Automated Chiller Temperature Control System

[David Gin]

Hi All,

I'm working on an automated chiller control system for final year project. It is a logic control that automatically set the chill water supply temperature based on current demand load and the chillers performance curve to achieve the best efficiency COP for the system. For a variable primary flow system where at least 2 chillers are running, how would one go about starting this?
Is there a set of algorithm for guidance?

 

[CWatters]

As it's a final year project (aka Homework) we shouldn't really just give you an answer/do it for you. What are your own thoughts?

Perhaps start by defining what exactly it is you need to optimise and how you would measure/monitor it.

For a given load what chiller parameters can you change that might effect the COP?

 

[russ_watters]

Have you researched how these things are currently done? Consulted with the vendors of such products?

[Edit]
I see some potential conceptual issues with what you want to do, but it does depend on the details. Where did you get the idea for this project? What is the chiller making chilled water for? How much do you know about the needs of the system it is supplying? Is this college or high school? Do you have an advisor?

 

[David Gin]

Hi all,

Thanks for your reply. The control logic would ideally be system-free which means that it does not depend on a system to work, however I am currently working with a variable primary flow chilled water system.

What I have so far is feeding the controller the historical data of each chillers, the performance curve of each chillers, the current parameters, the environmental parameters (outside drybulb, etc) to the control logic. Based on these data, it would config the chilled water supply set point of each chiller to meet its maximum efficiency that meets the current cooling loads.

I would like advise on the feasibility of this and is there increase benefit utilizing such control logic? Thank you for your time.

 

[CWatters]

I know very little about chillers but how else could you do it if you don't have some sort of control logic?

You list several things that you plan to feed into the controller but are they necessary? Which have the most impact on the COP? I a believer in KISS.

 

[russ_watters]

The control logic would ideally be system-free which means that it does not depend on a system to work, however I am currently working with a variable primary flow chilled water system.

I'm asking about the system(s) that the chiller is feeding cold water to. It sounds like it is an HVAC system, but the details of the load matter.

What I have so far is feeding the controller the historical data of each chillers, the performance curve of each chillers, the current parameters, the environmental parameters (outside drybulb, etc) to the control logic. Based on these data, it would config the chilled water supply set point of each chiller to meet its maximum efficiency that meets the current cooling loads.

I would like advise on the feasibility of this and is there increase benefit utilizing such control logic? Thank you for your time.

Generally the chilled water temperature is kept constant, but again it would depend on the details of the requirements of the system(s) it is feeding. There are other parameters that often vary though, such as condenser water temperature and the number of chillers operating (and therefore the flow through each chiller). And yes, there is a benefit.

Anyway, you left most of my questions unanswered...

 

[Ronie Bayron]

Hi All,

I'm working on an automated chiller control system for final year project. It is a logic control that automatically set the chill water supply temperature based on current demand load and the chillers performance curve to achieve the best efficiency COP for the system. For a variable primary flow system where at least 2 chillers are running, how would one go about starting this?
Is there a set of algorithm for guidance?

Concentrate on the return and supply temperatures of main distribution line Qm=mCpΔT/dt, each chiller unit has a Q capacity
Say, Qm=Q1+Q2+...+Qn
If Qm<Q1 then Q1
If Qm>Q1 then Q1+Q2
So on and so forth.