1. Air Flow Analysis in Pharmaceutical “Clean Rooms”
Patrick Phelps
( Flowsolve )
and
Richard Rowe
( Clean Room Construction Ltd )
IPUC Luxembourg - May 2000

2. Air Flow Analysis in Pharmaceutical “Clean Rooms”
Industrial Context
Health and Safety Issues
Application to an Existing Room
Application to New Ventilation Designs
Conclusions
Experimental Verification

3. “Clean Rooms” Areas concerned with the
preparation, processing and packaging
of pharmaceutical products
Strict codes of practice employed to eradicate risk of product contamination

4. “Clean Rooms” Personnel access controlled
2-layer sterile over-clothing
Equipment sterilised before entry
Strict cleansing procedures
Particular attention to ventilation

5. Ventilation of “Clean Rooms”
Design and performance of air supply, filtration, and extraction arrangements must meet exacting standards
Positive pressure areas
Use of Laminar Flow Units (LFU’s)

6. “ Laminar Flow Units ” Devices which deliver a controlled
down-draught of re-filtered air
over sensitive regions
preparation areas
processing areas
packaging areas
storage areas

7. Air Flow Analysis in Pharmaceutical “Clean Rooms”

8. CFD Application to Air Flow Analysis in “Clean Rooms”
Context
Upgrade of ventilation system in
a suite of “ clean rooms” at a
UK pharmaceutical company

9. CFD Application to Air Flow Analysis in “Clean Rooms”
Objective
Use CFD to ensure exacting requirements
can be met , following installation of a
number of LFU’s .

10. CFD Application to Air Flow Analysis in “Clean Rooms”
Workscope
3 rooms selected as representative
examples of areas being upgraded
Room 30
Room 23
Room 42

11. Modelling Considerations - 1
Irregular shaped domains
Internal Obstructions to flow
Cartesian grid employed
K-e (Chen) turbulence model used
Buoyancy important in transient
(otherwise assume isothermal)

12. Modelling Considerations - 2
Dependent variables solved
pressure, p
lateral velocity component, U
vertical velocity component, V
Longitudinal velocity component, W
Turbulence kinetic energy, k
Turbulence energy dissipation rate, e
residence time parameter, tres
(temperature, T (for transient)

13. Modelling Considerations - 3
Boundary Conditions
Air supply ducts
prescribed sources of mass, momentum, turbulence and residence time
Air Extract ducts
specified fixed-pressure outlet sinks
LFU’s
inlets and outlets specified as above

14. Representation of LFU’s
Air flow in individual units not solved
treated as internal blockages in domain
air discharged from base at prescribed rate
matching intake from front face
interactive updating of discharge residence time
constant internal residence time assumed

15. Room 30
Small room x 5.94 x 2.92 m.
Used for processing, filling and packaging of products
Contains central plinth with filling-machine enclosure above
Conveyor-linked trays outboard of enclosure, for containers & finished goods

16. Room “ Before ”

17. Room 30 Simulations Steady flow patterns “ before ”
LFU’s in enclosure only
Steady flow patterns “ after ”
following fitment of 10 new LFU’s
75,000 node 3-D model
Distribution : 50 x 30 x 50

18. Room “ After ”

19. Room 30 Simulations Objectives Check for “dead zones”
ensure ventilation criteria met
Criteria
25 Air changes per hour
(residence time 144 seconds)

20. Room 30 - Sectional Planes

21. Room “ Before ”

22. Room “ After ”

23. Room “ Before ”

24. Room “ After ”

25. Room “ Before ”

26. Room “ After ”

27. Room “ Before ”

28. Room “ After ”

29. Room “ Before ”

30. Room “ After ”

31. Room “ Before ”

32. Room “ After ”

33. Room “ Before ”

34. Room “ After ”

35. Room “ Before ”

36. Room “ After ”

37. Room “ Before ”

38. Room “ After ”

39. Room “ Before ”

40. Room “ After ”

41. Room 30 Ventilation Summary (after refurbishment)
Complex flow paths between inlet and outlet
Pattern complicated by LFU intake extraction flows
air has to pass around central plinth and filling-machine enclosure
air in near-side passage has to “run gauntlet” of LFU’s

42. Room 30 Ventilation Summary (after refurbishment)
Outflow-weighted residence time of 121 seconds meets client criterion
“near” extract air considerably “older” than “far” extract air
some “dead zones” still apparent
evidence of entrainment into more than one unit

43. “ Dead Zones ” - before

44. “ Dead Zones ” - after

45. LFU intake times

46. LFU Unit intake times

47. LFU Unit intake times
Units D1, D2, B2, C2, A1 receive considerably fresher air than their neighbours A2-A4, B2 and C1
Air entering A4 is 2.2 times older than that entering D1
Air leaving A4 exceeds air-change criterion en route to outlet

48. Conclusions
“Residence time” analysis concept adds considerable value to vector and contour plots in assessment of complex ventilation flows.