HVAC Chiller Plant 1

[kepharda]

Hello, I have a new project to install some Water Cooled Chillers for an HVAC project. This is my first HVAC Chiller install. Previously I have installed Chillers, but for Process Systems.

I was wondering if anyone could offer advice or tips to ease the install.

I will be working with a very experience HVAC Engineer who will be responsible for the Air handlers for the project.

We will also be replacing the supply and return piping for the project, so any tips in that regard could be helpful.

thanks,

dave

 

[quark]

Installing HVAC chillers is much easier than process chillers and can take care of design inadequacies more flexibly, or atleast that is my experience. The load on a HVAC chiller is more smoother than a process chiller.

However, you have more scope for energy optimization incase of HVAC chillers. You have to first decide on what kind of controls you want to incorporate into the system. You can consider the following (but not limited to) schemes.

1. Variable speed systems at chiller, primary, secondary and AHU levels. Go for two-way valves if you are considering variable flow systems.

2. Ambient temperature reset.

3. Splitting the quantity if your load varies drasitcally.

4. Balancing valves at inlet to chiller for parallel installations, to avoid overloading.

5. Venting pipelines from the highest points.

6. A pressure balancing tank.

7. Type of chiller to suit your load profile (turndowns).

8. Safe minimum flow lines for the pumps (i.e direct connection supply and return headers)

9. etc.,

 

[Drivewizard]

Here are some sugestions that will allow you to generate some good will and respect as an Engineer among the Customers you have and more so among the Service Companies that have to work on your profect for the next 20 years.
Trust me, word gets around who designs the good plants and which ones are to be avoided

1. Try to use delta P switches instead of paddle switches for proof of flow. Paddles break easily
2. Make sure you spec that Delta P switches are wired in series with pump starter aux contacts.
IE: Chilled water proof of flow switch should be wired in series with that pumps starter N.O. aux contact.
3. Most Chiller manuf do not like a proff of flow coming thru a software point on a control system, should be hard wired
4. Try to allow as much room as possible above and around the chillers as possible. They WILL need to be overhauled eventually. You may save them money on th footprint of the plant upfront, but when they have to have special rigging made or overhaul labor is doubled due to no clearances customers will remember who designed the plant.
5. Try to put a means of checking water side pressure as close to chiller evap and cond bundles as possible. And do not use two seperate gages. Use one gage and manifold together using valves, if possible. If piping is on opposite ends, use a petes plug. Service Guys need to know delta P across bundles to calc flows and tonnage, we don't care that the pump has gages on it, we have flow charts for the chiller not the pump curves.
I am probably forgetting something, but these are the things that come to mind easily. Oh, one more: READ the IOM manual for the equip you are spec.ing. I can not stress that enuff, that should have been #1 on the list.
Good luck, I know these are not real technical things, but seem to get lost in the shuffle of designing a plant and can cause headaches on start up and comissioning.

 

[atlas06]

"Service Guys need to know delta P across bundles to calc flows and tonnage"

Drivewizard, (thanks for the input) but how do you calculate tonnage and flow using delta P accross bundles? we typically call for flow measuring device and delta T (not P) to calculate tonnage.
I just wish that service guys actually review the design before it goes out for bid. Indeed field guys can be invaluable in the design, but for some reason, they are nevr involved, and those that do, don't seem interested in doing a real review, may be it's because everything is contracted out.

We engineers need the field guy to tell architect to make more space for the chiller plant (they never listen to engineers). When it comes from the owner, it is always implemented, if not, you get the space that Mr. Architect has assigned to you.

Quark, what do you call a pressure balancing tank?

 

[Drivewizard]

To calc actual flow thru a bundle we have been using this formula. ((Design GPM)X (SQRT(actual delta P/Design deltaP)))
Below is the EXCEL spreadsheet formula out of a log I use to log chillers. E204 is design GPM off of submittals
C205 is entry for actual delta P read off gages
C204 is design Delta P off submittals.

=((E204)*(SQRT(C205/C204)))

Tonnage is just the standard formula but then plug in the actual GPM from above calc. Also have logs to calc KW/ton if chiller display reads out KW or if Starter has a IQData display which reads KW
Hope that helps.

I hear ya about the architects, if they wont listen to you they sure has hell won't listen to us. We don't seem to travel in the same circles anyway. LOL

 

[tbedford]

DriveWizard,

Does your formula provide a true measurment of water flow or is it just a way to verify adequate water flow thru the chiller? I use one gauge at the pump and calculate measured flow head delivered by the pump.
Pump curves are easy to obtain especially over the internet.

 

[quark]

atlas,

It is expansion tank.

The flowrate can be approximately calculated by the formula given by Drivewizard in normal circumstances but a better method is as suggested by tbedford.

Same pressure drops can be manipulated across a chiller by manipulating the inlet and outlet valves (for example, 2barg inlet and 1 barg outlet or 3barg inlet and 2barg outlet), however pump flowrates differ at 3 barg and 2 barg.

I would rather suggest two gauges, one in suction and one in discharge (to check actual dp across the pump and then reading the flowrate from a pump curve).

Flowmeter suggested by Atlas is the best option though.

You can also calculate the flowrate across the HX, if you know the geometry, by the formulae given in Process Heat Transfer by DQ Kern or go to

http://www.processassociates.com

and use their calculators.

I agree with DW on DP switch rather than flap type flow switches. Alternately, you can go for inexpensive thermal flow sensors by IFM or similar types.

 

[krb]

McQuay has some literature on system design.

http://www.mcquay.com/McQuay/Literature/Literature

Happy reading.

 

[krb]

When you click on the link, select application guides, you will find the info there.

 

[daly1]

great stuff above and don't miss the opportunity to right size the pipe. All pipe charts are from the 30's or 40's when pipe was expensive and power was cheap. Now the converse is true. Evaluate the first cost and reduction in friction by oversizing pipe.

I'd also strongly avoid the temptation to size the pipe based on a 13-15F delta T. The purported savings in first cost are paid for by the customer for the life of the building.