill with liquid over time. A liquid line solenoid wired to shut off when the compressor shuts down will prevent this problem unless the compressor valves are bad and allow refrigerant to flow backward into the low side. I have also found that liquid in the evaporator aids in defrosting the coil whether it be electric or air defrost.

09-10-2001, 01:04 PM
[i]. I have also found that liquid in the evaporator aids in defrosting the coil whether it be electric or air defrost. [/B]

I have been told this and blindly trust that it is true. But blindly, mind you. Is this a convection sort of thing? Fluid running around the tubes, whereas vapor fails to do so? Or is it conduction, for similar reasons compared to vapor filled tubing? Is it even possible that there is a latent heat transfer?

Regarding cold starts, it is a common solution to put an overriding 2-minute time delay that takes the low pressure control out of the circuit during start-ups when ambient is colder than the temperature we are trying to achieve. A bootstrapping sort of thing that has to make you smile, because tossing heat from something indoors to something outdoors when it is colder outdoors than it is indoors is just clumsy as hell.



Mike Hopkins
22-10-2001, 11:34 PM
Is anyone familiar with the timed, periodic pumpout? I have seen this used where the system has a long off cycle and is using a one time pumpdown, or weather conditions favor economizer
over mechanical in the morning but bring on mechanical during the afternnoon. Basically a timer allows a pumpdown every whatever you set the time interval for to satisfy the low pressure switch but not before the timer says so. Normally I would set them up for a 2 hour interval. Kind of an elaborate control scenario but accomplishes a good pumped down evaporator but not too excessive for the motor to deal with. Good for leaking solenoids and the such. I believe Trane takes credit for it, at least
that's where I got it from.

Mike Hopkins

23-10-2001, 11:38 PM
Interesting Mike. I am weak in a/c although we service a lot of Trane equipment. I would be curious to learn more about the timed pumpout.

Is its purpose to take advantage of an economizer but still operate sufficiently to control humidity? Is normal operation a pump down when temperature is satisfied? You have me curious there. From your brief description, it sounds like an anti-compromise as a solution for a design compromise.


Mike Hopkins
24-10-2001, 12:27 AM
This control scenario works in conjunction with the low pressure switch used during pumpdown. Say compressor pumps down and shuts off and it is a one time pump down system. The timer will allow additional pumpdowns at a preset interval. If you set it for every 2 hours then every 2 hours, if the low pressure switch is made, then the compressor will come on and pump the system down. If low pressure switch is not made then no pumpout will occur. Typically used on larger systems with large refrigerant charges and rather susceptible reciprocating compressors. Where it comes in handy is when you have a one time pumpout system and an economizer. Say you have a warm day followed by several cool days and then another warm day. The economizer handles the load on the cool days but if you had leaking valves or liquid solenoid you could have considerable charge migration to the evaporator. The warm day comes again and the compressor comes on and gets a dose of liquid and bye-bye she goes. Whereas if it had timed pumpout the amount of liquid would be negligible and not neccesarily cause major problems. You could modify an auto pumpout system with this control setup and avoid constant recycling on the low pressure switch which could cause oil loss, motor overheating, starter wear, etc. I have a schematic of this control system and could get it to you if you like.

Mike Hopkins

24-10-2001, 11:24 PM
Originally posted by Mike Hopkins
I have a schematic of this control system and could get it to you if you like.
Sorry, haven't been back to this thread lately, yes I like the idea of this one Mike and I would certainly like a copy if possible.

Mike Hopkins
25-10-2001, 01:24 AM
Well Brian let's try this attachment deal. I minimized this file as much as possible, but it is essentially the setup. The timer you would use is rather special and was expensive and difficult to get at one time but now with the electronic age is a simple matter.
Mike Hopkins

Mike Hopkins
25-10-2001, 01:29 AM
Duh, didn't go through that time, let's try again Brian.
Mike Hopkins

Mike Hopkins
25-10-2001, 01:40 AM
Well as pathetic as it is let's give a little clarity to it.
At the top of the schematic would be the non recycle relay contacts, below that in parallel are the auxilliary pumpdown contacts of the compressor starter. Below that are the timer contacts. As you can see if the timer pulls in and the low pressure switch is closed the downstream (not shown)starter coil can pull in and run the compressor until the low pressure switch breaks again. I may be able to get a more detailed diagram that shows the whole picture and shrink it down to fit this forum.
Hope this helps a little.

Mike Hopkins

25-10-2001, 10:39 PM
Thanks Mike, something else to add to the 'notes' folder. :)