Flooded R22 system

[Steve Wright]

Does anybody have a recommended shutdown and start-up sequence of operation for a flooded r22 system as relates to sequencing Compressors, recurculation pumps ect.

Steve

[US Iceman]

Hi Steve,

Is this related to a flooded system, such as a chiller, or to a liquid overfeed system using refrigerant pumps to recirculate the R-22?

[Steve Wright]

Flooded Air over Evap with screw comps similar to NH3, liq overfeed.

[SteveDixey]

Flooded Air over Evap with screw comps similar to NH3, liq overfeed.
The main issue is usually the rate of pull-down to prevent pump cavitation, usually expressed as so many degrees over a time period. What pumps are you running?

I think Witt recommend no more than 3 deg C per minute and I think that was for ammonia but we also applied it to R22 without problems.

I've come across systems needing to be pumped down to their evaporating pressure before pumps are brought in and others that start pumps up at around 1 bar over the final evaporating pressure. Some of this has been arrived at by a "suck it and see" approach. The FrigoPacks running Grasso recips I've worked on have usually brought the pumps on as soon as the compressor has hit around 2 bar suction.

The point at which to bring in the pumps may depend on your system's ability to deal with the rush of vapour down the suction when the pumped cold liquid hits the warm evaporator. If you load up too early, you may also trip out on HP or oil temp if you give the screws too much to do too early. What level of loading control can you get over your screws?

Shut down usually means shutting off the pumps and then allowing the compressors to pull down to the unload point to pull liquid back out the evaporator. When you hit your lower suction limit, set the system to shut off the screws.

Steve

[Steve Wright]

Thanks Steve That was good stuff.

This is a new account for us, it is a conversion from NH3 to R22, we did no engineering we are fourth or fifth service contractor on the job, several other problems exist.

The system has been poorly maintained many safeties and operating controls not in the system software. A local guy installed an AB controller and does not have it programed correctly he just arbitrarily brings on compressors and pumps.

I think compressors are Frick with full modulation capacity control.

Steve

[US Iceman]

Steve,

This is a very important point...


The main issue is usually the rate of pull-down to prevent pump cavitation, usually expressed as so many degrees over a time period.

The pull down rate limit I normally use is 1 psi per minute. I try to have the controls load very slowly to allow the pumps to stay out of cavitation at any time.

Unloading is not that big of a deal, so you can unload a system as fast as you want, as long as the control system does not overshoot.

For start-up, I like to have one compressor start at minimum load and then run some time to stabilize. For hot start-ups, the compressor should be operated first, to cool the system down. The refrigerant pumps are off as are the evaporator fans. In other words, no load except the boiling refrigerant.

This allows the pumps and refrigerant in the low-pressure receiver to cool down. After the liquid temperature is close to the design operating condition, the refrigerant pump can be started.

What I recommend is to open the liquid line solenoid valves with the fans off. This allows the warm evaporator to cool down. This is also a good time to check the hand expansion valve setting.

After the piping and evaporators are brought down to temperature, you can incrementally start the evaporator fans gradually.

This allows the system temperature to be brought down slowly and helps to prevent cavitation during a hot start-up.

For cold start-ups (the system is already down to temperature, but is shut off), the evaporator fans can be started (if they are not already running). The refrigerant pumps can be started.

When the suction pressure starts to increase above setpoint, start one compressor. Again, let it load slowly. After the system stabilizes, the other compressors can be started and placed into automatic control. Again, the load rate o the compressors should be very slow.

If this is a two-stage system, the high stage compressor should be started first to pull down the intercooler pressure. In some cases, the system may have a valve that has to be opened to pull down the booster (low-stage compressor) suction pressure through a bypass line.

After the intercooler pressure is down to it's normal range, the boosters can be started. If this is not used, the boosters or their motors may be overloaded and shut down.

If you need some help with the controls, send me an email and we can discuss this.

[SteveDixey]

Thanks Steve That was good stuff.

I think compressors are Frick with full modulation capacity control.

Steve

US Iceman has also replied with a similar set of proposals. It can depend on how much "human" input happens, say by manually opening and closing valves, or if it is a totally automated setup that just needs the "go" button pressing.

How many freezers? You can as Iceman says, put on a pump to pump over cold liquid to one freezer and then when the evaporator is iced over, bring on the fans, leaving the others off on their valves, and do one at a time.

If your screws appear to be only capacity regulated rather than speed regulated through a VFD, screws tend to be all or nothing. Little happens below 80% capacity and then it all happens in the final few points.

If you are suffering overloads, you could link pressure transducers in, or pressure switches with dead bands, to modulate the screws, by preventing discharge pressures climbing too high in response to high suction pressures, or by sensing your motor current through CT's and allowing a high current set-point to either modulate the capacity back to off-load to say 80%, or hold loading below 100% until suction pressure drops off.

In tandem with that, you may also want to consider a pump shut-off set-point that stops pumping liquid over if suction pressure does not drop (from say 2 bar),within a certain time frame, showing the compressors are not coping. You can then set the pumps to provide liquid only in response to dropping suction pressures.

You could also program the controller to be able to select a warm start, and it takes the plant down real slow, or cold start, when it can ramp up the load quite quickly.

Steve

[US Iceman]

All of this is very good information for developing control strategies. One of the problems involved with this, is the level of understanding of the desired operation.

In conjunction with this it is imperative to build in the safety and control logic to meet all of the anticipated operating conditions.

All of this can be controlled by the main control system providing the control system has the proper logic in it.

Ideally, the control is automatic and covers all potential issues as I feel it is better to have a fully automatic plant, rather than one where human interface is constantly required.

When the system is being shutdown for service (pumps outs, installation, etc.) this should be done by personnel to make sure the affected area is supervised.

Same thing for hot start ups. I think these should also be supervised.

Otherwise, the control system should be capable of operating plant in all conditions (normal or upset).

I agree with all points listed.

[ecclesk]

At what point in the sequence would you bring on the surge drum work level solenoid?

the systems i have, have a hp receiver feeding the surge drum via econimizers & the feed is controlled by a surge drum level controller that open/closes the solenoid

[SteveDixey]

At what point in the sequence would you bring on the surge drum work level solenoid?

the systems i have, have a hp receiver feeding the surge drum via econimizers & the feed is controlled by a surge drum level controller that open/closes the solenoid

You would usually energise the level control system when the compressor plant starts, otherwise you will back up condensed liquid into the condenser and also risk flooding over liquid from the economisers. The level controller then controls the surge drum level until the plant is switched off. I am assuming you have a normally closed valve that opens when the level switch breaks.

Is there a solenoid valve and level switch on the HP receiver?

Steve

[ecclesk]

You would usually energise the level control system when the compressor plant starts

This is how we control it at the moment, but i'm having some PLC program mods done in the near future so this would be the ideal time to make any other changes if required.


Is there a solenoid valve and level switch on the HP receiver?

No, the only way we can adjust the HP receiver level is to adjust the work level set point on the surge drum level controller or to adjust the setting of the hand reg valve that controls flow from receiver to surge drum

[SteveDixey]

This is how we control it at the moment, but i'm having some PLC program mods done in the near future so this would be the ideal time to make any other changes if required.

You might want to look at running the condenser water pump for maybe 30 minutes after the plant stops. Helps to stop the residual heat baking scale onto the tubes.

Steve

[ecclesk]

You might want to look at running the condenser water pump for maybe 30 minutes after the plant stops. Helps to stop the residual heat baking scale onto the tubes.

We have air cooled condensers.

At present the start up sequence is:

liquid pump & evaps on,

30 sec delay then compressor starts.

Any ideas why it may have been sequenced like this? i may consider changing the sequence as we do have cavitation problems now and again at start up

[SteveDixey]

We have air cooled condensers.

At present the start up sequence is:

liquid pump & evaps on,

30 sec delay then compressor starts.

Any ideas why it may have been sequenced like this? i may consider changing the sequence as we do have cavitation problems now and again at start up

You will get some cavitation problems, but if it settles out, there is not much to worry about. However, if the gauges are hammering up and down and you can hear the liquid rattling continuously, you need to address it.

The sequence as described is OK and is a pretty standard way of operation. You need to open the evap valves and start the pump liquid to give the compressors work to do. There will generally be a 30 second time out on the pump pressure sensing software to make sure pump pressure is established before starting the compressor. If the 30 second delay senses pump pressure has not made the set-point, it will not start the compressors, and should set an alarm as well as stop the pump.

I have set up FrigoPacks that suffer from pump cavitation to start compressors to run down to LP off-load, stop, and then kick in the liquid pump after the pump down. The compressor then starts up when sensing the suction pressure is rising above the upper set-point pressure. Leave the evap fans off until the compressor starts to pump down again.


Steve

[Andy]

Hi:)
what about the Frick compressor controller, usually this will carry out all the compressor loading/unloading sequence.
Main thing is to keep the fans off during the pulldown (blasts I'm talking about not chills) then start the fans incrementally as the suction pressure drops between -20 and -30 deg C.
If you can start the pump straight off without overloading the plant it will prime it'self better.
On the subject of overloading, the Frick controller should take care of this this, by restricting loading on amps and discharge pressure.
Kind Regards. Andy.:)

[Lazarus]

The Unisab 11 is a very good controller for retrofitting to most types of compressor

[SteveDixey]

What controller have you. Was it Mitsubishi series with E200 interface?

Steve

[ecclesk]

What controller have you. Was it Mitsubishi series with E200 interface?

Yes mitsi FX2n's & E200's Hmi, i'm having 3 plant rooms networked back to a workshop PC for a dial in/out monitoring system also a few other program mods, so this would be the ideal time for me to have any other changes made if needed.