Static pressure to measure airflow? 3

[chims]

Has anyone ever used static pressure as a method of determining airflow?

Example: Seal the outside air damper up tight, so the return will equal the supply airflow. Measure the static pressure in the return air duct at the inlet of the AHU. Then use the fan laws Q1/Q2=square root of (P1/P2). If you need a specific volume of outside air then you can determine what your return air needs to be in order to get the require outside air. Then you control the static in the return duct to the new static pressure determined by the fan laws. It seems pretty simple to me and cheaper than airflow measuring stations, but I'm wondering if there other caveats to be aware of.

 

[MintJulep]

You mean caveats like it won't work?

Fan laws can be used to predict the behavior of a specific fan at other conditions if its performance is known at a condition, or to prediect the performance of a geometrically similar fan subject to the same conditions as a known fan.

They cannot be used to predict the performance of a system.

To do what you want to do, you need to know the system curve of each portion of your system, and the fan curve.

 

[chims]

Thanks for the input, Mint Julep. Airflow varies with the square root of the pressure. If we know three of the variables we can calculate the fourth. By knowing the air flow and static pressure at a particular point in the system(i.e. the return airflow and static pressure) I can now calculate the pressure required for the new airflow. The O/A and R/A dampers can then be adjusted to obtain the static pressure as calculated to obtain the new airflow. This should give the require O/A. (Refer to ASHRAE Journal article May 2001, page 20). It seems simple, I'm just wondering if anyone else has tried it.

I don't quite understand your comment on knowing the system curve of each portion of your system. At a constant operating condition, the system only has one system curve.

 

[MintJulep]

Chims,

You question appears to be asking about the control of a damper, which implies a non-constant operating condition.

You are correct that the airflow of the fan varies with the square root of the pressure across the fan.

However, the pressure that the fan is working against is a function of the system that it is connected to. And the pressure (resistance) that the system produces is a function of the flow rate.

If you close the outide air damper tight you have zero flow in the portion of the system from the damper to the point where the O/A and return mix. However by sealing the O/A damper you have changed the flow in the return and supply portions of the sytem, and hence have changed the pressure drop in those sections.

To do what you want to do, you must simultaniously solve the system equations for each portion of the duct system and the fan curve. Not impossible, but not easy.

 

[ChasBean1]

Chims, from what I understand so far, I tend to agree with Mint. The fan "laws" should still be postulates. They very vaguely represent actual conditions. A couple quick questions:

1. Is there a return fan & relief dampering?
2. Is it a VAV system?
3. What are the terminal devices in the supply and return duct (e.g., are flows controlled at terminal points)?

Thanks, CB

 

[lilliput1]

Then all goes out the window when on outdoor air economizer controls

 

[chims]

Mint,
I don't think I've explained the conditions clearly enough. Given a hypothetical fan delivering 10,000 cfm at a specific pressure. No return fan and the O/A duct is sealed tight. Therefore 10,000 cfm in the return duct at a specific pressure. I need 1000 cfm of O/A and 9000 cfm of R/A. Supply is still 10,000. I reduce the pressure in the R/A duct to the calculated pressure by closing the R/A damper and assume then that I have the 9000 cfm, without actually having to measure it. I can therefore assume that I have 1000 cfm of O/A. Meanwhile, the supply is still delivering the 10,000. This is just a hypothetical situation. My actual condition is different, but I'm trying to understand the simple situation before making it more complicated.

Chas,
1. My system has does not have a return fan and relief is through a remote exhaust fan.
2. The system is VAV.
3. There are VAV boxes as the terminal devices.

 

[lilliput1]

Without return/relief or exhaust fan how do you prevent the bulding (or process) from pressurizing and not letting in the additional OA? Do you have ventilation requirement for minimum OA? If there will be no pressurization (open system) you can modulate the RA damper close while modulating the OA damper open providing a manual balancing damper is installed at the OA duct & balanced for the minimum OA CFM position. You may get noisy flow in the OA & water entrainment if nor properly sized. For this case (no pressurization effect) you don't neet to set up SP control. Just balance for the minimum OA case & accept flows when O & return dampers are modulated in opposition to each other. It is in critical VAV systems when air flow measurment and return fan tracking of spply fan with fixed offset (equal to minimum OA CFM) is used.

 

[MintJulep]

Chims.

It doesn't work that way. You said it yourself. Hypothetical fan delivers 10,000 cfm at a specific pressure. At a different pressure the cfm will be different.

The pressure that the fan is working against is the resistance provided by the duct system.

The pressrue drop of the duct (more accurately, each part of the duct system) is a function of how much air is going through it.

To use your latest example:

O/A sealed. Flow in supply = flow in return = 10,000 cfm.

Now, you close a damper in the R/A duct. This INCREASES the resistance of the system. If the O/A damper is still sealed the flow of the fan will change according to its FAN CURVE, not the formula Q1/Q2=square root of (P1/P2). That formula applies only under a limited set of specific circumstances. Think about it. If that formula described the behaviour of every fan, then all fans would behaive identically. That isn't the case. Backwards curved fans, forwards curve fans, inclined blade fans, propeller fans all behaive very differently as pressure changes.

I know this because I was burnt on this subject a while ago on this forum. I have since educated myself.

 

[chims]

Mint,

I don't disagree with your analysis, but I am still supposing that the pressure in the system is constant to guarantee the delivery of 10,000 cfm. This could be by dampers or VAV boxes or whatever it takes. Yes, the pressures upstream of the fan have changed, but I can make changes elsewhere to ensure my static is the same. The bottom line is, that I'm going to deliver 10,000 cfm, come hell or high water. That is, I'm still on the same system curve.

So, having said that, does my static pressure analysis of determining R/A and O/A now hold true?

 

[MintJulep]

Ok, in that case then yes I think so.

But its no longer a VAV system then is it?

 

[chims]

Correct. That's where the challenge lies and maybe where this whole thing falls apart. That's why I was wondering if anyone has ever done it before.

 

[ChasBean1]

Chims,

I like it conceptually, at the unit. With the fan maintaining say, 2" in the supply duct, OA damper closed, and return duct pressure at 1", close the return and open the OA incrementally to achieve a return pressure of:

0.81", correlating with 9000 cfm
0.64", 8000 cfm
0.49", 7000 cfm
0.36", 6000 cfm
0.25", 5000 cfm
etc., probably fairly accurate, not perfect..

Like you said, though, attach a VAV system to it and all goes out the window. The system curve would have to be constant for this to work, I think.

 

[SlipMatWax]

TSI, a manufacturer of fume hood controls use this principle to maintain equal face velocity across an adjustable sash. The static pressure within the hood is always monitored and controls a damper to maintain setpoint.

Ckeck out the following pdf, it may be helpful.

http://www.tsi.com/hvac/lab/downloads/LabBro2980178RevA10_03WebSecure.pdf

 

[MintJulep]

Maintaining a constant delta-p across a variable orifice will most assuredly NOT maintain a constant face velocity.

Face velocity is proportional to the square root of the delta-p.

The TSI fume hood controlers referenced by SlipMatWax both mention velocity measurement.

 

[MintJulep]

Chas,

Think about that constant system curve statement a bit longer please.

For any fixed physical geometry, flow is proportional to the square root of the pressure drop through the system and inversely proportional to the "resistance" of the system. (The resistance is a combination of factors descriptive of the physical geometry of the system.)

If the geometry does not change, if you increase the flow, the pressure drop must also increase. No way around that.

If you add the abilty to change the geometry then yes, it could be possible to design a system that maintains a constant pressure drop as the flow is changed, but the control would by necessity have to alter both the geometry and the fan simultaniously. Such a control would be inherently unstable, and prone to control system induced oscillations. I see no advantage in attempting to devise such a beast.

 

[paulkeating]

Well, I know now that you guys are all engineers in the truest sense that you never seemed to get bored of the "puzzle" nor angered by the criticims of the other writers.

Me, I got bored of the tech stuff (mustn't be a real engineer, me) but enjoyed the human interaction!!! (anyone fancy a pint?)

 

[wilg]

A pint or two would be great right about now.

 

[ChasBean1]

Mint, sorry about the late reply as I didn't see your post. Point taken about the change in the system geometry. However, we were talking about a constant supply air volume (yes, I realize it’s a VAV system, but bear with me).

Terminal boxes in the system control air volume. If, for example, we shut the return (mixing) damper and open the OA damper within the AHU, the supply fan will have less resistance at its inlet and tend to deliver more air. Because flow-controlled supply terminal boxes prohibit further air delivery, the fan total static remains constant, although at a higher discharge pressure and less negative suction. Initially we were moving 10,000 cfm at a given fan RPM, after the damper modulation we’re still moving 10,000 cfm at the same fan RPM…

Looking at Chim’s analysis and temporarily forgetting the fact that the fan might have a speed drive or that terminal devices will vary flow, I think it’s feasible.

For the purpose of this exercise, think about only the return damper and the piece of duct upstream where the pressure sensor is located. If we’re measuring -1 in. w.c. with a known 10,000 cfm flow and we slowly close the return damper until return duct pressure becomes -0.5 in., I would estimate the return volume to have lowered to a ballpark of 7,071 cfm. Continuing to close return dampers to a duct pressure of -0.25 in., it should correlate roughly with 5,000 cfm.

Does this make sense? It won’t exactly define the behavior, but it should roughly represent a standard flow versus pressure relationship.

Vr, CB

 

[MintJulep]

Chas,

I think your case applies only to a very simple system consisting of one inlet, one fan and one outlet. It is correct that if you close the outlet damper and open the inlet damper to keep the total pressure drop constant then the flow will remain constant.

However, if the complexity of the system is increased things start getting more involved. Add one more outlet to the system.

Restrict the damper at outlet 1, but leave the damper at outlet 2 unchanged. This will increase the resistance in branch 1 and the entire discharge side of the fan. The fan will respond along its curve and the cfm delivery will change. As a result, the flow at BOTH outlet terminals will change, even though only one damper was moved.

Now open the fan inlet damper to return the total pressure drop to the original condition. The total flow will increase back to the original value, but the ratio of flow at outlet1utlet2 will be different than the original ratio.

If the fan flow is held constant, that air must go someplace. If you increase the resistance in one branch the air will flow down a branch with fewer restrictions in proportion to the system curve of each branch. This is the basis of the equal pressure loss method of sizing ductwork.