New Chiller Design

[SRBKDW]

A client has asked for a new chillled water plant for a medium sized building. Air cooled chiller feeding the chilled water plant will be sized at approximately 100 tons. This chiller will feed 5 or 6 separate chilled water air handling systems. I'm evaluating pros/cons of designing the system with 3-way valves or 2-way valves. If one were to go with 2-way valves would you recommend a VFD on the pump or just riding the pump curve. I'd like to get input from some of the smart folks out there about items to consider and what they would do in this situation. Thanks in advance for your help.

 

[quark]

Though I would go with VFDs ideally(and I did) to get the energy benefits, it all depends upon your load pattern and corresponding budgetary limitations.

Generally a centrifugal pump should not be run below 25% of the flow rate at BEP (i.e below safe minimum flow). If you reduce the speed of the pump, safe minimum flow also gets reduced.

You should tell us the maximum, minimum flow rates, static discharge head and total discharge head at full flow to get a better solution.

Regards,

 

[TFurst]

A VFD and two way chilled water valves is a great system, in addition to this I would recommend using Griswold or Flow Design flow controllers at each coil as they will modulate to maintain the correct flow rate regardless of where the pump is operating at on its curve and how many other coils are calling for cooling. The other big issue to deal with is maintaining the correct minimum flow through the chiller, you might want to go to Tranes website and do a search on VariPrime which details the requirements for a variable primary system. Depending on how much piping you wish to change you can aslo look at a decoupled system so that the design flow is maintained through the chiller at all times.

Tony

 

[imok2]

From where I stand the decision is basically "how much money are you willing to spend(waste?) If the cost of energy is not an issue, then by all means go the easyist way with constant volume pumping and 3- way valves. If energy cost is an issue the the two previous posts are better and more cost effective.

 

[SRBKDW]

Thanks for the replies. A couple of thoughts.

Imok2 - How much money are we really talking about saving in reality? Chiller electrical costs will remain essentially the same I believe so the main area we are saving energy cost is in pumping costs. The pump minimum threshold I think will be determined by chiller's minimum flow threshold as opposed to the pumps minimum flow threshold. Assume for example that a 100 ton chiller's minimum threshold is ~150 gpm versus full design flow of 240 gpm. Is the energy savings really that significant on a 5 hp to 7.5 hp motor? Please let me know your thoughts.

TFurst - You lost me on the Griswold flow controllers. If I have a two-way valve it's going to modulate based on a variable I'm trying to maintain (room temperature or supply air temperature or similar). I believe I just want to balance it initially so that my full design flow is obtained when the valve is full open. After the initial balancing it should modulate based on whatever variable it is attempting to satisfy. Let me know if I'm missing something. You also lost me on decoupling the system. If I create a bypass so that my chiller is always getting full design flow what have I accomplished? My pumping costs remain the same as the pump is always pumping the full design flow even though it's a 2 way system. Let me know what I'm missing. Thanks.

 

[imok2]

OK, lets take an example:
HP = 7.5 x 746 watts/hp ~ 5600watts..now with a turn down of 50%
We know that the power consumed = Speed cubed so, with a 50% turndown we have 5600 watts x0.5 x 0.5 x0.5 = 700 watts.
Now the curve on a typical office building heat load looks like an IQ curve. (did i say that)! Anywhy, the pumps will be mostly in the 50 to 60% speed range. If you consider what you pay for a KW and the hours of operation .. well I leave it up to you.

 

[TFurst]

In response to the question about the Griswold flow controllers, they are actually a variable orifice balance valve. One of the problems with any hydronic system that uses two way valves and A VFD is that as the different control valves modulate in response to load changes the actual system dynamic pressure differential changes as the pump speed changes through the VFD. As this happens you can get widely varying flow rates through your coils. Installing an automatic flow controller (balance valve) eliminates this problem as the valve will maintain the design flow rate regardless of DP (depending on cartridge). Now on the matter of decoupling the system the one area of greatest concern is maintaining proper flow through the chiller whenever the chiller is operating, using two pumps as in a decoupled (primary/secondary) system assures that you will always have the correct flow through the machine while still allowing you to capture the savings through reduced pump horsepower with the VFD. No it is not the greatest system in the world and yes there are a myriad of ways to design any system this is just one way I have done what you are doing.

Tony

 

[Bittooria]

VFD energy benefits shall acrue only if 2-way valve are provided. Otherwise, if VFD is not considered any of 2-way or 3-way valve may be used. One problem noticed with 2-way valve is balancing of system particularly in large network few circuits may get very little or scarce water.
You can learn more on a paper M150, titled design considerations of pump hydronic system at PDHonline.org.

 

[khatra]

I would use 2-way control valve with VFD for the secondary pumps. Make sure the primary pump are constant flow.