Predictive Chiller Plant Control

[digi]

Hi, Anyone knows about the Predictive Chiller Plant Control? What will be the control's strategic?
Got one customer he request for that control sequence. He explained that the controller should have the learning capability to knows at what time of the day to start how many chillers at one time instead of kick in the chiller one by one. This concept looks interesting, but what will be the advantages and disadvantages?

[amigo]

sounds like optimal cooling start,
Here is a brief glimpse of what needs to be considered.
If you do not have thermal storage, the pricipals of load consideration should still apply.
http://www.stacenergy.org/projects/03-STAC-01/05-2004-03.pdf

[digi]

Thanks amigo.
I need to find out more on the details from the customer.
Their intention is to let the controller learn the chiller plant sequential control for whole year. So when come to following year, the controller will follow the previous year sequence to start/stop the chillers.
This building is in Asia and do not have any thermal storage. So the load is quite constant as only the hot day and rainny season.

[junkman]

You would probably have better luck tapping into a weather forcasting service on the internet and feeding the data into your automation system.

But, if the weatherman can't get it right, what makes you think your program will?

and like the stock market,

past performance is no indication of future performance..

Just something to think about.

What you will learn is how long it takes to cool or heat your space per-minute from what temperator and at what outside temperature. The data would be interesting.

[Edited by junkman on 07-01-2005 at 11:35 PM]

[jimmyj]

My 2 cents, If the Chiller plant has no thermal storage, than why do they need such advanced load profiles???? This is a simple load sequencing and runtime equalization thin. yes/no?