when Cv and equivalent length pressure loss calculations yield very different results...

ginahoy

I don't deal with hydronics often so I could use some help. When looking at head loss on valves, I noticed some provide Equivalent Lengths and some provide the Cv. The Taco Heat Motor valve spec lists both. I know that these numbers are estimates and may be sensitive to velocity. I'm looking at the 573 for an application and it lists a Cv of 7.2 and an equivalent length of 130 feet.

For my particular max flow rate (13 gpm, 3.3 fps), the EL yields a pressure drop of about 5 feet, whereas the Cv yields a pressure drop of 7.5 feet. That's a big enough difference that I'm wondering which one is likely to be more accurate?

JStar

The CV rating is how many gallons need to flow through the fitting in order to produce 1psi of pressure drop.

Content deleted to remove false information.

Rich_49

Joe ,

It's early in the morning . Hatt is correct with the square function .

1.8' x 1.8' x 2.34 = 7.507 '

Ginahoy ,

What type tubing or pipe are you using , what temp fluid , what type fluid , what is the actual TEL of the tubing less the valve ?

JStar

Well then, I've been doing this calculation wrong for a long time. Good to know.

Why is the quotient squared?

Rich_49

Here ya go

JStar

That shows the formula but doesn't answer my question.

Zman

Joe,
The relationship between flow and resistance is not linear.
The formula with the quotient squared reflects that relationship.
Does that help?
Carl

JStar

Bu hasn't the square function already been accounted for as the CV rating itself - CV being based on Pressure Per Square Inch.

JStar

Post deleted

JStar

So, by that formula, CV = Flow

1/1 = 1
Sq. root of 1 = 1
Flow rate x 1 = Flow rate
CV = Flow rate

But, that's not accurate.

ginahoy

Said differently, Cv is flow (gpm) through a valve or fitting that would have a pressure drop of 1 psi.

http://bit.ly/1PjqR2z

So if the flow is such that the pressure drop is 1, then Cv does indeed equal the flow rate (that's inherent in the definition of Cv). But you can't then say that Cv = flow rate when ∆P is not 1 psi.

ginahoy

Hatterasguy wrote: "I would certainly use the Cv calculation"

Thanks for that. You're the only person who offered an opinion in response to my question.

The reason I asked is because some valve and fitting specs qualify the Cv as being the "average Cv", implying there's some slop in that method, compared to the equivalent length method. But admittedly I'm probably reading too much into that. In any case, when given Cv and EL, I typically check both ways. Sometimes the results are close but it's not unusual to see differences as large as in my original post (in that case, 50%). I'm hoping someone who understands fluid dynamics at a theoretical level might chime in.

ginahoy

Hatterasguy wrote: "equivalent length is dependent on the material and the the ID of the pipe"

When calculating a valve of fitting pressure drop based on EL, we should always use the reference pressure drop for the piping that's most similar to the fitting material and nominal size, since that's what the mfr's EL calculation was based on. So if I'm using a 3/4" valve or copper fitting with 1" pex, I use the PD for 3/4" copper to calculate the PD for the fitting.

JStar

Hatterasguy said:

JStar said:

So, by that formula, CV = Flow

1/1 = 1
Sq. root of 1 = 1
Flow rate x 1 = Flow rate
CV = Flow rate

But, that's not accurate.

Cv=Flow rate at only one point: When the SG = 1 and there is a pressure drop of 1.

This just happens to be the point at which the manufacturers specify the CV. But, it can be any value depending on the two variables and the flow rate.

Yes, I got it now. I didn't focus on the variability of the delta-p, as it's something that I can't normally test in the field and doesn't occupy a large reserve in my mind. A valve with a CV of 7.2 will flow 7.2 GPM with 1psi delta-p. If you intend to raise the flow, the pressure drop will increase, but then it's not immediately intuitive what that exact pressure drop will be.

The mathematical formula didn't make sense when I tried to solve for multiple variables. Although, my written math skills are only good enough to make me upset that they aren't better.

If:
CV = GPM (Sq.Root of 1/DP)
DP = 1

Then:

CV = 2GPM(Sq.Root of 1/DP)
DP = ???

JStar

I knew that's where I went wrong. But in the OP's example, entering 1.8 or 3.2 or 7.6 doesn't make the equation correct either. I like solving the equation backwards to verify. I'm missing something.

ginahoy

Hatterasguy said:

Do we know this for certain?
I suspect somebody has used the wrong pipe diameter. If it was done properly, the 130' EL's pressure drop would have matched the Cv.

I disagree. Consider an example where the ∆P based on Cv works out to the same ∆P when using the EL method. However, at substantially different flow rates, the ∆P's will diverge. Try it.

So this begs the question, which method tracks actual pressure drop more closely?

On one hand, I sometimes see Cv's caveated as being based on a given velocity such as 4 fps or 8 fps, suggesting that they might not hold for other velocities.

On the other hand, the EL's listed on Taco's heat motor valve spec appear to be rounded to the nearest 10 feet(!)... the smaller valves have an EL of either 10 or 20, so a rounding difference of 5 could cause a 50% error. In my example, since the EL is so large (130), rounding to the nearest 10 feet could only account for a small error.

bob_46

If the blonde with the big rack hadn't sat next to me in algebra I might understand what I know about this stuff.
Gina , the TEL for the Taco 573 is 150 feet. I'm not sure how much difference that makes.
Hatt, your reference to area / diameter is accurate of course but I don't think that is where the referance to the square function comes from.
B&G says "system dynamic head will change APPROXIMATELY as the square of the change in water flow"
Crane Corp. says the value of the exponent of V has been found to vary from about 1.8 to 2.1 in reference to the Darcy equation.
Siggy says 1.75 , he appears to be a stickler for accuracy which is the way to go if you have his software.
B&G admits that the Syzer tends to over estimate head but refers to it as a safety factor. I got my Syzer in 65 and have never been burned by the results it's compact and easy to use.

ginahoy

Bob, that pressure drop table looks like it's from the 1982 heat motor instruction sheet. In any case, here's what's @ Taco website now: http://bit.ly/1J0Fmmx (rev. 2013).

bob_46

I wonder how Taco comes up with two different TELs with the same Cv ?

ginahoy

Exactly.

hot_rod

I think it comes down to how accurately you want to crunch the numbers. An entire chapter in Modern Hydronic Heating is dedicated to Fluid Flow in Pipes, and it is a topic that could fill an entire book, and has

I like how Siggy draws the analogy to electrical devices, Ohms Law to help understand the terms and equation.

In electrical circuits the voltage drop is proportional to the current flowing. This is not the case with fluids and when you look at it graphed to electrical graph is a straight line, whereas the fluid flow shows as a bell curve, similar to some pump or system curves.

To crunch the numbers tightly the formula takes into account the Darcy-Weisbach equation but also Moody friction factors. for hydronic calculations is could be best to assume turbulent flow in smooth pipes.

I suspect not knowing all the specifics to all the calculations, like the pipe and fittings, etc as far as resistance, some formulas have more wiggle room built into them.

Same as the discussion on the Sizer wheel compared to the long math approach to the calculation.

This is always the challenge for an author, trainer or the folks writing the tech formulas into their design manuals. How deep can you drill down into the numbers before you lose the audience. It needs to be presented in an understandable formula or math equation or designers and installers throw up their hands in confusion and ignore the whole concept.

The toughest group to teach in front of is a mix of experienced P.E.s and out in the field installers and designers. They are often at opposite ends of the spectrum regarding expectations.

jonny88

This is why I enjoy Dans books he can write so the wrench turners can understand.And I know that is through my own fault for not studying more but when it gets to advanced i am reading but brain has switched off.

ginahoy

Hatt, boy do I share your pain regarding some manufacturer's attitude toward product data. My original example, with 50% difference between Cv and EL, seems likely to be the result of sloppy work.

I'm attaching an excerpt from Uponor's ProPex design manual. Based on the introductory text, it's clear that whoever wrote this cares about accuracy. Now notice that on the tables that follow, Cv is expressed as an average value, whereas EL has no such qualification. I've seen that in other manufacturer's tables as well.

We know from piping pressure loss tables that PD varies with water temperature. For example, the PD at 40F may be as much as 50% greater than at 180F. The Cv therefore must assume not only a reference velocity, but a reference temperature as well. On the other hand, when using EL, temperature is accounted for in the reference piping pressure loss.

I realize all of these are just estimates, but what I'm seeing suggests that when both Cv's and EL's are provided, EL may be the better estimate (notwithstanding errors in the tables).

ginahoy

Hatterasguy said:

The calculation for EL must be done SUBSEQUENT to the Cv determination...

I disagree. Both are derived from the empirical data, but one does not follow the other.

So let's start with what we know... To be reasonably accurate, we know that Cv and EL must be derived from actual measurement data. We also know that PD (friction loss) is strongly dependent on water temperature (see attached reference). And since the EL of a fitting is dimensionless (it's simply a ratio), then if we use a piping PD reference that's broken out by T, then an EL based on that reference will automatically account for T.

(BTW, I knew that PD depends on temperature, but until I studied the attached tables this evening, I never realized how strongly that dependency was.)

On the other hand, the Cv calculation (F√(SG/∆P) does not account for temperature*, so a Cv is only valid at the temperature at which the PD was measured. No one publishes Cv's broken out by T, but rather, Cv's are typically qualified as being an average. So there's no way for the designer to account for T when using Cv to estimate fitting losses.

Make sense?

* SG is slightly affected by T, but it's virtually a constant in the range we're interested in

hot_rod

Also know there are no Cv police. A manufacturer can pretty much claim whatever they want for that number. The number can be derived with CFD computer fluid dynamics, very expensive complicated software, or by testing. And like any software program, garbage in, garbage out.

There is a specific test method defined for testing Cv with venturi tubes and actual flow measurements. It will develop a curve when plotted, unless someone try to plot it on a logarithmic graph paper. The actual flow test, I feel, is a better method.

So the Cv number published for a component could be accurate or way out in left field. We know this at Caleffi because we test a lot of our competitors products. We have one of the finest labs and a staff that is dedicated to testing and establishing data for flow, air removal, dirt, removal,pressure ratings, etc. We can break, stress, salt spray, pretty much any condition that the component may encounter from use, or misuse.

Test equipment needs to be lab quality, certified and re-calibrated and re-certified, constantly. Different teams will run the tests to eliminate human error, it is a big commitment in time and money.

Also more and more hydronic manufacturers buy products and re-label them, and they depend on the specs from the actual manufacturer, and do not test or confirm the data. Nothing wrong with partnering, as long as the numbers are confirmed..

So crunching numbers to the th degree really doesn't tell you much if the data you entered is wrong or pulled from someones a--.

hot_rod

My curiosity regrading gaming the Cv value would be the reason for it. The customers are depending on that value and gaming it would serve no purpose other than to try to make the component show less resistance than it actually has. Or the manufacturer is simply incompetent and cannot perform a proper test.

I don't know a reputable company would purposely "game" a rating number. Then again you know the old sale saying "don't let the truth get in the way of a sale"

I suspect sometimes numbers are created to match or exceed the competitors?

I may not trust the numbers or ratings posted on some of the product from PRC, would you?. China has a knack for fudging the facts.

SWEI

Hatterasguy said:

My curiosity regrading gaming the Cv value would be the reason for it. The customers are depending on that value and gaming it would serve no purpose other than to try to make the component show less resistance than it actually has.

There is a tendency in the residential hydronics market to think bigger is better when it comes to Cv. This is not the case for mixing, modulating, or any kind of proportional control.

ginahoy

Hatterasguy said:

...you'd still need to start with the Cv to get a known pressure drop at a specific flow rate and temperature to then go to an EL table.

Not sure I follow. The testing lab doesn't need to calculate Cv to develop a set of EL factors, nor does the designer need to refer to Cv to use the EL factors. The PD across the fitting or valve is directly measured at the reference flow rate and temperature (which must of course be measured). This PD can then be compared (as a ratio) with the PD/ft for the appropriate pipe, referenced at the same flow & temperature.

Yes, two empirical measurements are required to develop EL, but when the piping PD table is properly broken out by temperature, then EL accounts for temperature, which I now understand has a HUGE impact on the friction rate -- as much as 50% over the range we deal with.

I'll grant you this... if the piping PD tables are no good (or not broken out by T), then the EL would be wrong (as well as all the piping PD's you accounted for in the loop).

This issue of manufacturer (or lab) incompetence (or worse) is of course important, but my objective here was simply to gain a better understanding of what these numbers mean, ideally. I'm a graduate degreed engineer but this stuff isn't something I was exposed to in school. This discussion has been very helpful.

andreea

Hot rod mentioned that the problem can be solved with computational fluid dynamics software (CFD) which is very expensive. He's right about the traditional software like Ansys or SolidWorks. But if you want to try CFD, you could do it with this Cloud-based tool - SimScale. They have a free account for CFD analysis.