Air compressor intake filtering 1

[rawelk]

We have a mix of I-R and Kaeser air and water cooled rotary screw air compressors. Our plant's ambient air has enough dust (plastics fines, cardboard dust, and, since the equipment room is adjacent to the maintenance department, often influxes of other types of dirt), and in all cases we've added intake air filter boxes using an expanded metal mesh for filter backing, and 1/2" polyester filter media to reduce cleaning and replacement of the more expensive compressor-mounted air filters.

All but the Kaeser units use 3/4"-16 trade size expanded metal (78% open area) while the Kaeser filter boxes use 1/2"-13 trade size (57% open area) mesh.

The Kaeser FS400 compressors have two louvered doors, and each door is fitted with two filter boxes. I've never paid much attention to this, but while replacing a failed motor noticed that they block off three louvers (where the two boxes meet in the middle). Also, the interior sides of each box is fitted with two panels of acoustic baffling which also block off a percentage of the original louvered open area.

I'm wondering how much of an effect these added restriction of steel mesh and filter media have on air compressor performance, and whether it is worth the effort to re-work the Kaeser's filter boxes.

I'm contemplating using expanded metal with greater open area (I might be able to use 1-1/2"-16 trade size with 83% open area), remove or reduce the depth of the acoustic baffles to prevent them from blocking the louvers, and, instead of two boxes per door, use a single large one instead.

When comparing our water-cooled versus air-cooled compressor filter surface areas see there is a large and expected difference between air and water cooling, but a fair amount of range in the air-cooled ones of approximately 70 to 120 square inches per rated SCFM (at 110 PSI).

http://files.engineering.com/getfil...c&file=compressor_filter_comparison_table.jpg

Is there a rule-of-thumb pertaining to how much filter area to use per compressor CFM?

Are their any specific references discussing the effects of air compressor intake restriction I should be reading?

 

[crjones]

When you add exterior filter boxes you are adding pressure drop. If you are not using a manufacturer designed addition keep the flow are the same. If the compressor has an opening of say 100 sqin and your using expanded metal with a 60% open face your add on box needs to have a face area of 167sqin.

Restricting the inlet air flow reduces the capacity of the compressor and will cause overheating leading to premature component failure. The OEM fan is designed generally large enough to only support the basic unit.

One way to check, is to put a thermometer inside the enclosure while its running and compare to one outside the unit. Compare this to what the manufacturer says, generally they look for a 10-15 degF rise.

 

[rawelk]

Thanks, crjones, your observation about compensating for reductions in the effective open area between the OEM's original design, and our filter modifications makes sense to me.

It isn't difficult to build boxes to effect this surface area increase in the case of the I-R compressors, but there might not be enough room on the Kaeser compressor doors to make a 1:1 equivalence (although it'll be a lot closer than it is now).

 

[Bribyk]

Wouldn't any intake air restriction just be "throttling" the compressor's inlet? This is sometimes used to regulate screw compressor output, why would it be giving them overheating/motor issues unless the motor was trying to achieve a certain flowrate with a throttled inlet? The motor load/amperage should be reduced unless there's a feedback loop that's trying to achieve a specific setpoint.

 

[crjones]

Approximately 15% of the energy input into a screw compressor package is thrown off as heat within the enclosure. This creates a temperature rise. Electric motors and other electric components are rated at 104 degF(40degC). Assuming the control center is package mounted, you can have premature overload tripping, ambient compensation only goes so far. Say you plant temperature in the compressor area is 100 degF and your added restrictions have an enclosure rise of 25 degF. so the motor is seeing 125 degF (50 degC), by rule you have cut the motor life in half.

As far as controlling the speed of the motor to match requirements. Most installations use pressure. If I throttle the inlet into the compressor by pressure drops then ter is less mass flow through the unit. The pressure builds up more slowly. So the unit is running at max amp draw for a longer period of time building up more heat. And the ROTI calculation that justified the fancy controls went out the window.

 

[rawelk]

crjones, you have me in 100% agreement except for your description of how motor temperature rise works. Insulation class and service factor must be factored in as well as ambient temperature.

Toshiba application note AG-05 is one of the more concise explanations.

http://www.toshont.com/ag/mtrldesign/AG05 (Temperature Rise).pdf

However, the 10°C rule of thumb for halving (or doubling, depending on what direction you are going in) you note is the accepted standard. I've never seen anyone work it out from first principles because it requires experimentally determined values to plug into the Arrhenius equation.

http://en.wikipedia.org/wiki/Arrhenius_equation

One thing is certain, however. Decreasing air flow (and hence, cooling potential) has negative consequences.

 

[crjones]

Rules of thumb are meant to give general guidance not absolutes. They just see if you are pointed in the right direction.