Pressure losses and Recirc questions for Chilled Water Closed Loop System

[jtimmer]

Hi all, I am in the process of designing a chilled water system and have gotten to the point of running calculations to figure out the pump head I need. The system is a closed loop chilled water system that has a demand of 480 GPM. There are two chillers and two pumps, with one chiller and pump active and one on standby. The pumps and chillers are located on the roof (top) of the system. My first question is about the pressure loss through the system. We have 10 air handeling units spread throughout the system with two on the same level as the pump and the rest spread across the two floors below. When I am figuring out the loss through all the components I will want to do a sum of the entire system (supply and return) including the air handlers and the chillers is that right? Since that would be the worst case scenario with every air handler unit demanding water and the pump having to supply water to the whole system?

Once I have the losses throught the system I don't have to worry about elevation changes since it is a closed loop system right?

Secondly I am adding a recirculation line and orifice for the pumps and for the chiller so that they don't run without any flow. Does the AC plant orifice go at the bottom of the system so that if all the air handlers are closed and the system is running there will be the minimum required flow through the orifice back to the plant?

Also, is it customary to add in the recirc line flow rates to the system flow rate? I ask this because I am assuming the pump recirc is always open which means that this amount of GPM will be "lost" and sent back to the inlet of the pump before it goes to the system. So the pump will need to pump the entire system GPM + the Recirc line of the pump. Does that sound right?

Finally, what kind of factors are usually used for saftey in terms of pump size and Chiller size. The chiller we are looking at are rated for 522 GPM at our ambient temperature and design conditions which gives us a cushion of about 50 GPM. Is this too low? Anything more and we would need the next chiller size up.

Thanks

Josh

 

[LittleInch]

Wow, that's some post, but it's always helpful if you can add a schematic or flow diagram, but never mind, let's have a go at your issues, which you could have numbered....

1) Yes, you need to work out your max flow using all of the AHUs and chiller as this is your max flow case when e.g. you turn the system on on a warm day.
2) If your system is a closed pressurised system then yes you are correct, you can ignore head and concentrate on friction loss, but always remeber what the actual presure is going to be at the lowest point - roughly 10m = 1 bar PLUS what ever you pressuirse the system to at the top - prob 1 -2 barg
3) Recirculation. You have a number of options rnaging from the simple to the slighly more complex.
Simple - do as you say, but place the re-circ at the end of the system connecting between the flow and return. The problem is that you min flow value is likely to be some 15 - 20% of your max flow rate as you're only running with one unit (most pump vendors hate giving you min continuos flow, but if you push them they will). You don't say how yopur AHUs modulate temperature - is this by modulating the air flow though it, modulatig the chilled water or simple on/off water flow?

part complex - fit a flow meter - you can get good simple strap on ultasonic meters now which will do your job fine - and then a simple PLC which below a certain flow rate, say 20% of max, opens a valve with your orifice to take it to say 35%. Once flow goes past 40% close the valve..

Proper system - fit a flow meter as above and run it it a simple PLC running a PID controller and a control valve between your flow and return. Make sure your control valve at 100% open only lets through 20% of flow. Make the set point 20% of flow. As flow decreases towards 20% the valve opens to maintain 20%. As it increases due to more water demand, the valve will close until all the water goes to the AHUs.

Simple will mean your total flow is now 120% of your AHU demand as 20% will flow via your bypass all the time - maybe a bit less due to system losses, but not much.

Part complex - better than simple, but you will have higher pumping costs as the min flow actually needs to be higher than your min. However your max flow is only the amount needed for the AHUs as once you go past 40% the valve closes.

Proper system - Better for your electricity costs, but a little more complex and needs a modualting control valve - still pretty simple and most PLCs will handle this without a problem.

FOS - How do you expect me to judge? You or someone else have worked out what you need for your design conditions and I have no idea what they are or what allowances you have already included. Also depends if your clients / customers are happy to pay a bit more to be always cool on the 10 days a year it gets really hot or not. Your call I'm afraid.

A system like this could be v diffiuclt to balance as if the flow does change all the time due to water flow stopping and starting through the AHUs and hence unless you have flow meters and valves controlling on each AHU, then your system will only be balanced for the initial time when you set the valves and flow using some sort of portable flow meter during commisisoning. See my original point - how is your system set up - constant water flow and variable air flow or variable water flow?? It will make a huge difference to the operaiton of your system and compfort of your clients.

I'm sure some HVAC guys will come and put you straight as it's a common issue, but equally I've worked in many buildings where clearly no one gave it much thought and one room was like a fridge and the others sweltering.

Don't skimp on controls.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[jtimmer]

LittleInch thanks for the response. The system is using on/off solenoid valves. So when the thermostat in a room wants heat it turns on the valve and the AHU will get max flow until the thermostat turns it off. Also, the chilled water is set up using constant flow valves for each AHU. So each AHU has been sized to a certain GPM to cool the space it serves and the appropriate constant flow valve has been ordered for each AHU. So when the solenoid is turned on only the flow that the coil is designed for will go through the flow valve (ideally).

Also, on your point (2) from above, this is a closed and presurized system. So I understand the height of the water gives me my pressure at the lowest point, but the pressure at the top of the system comes from the expansion tank right? and this will be determined by the required pump suction pressure?

Thanks again

 

[LittleInch]

Constant flow valve?? For my education how does that work?? I've seen some constant pressure differential and back flow valves, but flow control normally needs a feedback loop from some sort of flow measurement device.

Pressure at the top will need to be as a minimum what the pump wants as min suction presusre yes, but normally you want ot aim for 1 - 2 bar to avoid any chance of either cavitation or low pressure as the water temperature falls. Don't take this too low - you'll regret it.

Glad to be of help

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[jtimmer]

The constant flow fittings we use work by a diaphragm that as pressure increases the diaphragm is compressed and closes off the flow area effectively regulating the flow as pressure changes. They have a 10% tolerance so they aren't incredibly accurate but they are mostly maintenance free.

 

[LittleInch]

Thought so - trust me they won't get you great flow control with 10 units all doing their own thing over two floors. Your pressure will change by a lot more than 10%. I'll have the room closes to the chiller please.... ;-)

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[jtimmer]

I agree, but this is what we were told they wanted so its what we will give them....

Thanks for your help

 

[LittleInch]

The low cost of UT flow meters and PLC controllers now adays should make you able to design something a lot better. Why don't you give them something better if you agree with me? You should never be satisfied with just the minimum. especially as it will probably keep you in the lower rated chiller size.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[jtimmer]

With time constraints we can't go out at this point and do a new control system with valves. I am going to try and do the control valve with the flow meter and PLC. Do you have a recommendation for this type of valve and controller? We have simple Ventui flow meters in the system already and may be able to connect these to a pressure transducer that can then control the valve. Would this kind of PLC be something that is commercially available or would we have to buy one and then program it?

 

[LittleInch]

To be honest I don't know that level of info, but an HVAC control guy should know. Its a simple system so I would have thought by now simple packages exist, but tide need to go looking. I design much bigger systems but the design I'd the same.

God luck .

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way