Refrigeration rack compressor sequence control

g.mccormick · Nov 25, 2015

We have a ref. rack that has 6 compressors. Each compressor is 3cyl with 2 unloaders. We therefore have 18 discrete "power" stages.

Currently the system that controls this rack is less than ideal.

My goal is to replace the controls with something different that I will be in control of.

The current system seems to bring on compressors in line, ie compressor 1 all cylinders, compresssor 2 all cylinders, etc, etc.

One main problem is that there are min on/off times to keep in consideration. Also, the sequencing allows that compressors that are running may be running fully loaded, while other compressors are being turned off. The problem then becomes that since a compressor must stay off for a certain amount of time, the system cannot react to a load.

It seems to me that keeping the compressors running unloaded would be a better option. The unloaders can be turned on/off basically at will with no adverse effects.

My idea is basically.
1. The first 6 stages are the 6 compressors running 2/3rds unloaded.
Example. 1compress/1cyl on at 5%, 2comp/1cyl each on at 10%, etc etc.
2. After all the compressors are on, start then coming on with more cylinders.

Depending on the setpoint of the system, I could utilize more/less compressors in my list, or perhaps utilize a slow integrator to disable compressors after the initial pull down.

Does this sound at all like something that would work?

g.mccormick · Apr 29, 2016

All,

Taking over of refrigeration equipment is looking more and more likely. Is there anyone that has experience with PLC controls on a refrigeration rack and air cooled condenser? I may need someone to bounce questions and ideas off of.

Equipment is 240hp R507a Rack.
6 compressors with 2 unloading heads each (18 total stages).
2X Sporlan Seri-J EEV expansion valves. I plan to use Sporlan IB boards to drive the valves.

Currently the rack is controlled by MicroThermo controls. There are various sensor boards, compressor boards, case controllers, etc. I plan to replace with PLC controls.

Condenser is 12 Fan air cooled condensing unit. Each 2 fans are controlled by same control signal s 6 stages of fans.

Basic controls strategy:
Currently this is how the controls work and I think I would target the same control scheme.
1. The compressor loading is targeting a suction pressure. The pressure setpoint is a function of the requested temperature that we want. The system is cooling a liquid through a heat exhcanger.

2. The EEV valves are controling superheat out of the heat exchanger. Currently this is done with the case controllers, but plan to do the control inside the PLC and utilize the Sporlan IB drivers to drive the valves.

3. The condenser fans controls stage on/off the fans to target a drop leg liquid pressure of 180PSI (80F temperature). I plan to control the same way.

Anyone familiar with industrial refrigeration controls have any lookouts and gotchas for me?

shooter · Apr 29, 2016

Find yourself an old book from danfoss, called automatic control of refrigeration.
( i do have it, as a book (sorry not scanned yet)
compressor can not run idle
there are 18 stages as you say.
starting and more is a simple sequence of all 18 steps, if more is needed and a compressor is on restarttimer the next one will get a startpulse auto, however the previous does not get a pulse as it is in restarttime.
decreasing is normally down to 33 % and the the previous is decreasing to 66% and sometimes to 33% and then the highest is stopped.
increasing should be double speed, while decreasing should be normal speed, with a quick stop when suction pressure gets too low.
This is all due to the control of the suction pressure.
The suction pressure is controlled at first just by a setpoint, however this setpoint can be autoadjusted by the EEV. when any eev is bumping against its setpoint(start at 7 C) the suctionpressure is decreased, when both valves are above setpoint lets assume 9 C) the suction setpoint is increased 0.1 Bar, This will give the highest possible efficiency of the set.
The Superheat is as low as possible. try to get it down to 3 C depending on the type of evaporator and installation.
If there is a gas/liquid heatexchanger to get the liquid cold by the suction gas. (often a pipe in a pipe) put the sensor just in front of the compressors. This way the suction gas is the coldest possible, giving best performance of the exchanger.

The condenser fans are a bit tricky, as it depends on efficiency of the fans. start on each condenser with 1 fan and build it up until the fanpower is higher as the profit of compressorpower.
example running 3 fans taking 20 kW. a extra fan is 4 kW and save on the compressor 10 kW, so switch it on, 6 fans are running and advantage on compressor is only 3 kW do NOT switch on.
it is a complex graph with capacities and outside temperatures.

you will also need a seqeunce strategy to give all compressors same running time (i do this never, to not have maint\enance all compressors at same tiem.
so i let one compressor running very much and that way i have a compressor without running hours, on standby.

valve drivers, i have replaced them with an arduino (yep only 3 dollars and some small parts)

g.mccormick · Apr 29, 2016

shooter said:
Find yourself an old book from danfoss, called automatic control of refrigeration.
( i do have it, as a book (sorry not scanned yet)
compressor can not run idle
there are 18 stages as you say.
starting and more is a simple sequence of all 18 steps, if more is needed and a compressor is on restarttimer the next one will get a startpulse auto, however the previous does not get a pulse as it is in restarttime.
decreasing is normally down to 33 % and the the previous is decreasing to 66% and sometimes to 33% and then the highest is stopped.
increasing should be double speed, while decreasing should be normal speed, with a quick stop when suction pressure gets too low.
This is all due to the control of the suction pressure.
The suction pressure is controlled at first just by a setpoint, however this setpoint can be autoadjusted by the EEV. when any eev is bumping against its setpoint(start at 7 C) the suctionpressure is decreased, when both valves are above setpoint lets assume 9 C) the suction setpoint is increased 0.1 Bar, This will give the highest possible efficiency of the set.
The Superheat is as low as possible. try to get it down to 3 C depending on the type of evaporator and installation.
If there is a gas/liquid heatexchanger to get the liquid cold by the suction gas. (often a pipe in a pipe) put the sensor just in front of the compressors. This way the suction gas is the coldest possible, giving best performance of the exchanger.

The condenser fans are a bit tricky, as it depends on efficiency of the fans. start on each condenser with 1 fan and build it up until the fanpower is higher as the profit of compressorpower.
example running 3 fans taking 20 kW. a extra fan is 4 kW and save on the compressor 10 kW, so switch it on, 6 fans are running and advantage on compressor is only 3 kW do NOT switch on.
it is a complex graph with capacities and outside temperatures.

you will also need a seqeunce strategy to give all compressors same running time (i do this never, to not have maint\enance all compressors at same tiem.
so i let one compressor running very much and that way i have a compressor without running hours, on standby.

valve drivers, i have replaced them with an arduino (yep only 3 dollars and some small parts)

I'm having a hard time understanding/deciphering your post. Note that saving electricity is not really my main goal, this is part of an engine test cell. Performance/temperature control of my process is what I am concerned with.

shooter · Apr 30, 2016

I think i was a little too fast for you.
Yes converting to a PLC is not big problem.
the sequence of compressors is rather simple, as going up is just 1 step at a time, and going down is with a flipflop to reset a step in the previous compressor.
Make sure you have a mass to get a smoother control.

the valve drivers only need a simple 4~20 mA signal.
Superheat (for control is a pressure sensor and a temp sensor close to the suction of compressors. Make a table(array 10) of this line and control the valve to be 3 C above this line.

condenser is really simple , switching on/off is fine for them, in proximity of houses do NOT use VFD as it is far better to have a constant noise then this changing every time.

shooter · Apr 30, 2016

The performance can only be better controlled
and temperature control will be better as you can control the process with expertise in stead of several controllers running in a cascade.

g.mccormick · May 4, 2016

Here is what little I have so far. I'm struggling to think how to best stage the compressors based on PID command. If anyone cares to peruse and comment on what I have so far please feel free.

It is in Automation Direct Productivity Suite software

g.mccormick · May 5, 2016

I feel that I have made some good progress on this project now. I believe that I have a nice way of sequencing the order of the compressors as well as staging compressors taking into account if a previous stage cannot run. Uploads include full project as well as two pdfs showing where I do my ordering of compressors (which is 1st,2nd,etc), and how I turn on a compressor based on where it is in the list, what the requested number of stages is, and what the preceding number of compressors is.

I anyone would like to make suggestions please do.

Thanks
Garrett

NOTE: PDFs were too large.

Refrigeration rack compressor sequence control

g.mccormick

Lifetime Supporting Member

g.mccormick

Lifetime Supporting Member

shooter

Member

g.mccormick

Lifetime Supporting Member

shooter

Member

shooter

Member

g.mccormick

Lifetime Supporting Member

g.mccormick

Lifetime Supporting Member

Similar Topics