Clean Room Application 2

[friartuck]

I am working on a clean room manufacturing area where the client (A local Hospital) make their own tablets and drugs. The clean room area have to be of very high air quality for which I am proposing terminal HEPA filtration, but I need to create a cascading pressure from the cleanest areas to the 'dirtiest'. On average the differences from room to room will be 10 to 15Pa (.04 to 0.06"wg).

The cguidance in CIBSE is pretty poor as it doesn't have a table of losses through doorways ate various pressure differentials. There is something referred to in the hospital techncal memoranda as the CDP (Component data base), but I don't know where to get a copy.

Has anyone got a decent table of pressure drop and flow rate tables for various door and window openings or an equation to accurately calculate the losses and flow rates.

Any help would be appreciated otherwise I can feel a 'wet finger in the air' and start from basics approach coming. (Just joking about the last comment by the way)



Friar Tuck of Sherwood

 

[tbedford]

I haven't seen tables of pd thru various doorways etc. but would suggest that 5 pa differential from clean to less clean would result in a quieter system.
the best book i have seen on the subject is "practical cleanroom design" by raymond k. schneider.,but it does not include the data you are searching for.

tom

 

[friartuck]

Thanks for the book info. I have designed a few clean rooms and theatres (Just finished 20 theatres in a local hospital). But the info we designed to is in a manual called HTM2025 (Which I have a copy..obviously)

However the pressure regimes called for are pretty basic in a theatre, I have a complex labrynth f corridors and rooms with varying pressures and differentials etc. I wish to know if either any one has access to the CDP (component data base) or a similar table or an equation that gives accurate and realistic values of leakage for any pressures ---which in my case will be between 5 and 15 to 20Pa) The client is very pinickitty and I think perhaps over specifies his pressure differentials..Usually they are no more than 10Pa. But he wants higher. And I haven't got a suitable table or method of calculating.

Thanks in anticipation.

Friar Tuck of Sherwood

 

[imok2]

Friertuck: this may be of some help...I think

http://www.bfrl.nist.gov/IAQanalysis/Cdemo_work07.html

 

[quark]

10 to 15 Pa is pretty common in pharma industry and the people are very rigid with these limits. You can calculate the leakage through openings by the formula

Q = 2610 x A x (dp)^1/2, where Q is in cfm, A is area of opening in sq.ft and dp is differential pressure in inches of water column.

I would like to give you some suggestions though 10 to 15 Pa is not very difficult to maintain.

1. Use adjustable door sweeps with the doors (which go up when you open the door and go down when you close it - Dorma can supply you these items)

2. Generally windows should be sealed type with double panels to take care of contamination.

3. Compensating the leakage when door is fully open is quite redundant unless you want to maintain the pressure differentials even though the door is opened. However, this is too theoretical and not possible. If you operate the system via BMS, put a time delay for recording pressures.

4. Try to minimize the conveyor openings.

5. My practice is to consider 5mm gap below the doors, 2mm between the two door closures(if it is a double door) and 1mm from sides and top(i.e door closure to door frame)

6. Open the doors to positive pressure(you can easily force them against 10 to 15 Pa pressure). Otherwise you may have pressure problems with partially opened doors. It is better to take a trial along with door manufacturer.

Hope this thread403-43928 will be helpful to you.

Regards,

 

[lilliput1]

We use 165% of quark's CFM because we calculate PD in. wg as 1.5 x the velocity pressure through the door crack. The 1.5 coefficient cosist of 0.5 entry loss ant 1.0 exit loss through the door opening. Quark's equation has an equivalent loss coefficient of 2.48.
Make sure you have adequate pressurization/cfm differential to allow for valve innaccuracy, control tolerance and control span. Require sealing of wall penetrations, gaskets on receptacles, sealing of wall to floor & to slab above & sealing of sprinkler heads.
Provide access door to allow cleaning of air flow measuring stations on valves. Avoid use of flow straighteners because they trap dirt/lint. Use air flow tracking or constant volume two position (occupied/unoccupied)control. The Phoenix venturi type valves require the least maintenance and no straight run of duct. However they will require separate airflow measurment if VAV tracking control is used or if CFM monitoring is required.

 

[SAK9]

friartuck,

If the rooms in question have no openings for conveyors,it is pointless trying to calculate offset volume between supply and exhaust as you can never accurately estimate the crack area because crack area consists of door leaks as well as leaks on partitions,ceiling board, improperly treated penetrations etc.From my experience,assuming a 10% leakage based on supply volume works quite fine for a reasonably tight room(clean rooms need a decent level of tightness)with a single door.However if the room has a number of doors assume about 100 cmh per door,you should be ok.

If you have coneyor openings,multiply this area with the velocity equivalent of the required room pressure to get the theoretical maximum air leakage through this opening.Room pressure can be adjusted by either modulating exhaust flow rate or supply air flow rate the correct solution depending on the design and specific requirements.Whether to go in for active pressure control or volumetric offset for achieving the pressure differential depends on what industry you are in.Generally for the pharma industry active pressure control is the way to go whereas for labs volumtric offset may be more appropriate.

 

[friartuck]

Can anyone tell me what is the maximum differential pressure that I can have between rooms. If the PD is high, it will be difficult to open or keep closed the doors. I have 15Pa across most interfaces but I might have 30Pa across another???? Anyone have any info????



Friar Tuck of Sherwood

 

[quark]

In one of the plants I even maintained 45Pa dp (filling with respect to filtration area). This generally happens when one area has multiple door openings to various areas. In the above system, the filling area was preceded by three change rooms but filtration area was directly attached to filling area.

Though it will be rather easy to open a door against 30 Pa dp, it is very difficult to maintain a door closed against a dp of morethan 15 Pa. Your door closer also plays main role. Your option will be between momentary door opening and permanent door opening.

Lilliput!

Do you mean 1.5 x 2610 or 1.5 x 4005? For a fixed dp, the maximum flow rate will be 4005xAx(dp)^1/2. I couldn't find entry and exit losses in either case from you equation. Can you please let me know?

To all!

Recently I read a paper from ASHRAE Journal about pressure balancing in cleanrooms. Half of the paper describes about the difficulty and precision with which the experiment has been conducted and the later half describes about the experimental results. The author's final suggestion is to have a differential flow of 150 cfm for a PD of 15 Pa but to keep redundancy upto 450 cfm(300%!!!). I think this shows the inconsistency and difficulty of pressure balancing. Well.... I gotta catch my boss.

Regards,

 

[lilliput1]

PD = 1.5 x (V/4005)^2
CFM = V x A

Where PD is in inches wg
and V in FPM
A is crack area in SF

The 1.5 is sum of the 0.5 entry loss & 1.0 exit loss coefficient.

 

[friartuck]

QUARK

Which ASHRAE Journal, I can get a copy

Tucky

Friar Tuck of Sherwood

 

[quark]

lilliput!

OK, I understood it.

Friartuck!

Feb'2003 issue and the topic is Room Pressure for critical environments.

Also check Dec'2001 issue for the topic Air Tightness.

If you are a member, you can download articles from previous issues also.

Regards,