Dear daovinhhien,

The calculation of the pressure drop of a bag filter system is executed according the general pressure drop formula:

dp=1/2 * resistance factor * density * velocity^2

The total pressure drop is the summarization of all partial pressure drops.

-dynamic pressure drops

-filter pressure drop

-channel pressure drops

-fan exhaust pressure drop

The pressure loss of filter bag is depending on:

-filter resistance factor under operational conditions (dust load)

-density (absolute pressure, temperature)

-filter load to the square (filter velocity^2)

The filter resistance factor is a function of the degree of contamination (or cleanliness) of the filter fabric, which makes it very difficult to assess.

Therefore, the resistance factor needs to be known from field measurements on existing installations for the same material.

Channel pressure drops can be calculated, using the Fanning Factor (Darcy-Weisbach, Swamee-Jain equation)

These pressure drop calculations have to be done for the required air flow.

The required airflow and the calculated pressure drop form the operation point of the required fan.

Or, calculate the pressure drops for a range of air flows and combine the derived pressure=function(airflow) curve of the installation with the pressure=function(airflow) of the chosen fan and determine the operation point graphically.

For a proper operation of the self-cleaning filter under the prevailing circumstances (particle size, temperature, pressure, etc.) a safe filter load in (m3/min)/m2 = m/min, must be chosen.

The filter manufacturer can advise you.

(The compressibility of the air is often neglected in these calculations)

The start of such a design is:

- determine the required air flow

- determine the allowable filter load

- determine the filter resistance factor for the operational conditions from field measurements

- determine channel cross sections by assuming f.i. 20 m/sec air velocity.

- determine geometrical layout of installation (lengths, bends, silencers)

- calculate partial pressure drops and summarize those.

- select fan, based on the chosen air flow and calculated pressure drop.

If required, repeat this procedure until a satisfactory solution is reached.

Success

Teus ■

The pressure loss for the bag filter component of your calculations will be an arbitrary figure for the “final” pressure loss when the filter bags have reached the end of their life and are fully blinded. The filter manufacturer will advise you but, in the case of a feedmill system, an allowance of 2 kPa would be normal, (depending on the overall pressure losses in the system). Ask the bag filter supplier to advise you.

Michael Reid. ■

You say that the system airflow is 20,000 cu.m/h and that the filter cloth area is 30 sq.m. Is this correct? If so, the filter velocity is very high.

For that airflow, I would expect the filter area to be 4 or 5 times more.

The bagfilter pressure loss will vary during the life of the filter bags. Initially, it will be low. As the filters age, the pressure loss will rise to a "final" figure when they have reached the end of their life. You must decide, arbitrarily what this final figure will be based upon experience with similar systems. Your bagfilter supplier will have the necessary experience. 100 mm H2O is not an unreasonable allowance.

In your case, for a bagfilter handling this kind of dust, the filters may finally wear out through fatigue (flexing on the support frames) and develop holes and leaks. They should last a number of years in a well designed bagfilter.

Michael Reid. ■

Dear HienDaoVinh,

20.000 m3/hr = 333 m3/min and a filter area of 30 m2 results in a Filter loading (velocity) of 11 m/min.

As Michael already noticed, this is extremely high.

In addition, the permeability of 180 l/m2.min indicates a filter loading of 0.18 m/min at 200 Pa. (20 mmWC)

As for dusty products (corn dust), a filter load of 1.2 to 1.5 m/min is more usual, it must be concluded that the given figures are not correct.

Using 1.2 m/min as the allowable filter load and a filter area of 30 m2, the fan air flow is 1.2*30 = 36 m3/min = 2160 m3/hr

As for the maximum pressure drop over the filter (after reaching the operational service time), a pressure drop of 350-450 mmWC should be chosen.

The fan specification is then (approx.):

Air flow = 2160 m3/hr # 36 m3/min # 0.6 m3/sec

pressure drop filter = 400 mmWC

dynamic pressure drop = 24 mmWC

pressure drop rest = 100 mmWC

Total pressure drop = 525 mmWC # 5250 Pa

Fan power approx. 6 kW

Strangely, the fan power of 22.5 kW in your data corresponds with 20.000 m3/hr at 100 mmWC

(However a too low pressure drop, which leads to a short filter life time, as the filter elements have to be replaced, because the fan cannot bring the pressure)

Note:

Keep in mind that this is the operational working point on the fan curve when the filter is at the end of its operational service time. At the start of its service time, the airflow will be higher at a lower pressure drop.

Whether you have selected the right airflow or filter area for your application, we cannot verify.

Use the experience and knowledge of a well-known supplier.

Take care

Teus ■

Hi Tues,

Yes, I think the given figures are wrong in one or more places. By the way, the filter cloth manufacturer uses the permeability as a "standard" relative measure to compare different samples of clean, new cloth. I doesn't have much to do with the selection of "final" pressure loss of the bag filter.

The filter velocity could be as much as 2.0 m/min for this application.

Michael Reid. ■

Hi Michael,

Still, the given figures of permeability and “normal” filter load pose some problems.

Permeability:

180 l/min.m2 # 0.18 m3/min.m2 = 0.18 m/min (= Filter load)

At a pressure drop of 200 Pa

Bernoulli says:

dp = * resistance factor * density * velocity^2

hence:

resistance factor = 2 * dp / (density * velocity^2)

dp = 200 Pa

density = 1.18 kg/m3 (air)

velocity = 0.180 m3/min = 0.003 m/sec

resistance factor = 2 * 200 / (1.18 * 0.003^2) = 37664783 (calculated according stated permeability)

practice:

Filter load = 2 m/min = 0.03333 m/sec

At a pressure drop of 3500 Pa (Normal, usual filter pressure drop)

Bernoulli says:

dp = * resistance factor * density * velocity^2

hence:

resistance factor = 2 * dp / (density * velocity^2)

dp = 3500 Pa

density = 1.18 kg/m3 (air)

velocity = 2 m3/min = 0.033 m/sec

resistance factor = 2 * 3500 / (1.18 * 0.033^2) = 5349667 (calculated according operational conditions)

conclusion:

The resistance factor of a clean filter material is 37664783/5349667 = 7.04 times higher than a polluted filter.

????????

Where am I missing something?

Have a nice day

Teus ■

Hi Teus,

The permeability figure is only a relative number to compare different materials in the clean condition. It does not relate to actual operating conditions.

A final pressure loss of 350 to 400 mm is way too high. A practical allowance, in my experience, for this application, would be 2,000 Pa (inlet flange to outlet flange of the bagfilter). As I said, the filters are likely to wear out through abrasion before this figure is reached.

Our correspondent should ask his supplier to confirm the allowance.

Cheers,

Michael. ■

Hi Michael,

Triggered by this thread, I focused a bit deeper in the operation of a filter in combination with a fan and different degrees of filter pollution .

To start with the permeability figure.

As this is a reference number, it still resembles a (low) filter load at a given (relatively high) pressure.

These data result in a clean resistance factor, which is very high compared to resistance factors derived from field installations.

Being a reference figure, there must also be a standardized measuring protocol.

A dust filter installation can serve several objectives:

-creating an air flow in a hood, thereby keeping an area, where dust is generated, constantly supplied with clean air from the surroundings. The fan flow equals the filter air flow.

The filter load is then the fan air flow divided by the filter area.

(F.i. a welders bench)

-Creating an under pressure in a volume, while incoming air in that volume is sucked away.

The filter fan flow is then usually 3 to 4 times the incoming airflow through the pneumatic conveying system and/or gap leakages.

The actual filter load is then the pneumatic air flow + leakages divided by the filter area.

(F.i. a volume in a silo above the material, which is pneumatically filled or a belt take up casing)

The actual filter load is then of the fan-filter load.

Example cement silo with a pneumatic loading system.

Filter area in m2 = 2* conveying air flow in m3/min (actual filter load = 0.5 m/min)

Filter fan air flow in m3/min = 2* filter area in m2 (Fan filter load = 2 m/min)

The resulting air flow and pressure drop is where the installation pressure=funct(airflow) curve intersects with the fan pressure=funct(airflow) curve.

Depending on the application, a higher or lower pressure drop is acceptable.

For cement vacuum installations, interruptedly operated in a double tank system, this 350 mmWC was accepted in order to keep the filters small. (at cost of capacity loss due to pressure loss)

150 – 200 mmWC for normal installations is indeed a better value.

Dust filter operation is more complicated than at first sight.

Have a nice day

Teus

Attachments

filterfancurvepdf (PDF)

■

Hi Teus, HienDaoVinh,

The objective in system design and fan selection is to have the “system curve” intersect the fan pressure-volume curve at the required airflow for the system, as you say.

Calculation of the pressure losses for the fixed, static parts of the system is straightforward according to well established principles and can be learned relatively easily.

But, the filter losses are variable over the life of the filter bags and allowance must be made in the fan selection for the “worst case”, i.e., the final pressure loss. This figure is arbitrary and based on experience.

Filter velocity is another selection which takes account of dust fineness, gas composition, temperature etc. entirely experience based.

At commissioning, the system airflow will be higher than designed because the filters are new and highly permeable. After a period of settling-in, with the building up of a filter cake of dust, the airflow will stabilise.

In some difficult cases, (e.g., fine metallurgical fume) a damper will be fitted ahead of the exhaust fan to restrict the initial airflow to prevent premature blinding.

The subject application is not a difficult one and a filter velocity of 2.5 m/min and final pressure loss of 2,000 Pa would result in a reliable, durable system. I realise that economics might dictate a less conservative selection. In any case, the bagfilter supplier’s experience must be relied upon to confirm. The design of bagfilters is not a job for amateurs.

Good Luck,

Michael. ■

Hi Michael, HienDaoVinh,

Permeability:

stated by Mr. HienDaoVinh: Air permeability : l/m2/mim@200Pa:160-180

From suppliers data, I found that the definition is:

Air permeability : l/100cm2/mim@200Pa:160-180

The latter results in a calculated clean filter resistance factor = 188

The proposal of Michael results in:

filter resistance factor = 200/1.18*3600/2.5^2 = 100000 (520 times clean filter)

operational airflow= 2.5 * 30 = 75 m3/min = 4500 m3/hr (Not 20000 m3/hr)

Then, there is still the problem whether the long term operation resistance factor meets the assumption of the value of 100000.

I retrieved a number of installations, which I worked on and could measure:

Pneumatic vacuum grain unloaders:

Filter load: 2.7 to 3.36 m/min at .45 bar vacuum

filter pressure drop: 50-250 mmWc in clean products (low dust load)

filter pressure drop: 380-550 mmWc in very dusty soybeans (high dust load)

filter pressure drop: 380-1000 mmWc in very dusty tapioca meal (very high dust load)

(In this case, the filters blocked, due to the fact that the cleaned dust agglomerations had a too low suspension velocity in relation to the upward air velocity between the filters.

The system kept all the dust between the filters.

Pneumatic vacuum cement unloaders

Filter load: 1.2 to 1.6 m/min at 0.65 bar vacuum

filter pressure drop: 250-450 mmWc in cement (high dust load)

Pneumatic pressure cement silo filters

Filter load: 0.5 to 0.8 m/min at atmospheric pressure.

filter pressure drop: 150-250 mmWc in cement (high dust load)

Resume:

-The applicable filter load is depending on the dust load and the particle size

Whether your choice is the correct one is hard to tell.

Did you choose the air flow to start with?

Did you choose the filter area to start with?

What was the basis for your initial choice?

Do you just want to keep the corn dust inside?

Thanks to this thread and Michael’s comments, I evaluated the filter issue further through calculations, which fortunately are more or less in line with the rule of thumbs.

However, the rule of thumbs are very depending on material properties s.a. particle size, suspension velocity, dust load, filter fabric, can velocity between the filter elements etc.

Disregarding these parameters, results in the “Hit and miss” method.

It is true:

The design of bagfilters is not a job for amateurs

Have a nice day

Teus ■

Hi Teus,

Your fan curve shows what happens typically. The initial airflow is high because the filter resistance is low. As the filters are loaded, the operating point moves up the curve to the “design” airflow. If the allowance for bagfilter pressure loss is a small proportion of the overall system, then the difference is small.

Vacuum unloading systems and dust collection systems are entirely different animals; intermittent operation –v- continuous operation, extremely high dust loading –v- low. Re-entrainment velocity and so-called “can velocity” are traps for young players. You and I can discuss the relative details until the cows come home.

What our colleague Hein Doa Vinh needs is practical advice on the design of his hammer mill system. This is not a difficult application. I suggest he engages a specialist consultant to help him to design the whole system, including the exhaust fan and advise whether the proposals of bagfilter suppliers are realistic. That way, he can avoid blind alleys and speculation and sleep better at night.

Alternatively, he can ask three separate suppliers to quote for the design of the whole system, mill enclosures, hoods, ducting, bagfilter and exhaust fan, with guarantees of quality and performance.

Cheers,

Michael. ■

Hi Michael, HienDaoVinh,

I feel confident that Mr HienDaoVinh has now the required general information on the operation and design of filter/fan installations.

Actually, that is the purpose of this forum.

Have a nice day and take care

Teus ■

Mr. Teus,

Please let me know the procedure to design a De-dusting unit. Means what are the parameters need to be considered and what is the required data and what to be calculate?

Originally Posted by Teus Tuinenburg

Dear daovinhhien,

The calculation of the pressure drop of a bag filter system is executed according the general pressure drop formula:

dp=1/2 * resistance factor * density * velocity^2

The total pressure drop is the summarization of all partial pressure drops.

-dynamic pressure drops

-filter pressure drop

-channel pressure drops

-fan exhaust pressure drop

The pressure loss of filter bag is depending on:

-filter resistance factor under operational conditions (dust load)

-density (absolute pressure, temperature)

-filter load to the square (filter velocity^2)

The filter resistance factor is a function of the degree of contamination (or cleanliness) of the filter fabric, which makes it very difficult to assess.

Therefore, the resistance factor needs to be known from field measurements on existing installations for the same material.

Channel pressure drops can be calculated, using the Fanning Factor (Darcy-Weisbach, Swamee-Jain equation)

These pressure drop calculations have to be done for the required air flow.

The required airflow and the calculated pressure drop form the operation point of the required fan.

Or, calculate the pressure drops for a range of air flows and combine the derived pressure=function(airflow) curve of the installation with the pressure=function(airflow) of the chosen fan and determine the operation point graphically.

For a proper operation of the self-cleaning filter under the prevailing circumstances (particle size, temperature, pressure, etc.) a safe filter load in (m3/min)/m2 = m/min, must be chosen.

The filter manufacturer can advise you.

(The compressibility of the air is often neglected in these calculations)

The start of such a design is:

- determine the required air flow

- determine the allowable filter load

- determine the filter resistance factor for the operational conditions from field measurements

- determine channel cross sections by assuming f.i. 20 m/sec air velocity.

- determine geometrical layout of installation (lengths, bends, silencers)

- calculate partial pressure drops and summarize those.

- select fan, based on the chosen air flow and calculated pressure drop.

If required, repeat this procedure until a satisfactory solution is reached.

Success

Teus

■

Dear respectable honers,

Please help me to get the Design procedure of Dust filter system with calculation procedure, which will help to many newly upcoming Engineers.

1. What are the parameters to be considered

2. On what basis will select the Air volume, Pressure, Speed?

3. What are the losses to be considered?

4. Is there any Authors or publications which will provide complete solution for this?

Please mail me the related documents or Links related to this topics to my mail id

gangadharreddy224@gmail.com ■

Originally Posted by daovinhhien

Thanks every body

after see some your advices i think i understand how to calculate and make good bag filter design

i reseach some document such as dust filter hand book and i check again.

i use Darcy law and pressure drop in approximate 1500 Pa for all design system.

with the normal velocity air cloth ratio is 0.07 m3/m2.s

i use the velocity for corn and rice husk, some design system run ok.

thanks every body so much.

Dear Mr Hien,

Do you check the emission from your system exhaust? Seem your air to cloth ratio is too high with rice husk dust (0.07 m3/m2/s ~ 4.2 m3/m2/min). Maximum air to cloth for this dust is 2,4 m3/m2/min. I have a system design for dust from rice husk and air to cloth is 3m3/m2/min, the exhaust seem not good. With high air to cloth, you will meet high pressure drop, time clean, and decrease bag life, high power for fan.

The pressure drop in bag filter need control under 1500 pa for good operation and bag life, avoid bag clog and damaged.

Best regards,

Tuan ■