Sequencing VSD and constant speed chillers

[DJDatHTL]

We have an existing central utility plant consisting of two York 1500-ton constant speed centrifugal chillers and a side-car arranged plate & frame heat exchanger (750-ton). There are VFD pumps on both the primary and secondary chilled water systems (which are de-coupled) and constant speed condenser water pumps feeding two Evapco VFD cooling towers utilizing an OA wetbulb reset of setpoint from 85-57DegF CWT (based upon temperature in Binghamton, NY).

We are going to add four additional York 1380-ton VSD chillers along with similar pumps/VFD's.

I am wondering what the best sequence would be to maintain efficiency with regards to chiller sequencing based upon chiller type/load/pump speed/etc.

Are the strategies between constant speed and variable speed chillers completely different? I haven't worked with the VSD chillers yet, and I am just beginning to explore them.

Thanks!

 

[ChasBean1]

I hate to say this in this forum (because typically we are above this) but you should consult the manufacturer regarding variable 1° (primary) chilled water flow.

Apart from this, a 1°/2° system with decoupling should stress the importance of proper 2° CHW D/P and flow control and proper chiller staging based on 2° CHW demand.

 

[235zzzo]

Hi!

A few comments:
-Those kind of chillers, certainly have several thermodynamic groups working sequentially along the 1º('s?) circuit(s) temperature range.

-If we define/combine all the chillers we can put them to work sequentially to respond to all possible simultaneous demands. We have to exercise a global simulation to realise such "not so difficult and interesting data to get".

- Why to insist having VFD-pumps in the 1º('s?) circuit(s). Well, I can accept those expensive VFD-pumps in the cooling condensing circuits to the cooling-towers, because the atmospheric conditions can change a lot, as well.

- On contrary, it makes sense (at least to me) if we have VFD-pumps for the secondary/distribuition circuits in order to respond to the variable cooling loads.

- ...

zzzo

 

[MarauderX]

Gotta say my experience with constant vs. variable is to split the systems up. We had a courthouse addition job that we used new VSD chillers we wanted to connect via a kinda complex valve system to incorporate them with the CV system. Eventually we backed off and split the systems from each other due to potential problem scenarios with the CV starting & stopping driving the VSD primary & secondary pump demands crazy. Our rep didn't recommend it, so our controls guy redesigned the system to be isolated from one another with simpler demand controlling for each set of chillers.

On the secondary side the system operated with 1 VSD chiller on all the time, to handle the in-between loads. The CVs would fire up one at a time once the load began to get beyond the 1st VSD, then the 2nd CV, then the 2nd VSD would come on-line if the load between the two VSDs was above their combined minimum.

Anyway, it's an interesting control scenario, but we split the primary side completely to eliminate potential hazards, let alone confusing maintenance.

 

[imok2]

DJDatHTL
In order to get you familier with your situation I urge you to go to this web site:

http://www.automatedbuildings.com/

click on Tom Hartman and read his articles on All-Variable Speed Centrifugal Chiller Plants: Can We Make Our Plants More Efficient? Why the sudden change in attitude toward variable speed? and the 2 above it.Hope this helps
Roger

 

[235zzzo]

Hi IMOK2 !

I tried to acess to the site

http://www.automatedbuildings.com/,

and I was no able to reach it. Can you confirm that reference you gave?
Thank you!
zzzo

 

[quark]

zzzo!

Click on the link in your own post.

Eng-Tips.com : Solving your problems before you get them.

 

[imok2]

zzzo As quark says "click on your own link" mine has "click" in it but not supposed too.
Roger

 

[235zzzo]

Hi IMOK2 !

It has worked this time!
Thank you!
Nice weekend to the Forum!
zzzo