Flooding the Evaporater

[Hopeless1]

I work in a energy plant that supplies a larger complex of building, and sence i have been here i have been in an on going arguement about CHW flow rate threw the chillers. I say that the designed flow rate would give the best performance and make the most money for our company by minimizing the equipment online or trim flows down to design when there are more than one chiller online. most on my coworkers believe that the flooded evap is alright and will make more money due to higher tonage output due to haveing more CHW pumps running. I say that you lose anything made due to electrical useage. so i'm suck in this to prove one way or another.

Can you help?

 

[MechEngNCPE]

Well, I don't know your system, do you have two or three chillers of around 1000 tons each, where you are on a primary/secondary loop plant? Do you have variable speed secondary pumps? Are your chillers inlet guide vane or variable speed compressors?

Chillers stage on and off depending on chilled water return temperature. As long as the chwr is low enough for one chiller to handle the load, do not bring the other chillers online. Now when your chwr temp gets above say 55, you will need to stage another chiller on.

Do not operate multiple chillers at reduced load, even though you get a better kw/ton at reduced load (down to 25%, then it goes way up) Just having the chiller running, along with the pressure drop through the chiller, etc. you're less efficient. Chillers are meant to be operated at around 80% load. I say once you hit that mark, stage another chiller on.

 

[MintJulep]

Presumably you bill based on delta-T and flow rate and pay a combination of kW-Hours and kW demand.

 

[MechEngNCPE]

I have read your question five times and it is still unclear to me. Is this a manually operated chiller plant? Do you turn pumps and chillers on and off manually??? Or is this a theoretical question?

 

[Hopeless1]

the sysetem:
Primary loop; 5 chillers : 500T, 1000T, 2000T, 2500T, 3000T
none are Veriable speed; it's a Manually operated plant.

As a contracted comapany, we don't own anything just operate, that said we are paid by the ton and credit back KW used at EoM.

here is were things get tricky the loop isnt a real loop but two legs and the load in the middle with a diff-press that must be maintained (13-18 psi). so chw pumps are used to keep up the psi. which increases the flow threw the chiller, in turn raises up the tonage due to flow rate. this setup tends to increase motor current on the chiller and take longer to bring down chw temp.

i just want to trim flow rate threw the chiller down to its operating flow rate but i'm told that it doesnt hurt anything and doesnt matter. and there is the stand-off.

if this helps i hope...

 

[MechEngNCPE]

What a horribly designed plant, no offense. So the chillers' compressors stage on flow rate? not leaving chws temp? Bizzare.

If that truly is the case, run your pumps to provide the required pressure at your load, and manually stage your chillers to provide adequite chws temp.

By the way... your load will use the same tonnage at a ten degree delta T at say 1000gpm as 500gpm at 20 degree delta T, not that you should go over say 14 degrees delta T in a chw system. but just to give you an example.

 

[Hopeless1]

its a 30yr old design. so then the over flow really doesnt matter in it all?

i just have a problem in my head with running 5800gpm CHW threw a chiller designed for 3500gpm CHW. (2000T chiller)

maybe its just me?

 

[MechEngNCPE]

Look at it like this.. You need a certain chws temperature and flowrate. If one chiller can provide that cooling capacity with half the gpm, maybe consider a bypass so you are not running all that water through that one chiller? I'm thinking pressure drop and pump innefficiencies here. One would have to know your exact system to determine the best operation, but typically you will gain nothing and lose kW's overall if you are operating above design flowrate. Just some thoughts.

 

[Hopeless1]

i have trimmed down the flow rate to designed spec and found that there is a KW drop, as well as a more rapid cooling effect to the chw system b/iue have a psi drop from the trimming but i slowly goes back up as i satisfy the air-handlers and shut off then allowing me to shut down addtional chw pmps needed to maintain the pressure diff.

i was hoping that there would be a mechianial reason to want to do that all the time so that other shifts would follow suit (like saying that it would save wear on the compressor.) other wise ppl would still just run the chillers at any flow rate.

hopefully saving money would be enough reason but old habits around here seem to die hard.

 

[MechEngNCPE]

Let the big horse do the work.

The 2000 ton chiller has a better COP, so if you can satisfy your loads with that chiller only, do it. Another method of decreasing your kW/ton is to raise the condenser water leaving temperature, to say 100F, and therefore requiring less gpm through the towers. Also, if you have VFD's on the tower fans, running 2 cells as opposed to one can increase plant efficiency by lowering the bhp on the fan motors. You may want to check that out.

There are multiple things you can do to save energy through chiller plant optimization. I did an energy center chilled water plant optimization project where we converted the system containing 3 chillers 3 towers, no vfd's, manually operated, primary/secondary w/3-way valves on the loads into a fully automated primary/vfd secondary with no 3-way valves, and vfd's on the compressors and tower fans. We saved up to 30% yearly energy consumption.

 

[MechEngNCPE]

I meant the 3000Ton

 

[MechEngNCPE]

Anyone else want to chime in on this? Anyone else have chiller plant optimization techniques? I know there are hundreds out there. What are some more things he can do without purchasing new equipment?

Can the implementation of flow/temp sensors/controllers help his situation?

The main thing, I believe is to work the 3000ton chiller up to full load, then stage the 2500ton on etc. until load is met.

I still like my idea of raising the delta t on the condenser side and using multiple cells at part load if that is possible.