Automatic Balancing Valves in Chilled Water Header System 1

[keenfirengineer]

Chilled water is supplied to my building from a district cooling plant. I have heat exchangers forming a primary and secondary loop, with a header piping system. I have designed modulating valves at the heat exchangers, installed on the primary side, with temperature sensors measuring the temperature at the secondary chilled water return piping. These sensors will control the modulating valves on the primary side, opening when the return chilled water temperature is higher and closing when the return temperature is lower than the setpoint. I have also installed automatic balancing valves at the heat exchangers.

My concern is that during part-load conditions, the pumps will run with reduced flow. From my findings, automatic balancing valves are set for 100% flow. Therefore, during part-load conditions, the balancing valves may not be effective, and the first heat exchanger in the chilled water supply piping may receive more flow.

For example, if I have 4 heat exchangers at 1000RT each and the building load is 1600RT, I would only need to run 2 pumps supplying 2 heat exchangers at 800RT each. Will the flow through the 2 heat exchangers be equal?

Is my understanding of automatic balancing valves correct? If so, how can I balance the flow through each heat exchanger to ensure equal flow during part-load conditions? Installing a PIBCV would be too costly, though it seems to be the easiest solution.

 

[LittleInch]

Can you sketch this out please?

But I think you're right. It all depends on what flow is set as the controlling flow for the balancing valves.[pre][/pre]

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[HVAC-Novice]

Use Pressure Independent Control Valves (PICV).

 

[3DDave]

"how can I balance the flow through each heat exchanger to ensure equal flow during part-load conditions?"

That is not an easy task.

PICVs, which I had to look up (interesting video here:

https://www.youtube.com/watch?v=nAM5xU_KfzU

work to maintain a set flow through the valve but don't appear to work in concert automatically.

However, if there is a control system attached that modulates all the Pressure Independent Control Valves via communication/electronically that would work. The next layer of this onion is deciding what that flow rate should be and telling that to the system controller.

I suspect that will vary among the different PICV manufacturers.

As to the part: "during part-load conditions, the balancing valves may not be effective, and the first heat exchanger in the chilled water supply piping may receive more flow." depends on the vertical separation. If I understand correctly, if the overall flow drops, it is dropping pressure along with it and all the balance valves will open fully, making the distribution a matter of which one is lowest in the building or which has the lowest pressure losses due to flow / a combination of the two factors.

 

[HVAC-Novice]

During part-load you don't have equal flow since each coil will have a different part-load. If the each coil even has the same design flow to begin with.

PICV work automatically.... the flow depends on valve position. It is a mechanical valve for the simpler ones. the electronic ones use a flowmeter. You get exactly the flow the controller calls for.

Yes, you need a controller. That is how a control valve works. If you don't have a controller and the PICV is just open, it is an automatic pressure-independent balancing valve.

What flowrate and PICV selection depends on the design flow.

https://www.belimo.com/ch/en_GB/pro...mentation/pressure-independent-control-valves

 

[LittleInch]

If I understand this correctly, all the HX on the secondary loop are not individually controlled for flow or temp?

You might be better off running the secondary loop at full rate but use three way valves to control flow through each HX?

Or have the flow control signal sent to all the balancing valves.

A drawing would help understand your control system though.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Rputvin]

I was going to suggest a similar approach to what LI describes

Do not vary flow on the chilled water side, use 3-way valves on the heat exchanger. Choose porting to mimic the pressure drop of the heat exchanger through the bypass port of the valve. The whole system will operate with stable flow and pressure, and you'll just have heat exchangers in bypass when the load isn't needed. One control point for the 3-way valve, and one manually set balancing valve that doesn't need fiddling with unless you're rebalancing the chiller circuit or performing maintenance.

Whoever is in charge of that chiller will appreciate steady flow/pressure requirements instead of a constantly dynamic load.

 

[btrueblood]

"You might be better off running the secondary loop at full rate but use three way valves to control flow through each HX?"

Then you are mixing the bypassed flow with chilled flow and losing that cooling? Seems inefficient. PICVs is the right way to go.

 

[LittleInch]

You don't lose any cooling, only the return temperature falls.

PICVs in this instance need a variable control signal otherwise if they only start to control, i.e. close, when the flow limit is reached. So one unit could be at its max of 1000RT and the other at 600 instead of both at 800.

But we need to see how the control logic is set up and what pumps are where. Diagrams help a lot here.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[keenfirengineer]

Hi everyone, apologies for my late response. I've attached the schematic of my plant room for your better understanding. I hope this provides clarity.

I've designed the system to monitor the temperature and pressure at each heat exchanger.

Upon further review, I found that automatic balancing valves cannot be controlled via a centralized system, such as the Chiller Management System or Building Automation System.

I believe PICVs are the best solution, but the cost becomes exceptionally high with larger sizes.

Therefore, I'm looking for a more cost-effective alternative that works well for part-load conditions, if one is available.

 

[HVAC-Novice]

So the left side PFHE are your heat exchangers? which way is the blue flow? does it flow from right to left? Is the blue coming from chiller?

and why does the red flow turn turquois downstream of the pumps? Is that basically the water you want to cool?

what is the control sequence or what is the goal? Do you open up in the HX after the other until yo reach setpoint?

Are you looking into replacing both valves (blue and turquois). Ideally all on/off valves get eliminated to avoid low dT.

 

[keenfirengineer]

Resend with revised schematic for clarity.

@HVAC-Novice - the red flow downstream of the pump connects to the supply line as by-pass during very low load conditions.

Yes the Motorized Valve opens once the pressure in the system drops as the valves at the air side equipment opens.

Motorized (on/off) valves are required to channel the flow into the required Heat Exchangers.

Sorry I have not developed the control logic schematic.

 

[HVAC-Novice]

Unrelated to your original question, but VFD pumps don't need balancing valves.

On the left side you show BV and ABV in series. I assume you mean that to be automatic and manual balancing valves. If you use BV, you only need one. And PICV would eliminate both BV.

the valves on each side of the HX need to work in tandem. if isolated, both sides should be closed. You could also modulate both sides.

PICV would work even if you operate them in on/off mode. You still want to not exceed the design flow.

how is it decided how much flow you need on the secondary side? That would determine how many HX need to be open. You could use modulating actuators on secondary size and modulate open one HX after the next to get your required flowrate. the modulating valve on chiller plant size would be modulated to maintain each HX discharge temperature.

On cost: one PICV may cost less than a control valve = BV + space and piping tp install 2 instead of 2 devices. and you don't need to hire a balancer. if you add that up, I don't think ICV cost more overall. And you get the benefit to be able to monitor the flow and have accurate flow.

 

[LittleInch]

So do you actually have two separate systems here?

As far as I can see from this the "primary" supply just acts to cool the secondary separate loop?
So District cooling plant water is kept separate from your "secondary" cooling water?

What does PFHE mean?

But this is why a diagram is so useful as only now can I understand what you mean, but then maybe I'm slow on the uptake....

MV usually mean Motorised valve (on/off) but this looks like a control valve ( variable flow)?
BV I assume means Block Valve and not Balancing Valve?
Given that the primary flow into each HX is set by the temperature of the secondary water leaving the HX why are you bothered about balancing flows? So long as the flow doesn't go above max or below min rating of the HX, the flow of colder water on the primary side will match the flow of the outgoing secondary colder water (dark blue) and maintain a fixed temperature into your secondary system. If one is operating at half the flow of the other so what?

Do your ABVs as you term them actually mean PICV, i.e. they start to close once your flow rate goes past a certain flow?



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[BronYrAur]

So will all 4 heat exchangers be active all the time or will you be staging them? Your question seems to suggest that they will all be open, and you are worried about the closest one getting too much (or all) of the flow. If the load is small, you should only need one heat exchanger. There are various ways to stage them.

 

[HVAC-Novice]

You will want to stage them based on flow requirement to keep the fluid flow in each of the HX relatively high (for turbulent flow and better heat transfer as opposed to laminar flow). Would be good to know what flow can be. Can it be 20-100% of design flow, or could flow be 5% of design flow? the HX selection software will tell you minimum flows they expect.

 

[georgeverghese]

I've been looking at this sketch and responses so far, and it is all rather confusing.
For no further info, it looks like the piping arrangement at the PFHEs' is not correct if you want more or less equal flow of secondary loop fluid to the 4 units. The secondary return header feed to these PFHEs' should be from the bottom of the page starting with first feed to unit 4, and the blind flange should be at the top after the last tee off to unit 1. Piping arrangement on primary service to the 4 PFHEs' and the feed and exit headers at the pumps is correct.