DUCTINGS LOSSES 4

[hvacers]

HI,

Im an energy engineer of an HVAC company, i have a problem with one of my AHUs. I conducted Duct traverse to one of my ahus and measured the air flow of the diffusers to each area served. i measured 13000 cfm at the main supply ducting which just branched out of the AHU. While i just got a total 5000 cfm for measuring all of the air flow of each diffuser supplied by this ahu. the discrepancy is really large, can this be all accounted to loss??? or do i have i problem somewhere to my ductings?


JLV

 

[MintJulep]

i measured 13000 cfm at the main supply ducting which just branched out of the AHU

What is the duct configuration here? How far from the blower discharge? Transitions? Turns? How was your traverse done?

It's possible that there could be a very large measurement error here.

Or you might have one really big leak in the duct system, or a lot of small leaks.

 

[jvithayapun]

could be leakage somewhere or static pressure from ductwork. just by looking at random fan curves, a 1 inch increase in static pressure can cause an 8000 cfm decrease.

 

[MintJulep]

Uhmm....

There's this little thing called conservation of mass.

 

[hvacers]

the AHU is situated at the 4th floor of the building while the areas being served are located at the 1st, 2nd and third floor of the building.

the duct traverse was conducted at the main ducting from the air suppy of the AHU, where we took about 40 points spreaded all over the rectangular area of the ducting and took the average of all the 40 readings...

is there any rule of thumb about the conservation of the airflow?? example only 90% of the supply air will be delivered to the diffusers???

thanks for sharing ideas

 

[ChasBean1]

100% of what you measure in a traverse should be accounted for at diffusers; duct leakage should usually be about 5% or less. Instrument error is a factor but not that much. Are there VAV boxes and were duct traverses and diffuser measurements done at the same time under the exact same conditions (e.g., no concern about unoccupied control or VAV airflow)?

 

[BM4BRANES]

From a simple/practicle approach, what is the rated capacity of the A/H,how big is the area served?The hood measurements would be more difficult to goof up, agreed? what is/was your objective originally? How critical is the project?

 

[berkshire]

Have you got a future expansion branch duct that is not capped off between the floors somewhere ?
I would suggest a physical inspection of the ducting and re sealing where needed.
B.E.

 

[SAK9]

Analyse the duct traverse readings closely.If the readings are too far apart,the location for the traverse is not good enough and the airflow reading can be misleading.If there is not a good spot for duct traverse ,I would just sum up the flow hood readings of the diffusers to arrive at the AHU flow(allow for a 5% leakage as Chasebean suggested).Also check motor current draw and verify on the fan curve to get an idea aboout the possible airflow.Also check reurn air temperature against room temperature(if RA temperature lower than room temp it shows leakage into ceiling plenum)

 

[mauricestoker]

I would suggest looking at a standard traverse method, as velocity pressure across a cross section of duct is not linear. A typical method is given by the ACGIH, 23rd edition. A 10 or 20 point traverse, with second traverse perpendicular to first, should be sufficient. For the largest velocity measure pressure read, divide by 10. At least 75% of the traverse readings would have to be greater than Vp(max) in order to have an accurate reading. TABB or NEBB should have a similar procedure.

With the losses you are talking about, I would consider it a serious energy issue. Either traverse data was bad or instrument out of calibration, or you have a gaping hole. Spot manometer/dP readings should help localize, but at that flow you should be able to hear it. I would recommend physically observing every section of duct if the readings are correct.

 

[RossABQ]

It sounds like this "main duct" is in a vertical shaft? Can you take measurements at the branches off the main for each floor, starting at the top? I think you'll find your losses between floors, likely due to duct separation or a split seam.

 

[hvacers]

i performed the duct traverse reading myself, to validate the first reading of my technicians... i got an average of 4966CFM. the readings from the diffusers are 4179CFM. the discrepancy is around 15% which seems to be realistic than the previous readings.

but the problem is i dont know which traverse reading is correct. what my technician did is, he did not plug the static pressure(-) to the device. he only connected the total pressure(+) to the device.

while

I connected both the total(+) and Static(-) pressure to the device.


which setup is correct?

 

[mauricestoker]

You need both total and static in order to calculate velocity pressure. Velocity pressure is then converted to velocity and multiplied by duct area. If the tech measured total pressure then used the conversion table or calc, then a much larger flow would be expected.

 

[berkshire]

Ok so now your loss is just under 800 cfm compared to 8000 cfm in your first post.
Maurice is correct you need both total( pitot ) and static at the duct. If your tech simply measured pitot pressure in the duct and room static, your reading is going to be way too high.
B.E.

 

[hvacers]

thanks to everyone who shared ideas...

 

[briand2]

I'm sorry for stating the obvious, and I wouldn't wish to offend anyone, but, assuming your meter reads differential pressure rather than velocity, did you (correctly) work out the velocity at each measurement point and average the calculated velocities rather than (incorrectly) average the raw differential pressure readings?

Brian

 

[mauricestoker]

That would be at about the same step where you determine whether 75% of all Vp's are greater than 0.1*Vp(max).