Balancing Valves 1

[Dymalica]

What do you like to use? I've read some articles that say balancing valves or AFLV (Automatic Flow limiting valves or Griswold) may not even be needed. My boss (30 year) veteran loves AFLV. But they use a lot of energy to maintain accuracy. I see now many manufacturers have pressure independent balancing valves. Is this the solution to systems with VFDs or buildings with shut off AHUs?

 

[flyrfan]

My opinion is that if it is a variable flow system, balancing valves are not really serving any purpose, other than providing a means for measuring flow. If you have a constant flow system, you absolutely need a balancing valve.

 

[ChrisConley]

I disagree that you don't need balancing valves in a variable flow situation.

The problem with variable flow is low delta T. Coils served by chilled water have very small increases in capacity when 'overflowed', which result in a smaller delta T in the coil. Because the capacity is only slightly higher, the control valve on the coil (which just cares about setpoint, not return water temp) won't help with the situation.

VFD's help, but don't completely fix the problem as pressure fluctuates rapidly in the system, often too fast for the VFD to respond (depending on PID programming and location of the deltaP transducers). The location of the deltaP transducers also becomes relevant in complicated systems with multiple branches serving different occupancies and exposures.

This smaller delta T leads to chillers that appear to be undersized.

On hot water, in condensing boiler applications, delta T helps drive down return water temp, and an automatic flow limiting valve protects that delta T.

 

[flyrfan]

Chris -

James "Burt" Richel published many ASHRAE articles on hydronic systems design and efficiency. He address this very subject in on article. His conclusion, balancing valves on only increase the wire to water efficiency of a variable flow CHW system. You can log in to ASHRAE's website, search his name for articles and make your own conclusions...I did, and support his conclusions.

Flyrfan

 

[ChrisConley]

Wouldn't an increase in 'wire to water' efficiency be a good thing?

 

[flyrfan]

My bad...the additional PD of the balancing valves would decrease the wire to water eff. The key is proper sizing of the CC control valves, which a vast majority of engineers & control contractors lack...but that's another subject.

 

[ChrisConley]

I logged in to ASHRAE, searched for a number of combinations of James "Burt" Richel in the site, and in ASHRAE Journal and didn't find anything.

Possible to get a link to his articles?

 

[flyrfan]

Here's the link to his articles.

http://bookstore.ashrae.biz/journal/journalsearch.php?limit=10&order=high&query=rishel

Also, I've attached the article on balancing valves.

 

[ChrisConley]

Rishel not Richel.

 

[flyrfan]

Your Welcome.

 

[DRWeig]

Here's Gil Avery's opinion, which I share...

http://www.kele.com/Tech/HVAC/BalVFlow.html

Good on y'all,

Goober Dave

 

[flyrfan]

DR....I forgot about Gil Avery. I remember reading that article when it was first published, and, like you, I agree w/ his conclusion.

 

[DRWeig]

Yah --

I did the math once for some simulations -- on a hot morning startup with all control valves wide open, you'll get some over-design flow but the pump curve limits what can happen, so you expend a little more energy for the first hour (normally) until the nearest zones to the plant get into their control range. After that, it's smooth and nobody notices. If you use balancing valves, you're a tad more efficient on hot start-up, but then you have to pay for the pressure drop all day long.

Overall efficiency is best without the balancing valves.

DRWeig

 

[ChasBean1]

I don’t agree with the author’s conclusion and here’s why:

Circuit setters in a two way, variable volume system are for even distribution during a demand condition, so when heating or cooling is needed most, no areas are neglected.

The added pressure drop cost argument carries no weight when the control valve is modulating. If you remove the circuit setter, the control valve will just pinch down more to keep the same flow. Added cost from resistance is a moot point until the control valve is open and flow becomes limited by the circuit setter.

Regarding the article, picture two conditions: a building that uses fin tube radiation on a 10°F day and the same building on a 45°F day (which could represent the low-use condition that the author shows in Figure 2). What’s more critical – the top floors getting their share of heat during the 10°F day or the floor closest to the pumps only getting 58% of design flow on the 45°F day?

Look back at the given pressure drops in Figure 1 and carry the assumption that circuit setters instead remain wide open. The top floor will instead get 58% of its design flow under the demand condition in order to make sure lower floors can achieve 100% flow under the non-demand conditions.

I can’t say I agree with this. The circuit setters are important for this type of system too, especially if there’s diversity in the central system/pumping sizing (i.e., the pumps deliver less than the maximum load).

-CB

 

[quark]

I agree with CB. Both the articles are dealing with purely dynamic pressure loss systems. Things will change when static head comes into picture.

Secondly, Page 48 compares conditions with all coil flow and only 1 coil flow. The author calculates the change in flowrate through coil 1 by considering pressure drop in both the cases (i.e (25/89)<sup>1/2x100 = 53%). However this is not true.

The changed flowrate should be calculated at the pump end i.e (27/100)1/2 = 52% of pump design flowrate considering all AHU coils are running. Since, 52% of total flow can't run through one coil (corresponding to a DP setpoint of 25 ft), flow control becomes difficult.

In practical sense , this is not a problem as the cases discussed are hypothetical.</sup>

 

[Drazen]

From Europian point, the question is what do you consider constant flow system? System that has constant flow on all branches up to each terminal device?!

Normally I use term "constant flow" for system which has constant flow on main branches, but not on "consumer part" of terminal device branches; yet on these branches (that often have thermostatic valves to control terminal devices i.e. flow in that branches is variable all the time) balancing valves are strongly recommended.

It is these days that Danfoss presented what they call automatic balancing valve, which should leave no doubt about effectiveness?!

It will be interesting to see will it totally replace traditional balancing valves in some near future...