Staged Chiller Control Methology

[69FIREBIRD]

Hi can anyone recommend the best way to program a 4 stage chiller (actually 2x 4 stage chillers in parallel). should I keep it simple and look at differentials from a setpoint coupled with "interstage" time delays and dead bands to bring on each stage or is their a better / more energy efficient / accurate way? - PID with stages set a % of output?

I’m just trying not to reinvent the wheel - if someone has a tried and tested way to stage control I'd much appreciate some insight.

Cheers.

 

[TConnolly]

If you search the forum there are several threads that deal with chiller control - as well as some things that you should do to protect the compressors and avoid short cycling, etc.

By four stage do you mean that your compressor has an unloader and a hot gas bypass valve? This is less common.
Stage 0 = off
Stage 1 = 25% On, one cylinder unloaded + hot gas bypass
Stage 2 = 50% On, one cylinder unloaded
Stage 3 = 100%
That is what is typical for four stage.

If you only have a hot gas bypass and no unloader valve (which is more common) then you have three stage control.

The simplest way is the first method, use deviation from setpoint and timers to stage up and down. This way you can easily make sure that you don't short cycle your compressor and make sure you meet minimum on times and off times. Again, see the other threads for info on that.

 

[69FIREBIRD]

Hi Chillers are around 20-30 years old original control has been removed some time ago and current control is via PLC. Their are 4 stages (stage 1 comes on with the compressor) controlled via solenoid valves and compressor start up is via some sort or partial winding method. their is no money for a replacing the chillers at present but we are going to start from scratch re control as the existing setup has been modified so many times, their are no drawings, and the PLC control is over complicated an buggie. So I think best option is keep it simple and look at differentials from a setpoint coupled with "interstage" time delays and dead bands to bring on each stage. But I'm interested before I jump into it to get some feedback on other ways or more efficient or accurate ways to control.

have doe a quick search for chiller control on the site - can you point me ot something specific

Thanks

 

[TConnolly]

It sounds like you don't yet have a strong grasp on how that chiller works. Chiller control is not something that should be attempted by someone who doesn't understand chillers, unless you really do have money to replace it. If you don't understand it you will very quickly be replacing the compressors. If my assessment is accurate then I suggest that you engage the services of a local refrigeration expert who is well versed in chillers to help you create the control design parameters and to check your work - otherwise by the time you are done you may pay for a new chiller yet still have the old one.


http://www.plctalk.net/qanda/showthread.php?p=209322

http://www.plctalk.net/qanda/showthread.php?t=41037

http://www.plctalk.net/qanda/showthread.php?t=30095

http://www.plctalk.net/qanda/showthread.php?t=49761

http://www.plctalk.net/qanda/showthread.php?t=40918

http://www.plctalk.net/qanda/showthread.php?t=39557

Knowing the age of the chiller helps in that it narrows down how the chiller capacity is controlled. At that age you have two kinds of capacity control 1)Hot gas bypass. 2)Cylinder unloading.

Does this chiller have semi-hermetic compressors or hermetic compressors? (if its a hermetic compressor then you can have hot gas bypass only for pressure control)

If it has semi-hermetic compressors, is there a solenoid valve mounted directly on the top of one compressor head? (If yes you probably have an unloader valve for capacity control. In this case you may or may not also have a hot gas bypass valve. Each kind of valve cuts cap

How many solenoid valves are in the refrigerant lines? A sketch or photo would help.

You said 2 x 4 stage chillers. Do you mean two completely separate chiller systems?
Or is this two chillers on a single pump tank? If so, do the two chillers operate as redundant chillers, or are both required for capacity?


Because chiller cooling capacity is constant, and because compressors cannot be time proportioned, setpoint deviation and staged timers is the best way to go. Where your possible capacity steps are 0, 25%, 50%, and 100% (75% is not an achievable capacity for a conventional compressor configuration - and it has yet to be determined if 25% is achievable for your configuration) and where compressors must stay on for a minimum amount of time, and once switched off, stay off for a minimum amount of time, PID control would just add headache over and above the simplicity of just using set point deviation and staging timers without improving the quality of control significantly.

 

[lostcontrol]

The chillers should have their own safeties built in, you should re-instate these, or try and resurrect them at least.

These standard things like LP/HP & Oil Pressures should all be hard-wired, with the PLC only monitoring them, and alarming if there is the facility to do this.


PLC control is good for staging them in/out, & loading them.

A fridgy (?) should be able to provide all the required spec's on this system, what the ins/outs are. There must be someone that is familiar with this plant, & how it operates?

 

[69FIREBIRD]

Hi Thanks for you reply. I have a reasonable understanding of refrigeration control and safeties and am working with a refrigeration engineer on this project.

Not sure about the semi-hermetic compressors or hermetic compressors - I'll have to check with the engineer.

will attempt to get a sketch of the set up together today.

Their are 4 solenoid operated stages on each of 2 refrigeration chillers set up in parallel with each other. Other than the first stage that starts with the compressor the other 3 stages on each chiller load/un-load depending on the required cooling capacity.

The existing PLC control looks to do multiple control methods if both are available then it stages up/down one stage on each chiller at a time based on time and if only one is available (other faulted or summer winter mode) then only one will stage up, their is probably some sort of lead lag set up as well - very complicated and that I think is part of the problem compounded by undocumented erratic and fault prone ancient electrical control.

will have a look at the forums you refer to later - thanks again.