ENERGY EFFICIENT LOW OPERATING COST CLEANROOM AIRFLOW DESIGN [Paper presented at IEST’s ESTECH 2003 Conference, Phoenix, AZ May 18-21] Rajan Jaisinghani, Technovation Systems, Inc.

Biography Rajan (Raj) Jaisinghani is a chemical engineer with over 30 years of R&D experience related to fluid mechanics, particle science, including colloid and aerosol science, filtration and cleanroom technology. Raj has a BS in Chemical Engineering from BHU, India and a MS with additional graduate work from the University of Wisconsin. He is widely published in the above fields and holds 12 patents. Currently Raj is President and CEO of Technovation Systems, Inc.

Abstract Up to 40% of the non process related initial and operating costs of cleanrooms are dependent on the airflow rate used in cleanrooms. Three factors that largely affect the energy and operating cost efficiency of cleanrooms are: a) the airflow rate design method, b) the type of air handling system and air flow distribution and c) the performance characteristics of the filter system utilized. A critical review of airflow rate design charts is presented, followed by a review of analytical design methods such as the dilution and transient analysis models with an emphasis on energy efficient design. The effect of various air-handling and new lower pressure drop filtration systems are also considered. Finally, a few examples of utilization of the newer design methods and air handling/filter systems are presented showing the associated energy savings.

Keywords airflow, cleanroom, airflow design, energy efficiency, operating cost, dilution model, transient analysis, airflow distribution, energy efficient filters, air filters, air handling

Introduction and Scope Recent work (1) aimed at bench marking cleanroom energy consumption has shown that the biggest factor affecting non-process related cleanroom initial and operating (energy) cost is the airflow rate. And cleanrooms use a lot of airflow! Surprisingly little attention has been paid to this the most important aspect of cleanroom design. In fact cleanroom airflow design methods have not changed in decades and what is worse is that the existing method – charts showing required airflow as a function of class (e.g. Table I (2)) – have in fact no traceable technical basis. Figure 1 shows the variables affecting airflow design and thus cleanroom performance. Clearly charts such as Table I do not take into account any of the variables associated with performance. The truth is that engineers designing cleanrooms conduct no calculations for the design of airflow. Thus the most important variable of the mechanical air handling system is simply obtained from a highly dubious chart. Energy efficiency dictates a more technical approach. This paper summarizes currently available methods - methods that have a technical basis. Table I Typical Airflow Design Chart (2)

Cleanroom Class 1 10 100 1,000 10,000 100,000

Airflow Type Unidirectional Unidirectional Unidirectional Mixed Mixed Mixed

Av. Airflow Velocity, fpm 60-100 50-90 40-80 25-40 10-15 5-10 [corrected]

Air changes/hr 360-540 300-540 240-480 150-240 60-90 5-48

In general very little attention has been paid to the energy efficient design of air handling systems. Specific areas of improvement include duct design, air distribution, selection of energy efficient motors and filter systems and out-of-process (night) operation energy saving methods. The lack of progress in such areas is at least partially due to industrial standards and

standard and validation bodies that propagate non science based standards, which inhibit the development of innovative energy efficient systems. The LBL (1) study shares these view points and has culminated into a research plan for addressing such needs.

Design Variables Process Contamination Generation

Filter Efficiency

Makeup Air& Concentration

„ Energy Efficient Airflow Design Particle Transport Rate

Process Sensitivity

Performance Criteria/Class

„ Airflow Distribution

Re-entrainment

Airflow design depends on many variables. Design charts do not take into account the impact of these variables.

Figure 1 Variables affecting cleanroom performance This paper reviews alternate methods for airflow design for cleanrooms. Basic trends, regarding efficacy of filtration systems, room velocity and other design variables that are evident from such models are also presented. Additionally, airflow distribution and the use and energy advantages of using new ultra low pressured drop filter systems are discussed. Examples of cleanroom airflow design and associated energy savings are also presented.

Cleanroom Airflow Design Methods Design methods should incorporate the cleanroom performance variables illustrated in Figure 1. For energy efficient, low initial and operating cost and for trouble free (not under or over designed cleanrooms) it is important that these variables be considered. In Figure 1 the filtration efficiency is the combined efficiency of the entire system, including prefilters. The process contamination generation rate can be easily measured by use of a particle counter around various machines. In fact equipment manufacturers’ should be obligated to provide such estimates. Similarly, the make up air quality can be determined. The difficult design parameters are related to process sensitivity. The process sensitivity should determine how long particles generated in the room will stay (i.e. is related to transport). Each process has different sensitivity with respect to contaminants and only analysis of product QCA data and or yields can provide answers to such questions. Current Chart Design Method As stated above the current charts (Table I) have no traceable technical basis. There are claims that these charts are based on experience factors. However, there is no documented evidence of this. The IEST is currently reviewing their version of such a chart (3); however, it is not clear what the basis for such a revision is. A preview of this revision (3) clearly shows that neither the old chart nor the revision have any parameters that take into account the important design variables shown in Figure 1. Utilizing such charts can result in under or over designed airflow rates depending on the values of the internal contamination generation rates and the makeup air contaminant concentration. Essentially, the designer is left with guessing the airflow rate – a variable too important to be left to guess work! The Dilution Model In order to develop a rational basis for designing cleanroom airflow it should be understood that the primary purpose of the clean airflow is to a) dilute the contaminants produced in the room and b) to transport or carry away such particles.

Jaisinghani (4) has presented a simplified dilution model that is based on rational analysis. This is summarized and reviewed here. Figure 2 shows the schematic of the model as applied to cleanrooms. The cleanroom design variables are defined in

A Simple Model for Dilution Particle Concentrations #/ft3 Ca = Conc. in Supply Cd = Design Conc. or Conc. in EV exit Cm = Conc. in make up f = Fraction of make up G = Internal generation #/min Q = Flow rate scfm = EV Velocity, fpm EV = Elemental Volume 1 ft3, w/ 1ft2 cross section

Cm Make up air, f Bank of Filters

R e t u r n

Q Ca Cleanroom

G

EV, Elemental Volume

Cd

Figure 1 - Model Schematic

Figure 2 – Dilution Model Schematic Figure 2. The diluted concentration, Cd, in the elemental volume with internal particle generation rate G is given by: Cd=Ca+G/Q

(1)

From this Jaisinghani (4) develops the following relationship between cleanroom design concentration and the design variables: Cd = [f (1-E) Cm+ G/Q]/[1 - (1-f)(1-E)]

(2)

This model is applicable to a single elementary volume in a cleanroom provided the following conditions are achieved in the airflow distribution: a) The airflow rate is evenly distributed so that essentially the flow is unidirectional with negligible amount of dispersion. b) Transport or the rate of transport is not a factor here. c) Steady state conditions are assumed. The model may be used for an entire cleanroom using a marching technique, similar to that used in finite element analysis. The equation 1 may be re-written as: Cd(i) = Cd(i-1) +G(i)/Q

(3)

where, i refers the current elemental volume and (i-1) refers to the elemental volume above this element. Then equation 2 only applies to the first element directly under the filters, analogous to a boundary value equation, except that Cd would be replaced by Cd(i). For elemental volumes below this initial volume equation 3 applies. In other words beyond the first elemental volume the filtration system does not play a role in the cleanroom dilution process. Hence, using equations 2 and 3 the entire cleanroom volumetric concentration may be profiled, by marching from top down. However, since cleanrooms are classified based on the highest concentration in the room, only the dirtiest cross section of the cleanroom may be analyzed.

Jaisinghani (4) has applied this model to a simplified cleanroom zone or cross section. Although simplified there is a significant amount of information to be gleaned from this. The calculations are based on the following conditions: 1. There is only one source of generation within a single elemental volume (as shown in Figure 2). 2. The make-up air conditions are fixed - 3% make-up air (as a percent of total air flow circulation), with make-up air concentration, Cm, fixed at 500,000/ ft3. 3. Four filtration systems are considered: a) 99.9%, b) 99.97%, c) 99.99% and d) 99.999999% filtration efficiency at 0.3 Um. The last one is an extremely high efficiency filtration system such as achievable at reasonable pressure drop using electrically enhanced filtration2 as a primary HEPA in series with a terminal HEPA filter. The mid efficiencies, 99.97% and 99.99%, are common HEPA efficiencies. The 99.9% efficiency filter, a HEPA filters of the past, is used to consider a lower HEPA efficiency filter that will have a significantly lower pressure drop. Figures 3 – 5 show the results of these calculations.

Cd, 0.3Um Conc.#/ft3

1000.0

Cm = 500,000, f = 0.03 G = 10,000

FILTER EFFICIENCY 99.90% 99.97% 99.99% "99.999999%"

100.0 10

20

30

40

50

60

Velocity fpm Figure 3 Effect of filtration efficiency and velocity at high values of G

1.

2.

3.

The following are the salient observations from these calculations: Referring to Figure 3, for higher class (i.e., less clean) cleanrooms, lower (than HEPA) efficiency filters may be utilized for proper cleanroom performance at significantly lower pressure drops (and thus lower energy consumption) in such cleanrooms. Current practices however continue to utilize HEPA filters in such applications (ISO Class 7-9) although such filters have a higher energy cost and provide no advantage over more energy efficient lower efficiency filters which also tend to last longer (have higher life or dust holding capacity). Referring to Figure 4, for all values of G used, higher velocities result in lower diluted concentration. However, at value of G, this effect has a limiting asymptotic value at around 60 fpm average room velocity. In other words adding more airflow beyond about 50-60 fpm there is a diminishing return in terms of improved performance. Figure 4 also illustrates that the cleanroom performance is in the end limited by the cleanliness of the cleanroom processes and practices or G. Referring to Figure 5, higher efficiency filters can be utilized to achieve the desired cleanroom classification at lower airflow rates- for the same internal particle generation rates. Although the more restrictive filters may be required for this, the lower airflow rates required may more than offset this and result in a net energy savings.

None of these trends are apparent in the airflow charts used by most cleanroom designers. Clearly airflow requirements for a particular room class depend on the internal generation rates, flow velocities, and filter efficiency for given make-up air conditions. Design charts such as Table 1 do not show such dependencies. While this model is simplified, it nevertheless takes into account, in a manner that is borne out by experience, the effect of various design variables. It also illustrates some important design considerations (see items 1-3 above). It should be noted that the values of Cd in the above calculations are low because only a single generation source in a single elemental volume is considered. As additional elemental volumes are considered the values of Cd obtained by this analysis will be higher –depending on values of G and Q in accordance with Equation 3.

9 9 .9 7 % 0 .3 U m F iltra tion

G = 100 00 G = 100 0 G = 100 G = 10

1 00.0 #/ft3

Cd, steady state conc,

C m = 50 0,000 , f = 0 .03

10 00 .0

10.0

1.0 10

20

30

40

50

60

V elocity, fpm Figure 4 – Effect of Velocity on Dilution

Cd 0.3Um Conc. #/ft3

100.0

FILTER EFFICIENCY

Cm = 500,000, f = 0.03 G = 100

99.90% 99.97% 99.99% 99.999999%

10.0

1.0 10

20

30

40

50

60

Velocity fpm Figure 5 – Effect of Filtration Efficiency Transient Analysis Jaisinghani (5) has also presented another method for airflow design utilizes the solution to the material balance equation as applied to a well mixed room. Due to the high air changes per hour cleanrooms are quite well mixed although not to the extent that there are no spatial variations. Hence, this model results only in an average room concentration. This average value may be used along with an assumed distribution function, such as a log normal distribution, to enable calculation of the 95% Upper Confidence Level (UCL) concentration. It should be noted that the ISO 144661 (6) standard allows for using the 95% UCL for determining the classification level of the cleanroom. The transient analysis model (5) is outlined below. The overall material balance equation is, as applied to a room with re-circulating air is:

V

dC = - QvC - E (Qf - Qv)C - KVC + QVCV(1 - E) + GV dt

(4)

where V is the room volume, ft3 C is the volume average room particulate concentration greater than a certain size, #/ft3 E is the fractional filter efficiency for the size range of interest Qf is the filtered air flow rate, scfh Qv is the make up air or ventilation air flow rate scfh Cv is the concentration of particles, greater than the size of interest, in the make up or ventilation air K is the #/ft3 particle decay (not due to filtration) rate constant, h-1 G is the internal room generation rate of particles greater than the size of interest, #/(ft3-h), and t is time, h. The solution (5) to equation 2 is: C(t) = [1-e-Ot] I/O + Co e-Ot

(5)

where Co is the initial average room concentration at time t = 0. and I, the net ingression parameter, is given by, I = Qv Cv (1-E)/V + G

(6)

and O, the removal parameter, is given by, O = [(1-E) Qv + EQf ]/V + K

(7)

At steady state (assuming a pseudo equilibrium), the equilibrium concentration is given by, Ceq = I/O For a non re-circulating air flow room, Qf = Qv, and hence, I = Qf Cv (1-E)/V + G

(8)

O = Qf /V + K

(9)

Figure 6 shows a typical analysis for a room that has been initially contaminated and shows the room recovery with respect to time. K is estimated to be approximately 0.1/h from based on experimental studies of particle decay in chamber without airflow. The value of G is estimated based on an average or worse case value based on measurement of process equipment. The use of the model requires some heuristic criteria regarding recovery time. Basically, this is an experience based value and is based on the premise that lower class (cleaner) rooms should clean up or recover faster to their design value. The fundamental basis of the use of this model is that since rooms operate in a fluctuating ingression mode (due to material, personnel entry and process fluctuations). A simplified method of dealing with this design challenge is to calculate recovery time after a single large step ingression. It also stands to reason that cleaner rooms should need to respond to these mini ingressions of contaminants faster. So although this method does not eliminate some experience based factors it does take into the account the variables that affect cleanroom design. Computational Fluid Dynamics and Transport Computational fluid dynamics (CFD) is a well establish method for determining the stream lines or spatial velocity profiles in open and duct flows. The most common method used is by Patankar and Spalding (7). The velocity profiles can

TRANSIENT RESPONSE using EEF HEPA FILTERS - DOUBLE HEPA SYSTEM >0.3 Um ROOM CONC. #/ft3

1,000,000 400 FT2 ROOM, 250 SCFM MAKEUP, G

100,000 10,000

7400 SCFM, 18.5

1,000 100 CLASS 10

11,100 SCFM,

10 1 0

1

2

3

4

5

TIME min Figure 6 – Example of a Transient Analysis then be used with the conservation of mass equation to result in concentration profiles. Boundary value conditions that incorporate filtration and sources for particle generation can also be introduced, but with a very much higher degree of complication. It should be noted that these methods require specific and detailed knowledge regarding the three dimensional profiles of the various process equipment in the cleanroom. Inaccuracies in such inputs result in large inaccuracies in the results. CFD analysis incorporating filtration and point source generation requires highly trained personnel with advanced knowledge of fluid dynamics. CFD analysis on the other hand can be fairly easily utilized as a means of determining airflow patterns in cleanrooms especially to analyze the effect of airflow return openings and their placement within the room. CFD analysis also is useful for estimating particle residence time in the room. This is an important design variable. Increasing the airflow rate tends to reduce the residence time until the onset of turbulence which can cause eddies to trap particles for longer time periods. The average residence time, Tr, is equal to the reciprocal of the average room velocity.

Air handling Considerations The above methods of airflow rate determination should be used in conjunction with other sound engineering principles and innovative concepts to result in energy efficient cleanroom design. Just a few of these concepts and principles are presented here. Unidirectional, Laminar and Turbulent Flow Often cleanroom airflow is incorrectly referred to as being laminar. More often than not these rooms are at best unidirectional. Non laminar unidirectional flows may have small scale turbulent fluctuations which are usually acceptable. What is not acceptable is to have highly turbulent flows that create significant eddies, since eddies usually result in larger residence time of particles trapped in such eddies. Recent work (7,8) has shown that cleanroom performance deteriorates at about average room velocities of about 65 fpm probably due to turbulent eddies formation. This means that simply increasing

airflow rates in cleanrooms will not always improve performance and that velocities above 65 fpm not only add to energy consumption but also adversely affect performance. It should be noted that the FDA cGMP guidelines call for 90+/- 20% fpm velocity at the filters which is contrary to the authors experience and the findings of the above studies (7,8). This is because in most ISO Class 4 pharmaceutical applications 100% ceiling coverage with filters is the norm. This velocity guideline’s origin has been traced to the ad hoc value obtained as a filter velocity in one of the first cleanrooms at Los Alamos. It should be also noted that HEPA filters perform better at lower velocities than at higher velocities (9). This is another example of a non science based guideline. It should be mentioned however that the primary concern of the cGMP guidelines is that there should be a “sweeping flow “ that transports away particles viz. unidirectional flow. Often it is the validation companies that have made the 90 fpm requirement to be an edict rather than a guideline. Airflow Distribution and Air handling Systems It is intuitively obvious that higher ceiling coverage by airflow distribution devices such as terminal filters or supply diffusers will produce higher performance cleanrooms or for equal performance higher ceiling coverage will require less airflow and thus lower energy consumption. To apply this concept the filtration or air handling system should have what is known as flow rate and flow velocity independence. The centralized air handling system (typically one rooftop unit with prefilters and terminal room sided HEPA filters) meets this requirement. However, due to larger air flows than used in standard HVAC applications, these units are typically custom designed units operating at non standard and possibly less efficient coil velocities than standardized units. Additionally, such customized systems cost about 3-5 times as much as standardized air conditioners. In order to overcome this handicap a distributed air handling scheme, as illustrated by Figure 7, is used. This system uses multiple in-duct fan filter units, which are the prime movers of air. This allows the use of standard high efficiency air conditioners that simply pull in the rated airflow from the return ducts or plenums and make up air, condition this air and feed this typically smaller fraction of the total re-circulating air into the intake of the in-duct fan filter units. The in-duct fan filter units then act as mixing boxes. In this system for ISO Class 5 and above no terminal HEPA filters are utilized. For lower class rooms terminal HEPA filters are also used to result in double HEPA filtration with a system filtration efficiency of 99.999999% at 0.3 Um. The advantages of double HEPA filtration have already been established; double HEPA filters require significantly less airflow than single HEPA filtration.

D is t r ib u te d A ir H a n d lin g S y s te m In d u c t f a n u n its w ith p r im a r y U L P D ™ H E P A s

AC m a k e u p

• F lo w / v e lo c it y in d e p e n d e n c e • A C f lo w is n o t dependent on t o t a l f lo w

O p tio n a l B a n k o f T e r m in a l H E P A f ilte r s

C le a n r o o m F lo w S c h e m a t ic

Figure 7 The Distributed Air Handling System

1. 2. 3.

Other advantages of this air handling system are: The filters are replaced outside of the room – no contamination inside the room. With double HEPA filtration the terminal HEPA filters should never require replacement under normal circumstances. Less noise inside the room since there are no fans on the ceiling.

4.

Typically no roof re-enforcement is needed to install the standard air conditioners as opposed to the higher weight customized central air handlers.

Ultra Low Pressure Drop Filters During the last 6 years electrically enhanced filtration (EEF) systems have been successfully commercialized (10,5) for cleanroom and other applications such as residential and commercial buildings. These filters typically use a lower efficiency, lower pressure drop filter and electrically enhance the filter to perform at HEPA performance levels while still retaining the low pressure drop advantage of the basic filter material. Combining the cleanroom and other applications there are now over 16,000 units in operation, some of them for over 5 years. Clearly, the reliability of this technology has been proven. Recent advances in this technology (11) have vastly magnified the advantages of this technology, by utilizing the advances made in mechanical construction of commercial filters along with improvements to the EEF technology itself. The result is ultra low pressure drop HEPA filters having only 0.5”WC pressure drop at 2400 scfm in a 24”x24”x11.5” deep size filter! This is 60% lower than conventional mechanical HEPA filters of the same size. This means that these units will consume 60% less energy than conventional filter units. Another advantage of this technology is that these EEF HEPAs have about 2.5-3 times higher dust loading capacity. This results in significant savings in operating and maintenance costs. The EEF HEPA filters are available as filters that can be installed in any air handling equipment or in separate housings with or without fans. Thus they can be utilized in both central and distributed air handling systems and the advantages of low pressure drop can be used to increase the energy advantages of the lower flow required for double HEPA filter systems. Table II below shows the relative savings by using lower pressure drop primary filters in different class cleanrooms versus conventional mechanical primary filters. The pre-filters and terminal filters are always conventional filters in both cases. Here a centralized or distributed system is assumed such that in both cases the pre-filters and primary filters are 2’x2’ in cross section and are operated at 2000 scfm per filter with the same airflow rate (assuming properly designed airflow) in each case. Hence only the terminal filter pressure drop (PD) is dependent on the room velocity. For this analysis it is assumed that for ISO Class 1-3 there is 100% ceiling coverage of terminal filters. For Class 5-7 it is assumed that there are no terminal filters used. The relative savings due to system 1 in this case are given by: %Savings = [1 – (PD1/PD2) ] X 100

ISO Class 1 10 100 1000 10K

(10)

Table II Energy Savings Due to Ultra Low PD Filters at Same Airflow Rates Room Prefilter Primary Terminal Total Prefilter Primary Terminal Total Velocity PD1“WC HEPA HEPA PD1 PD2“WC HEPA HEPA PD2 fpm PD1“WC PD1“WC “WC PD2“WC PD2“WC “WC 60 0.25 0.4 0.3 0.95 0.25 1.0 0.3 1.55 45 0.25 0.4 0.22 0.85 0.25 1.0 0.22 1.47 25 0.25 0.4 0.125 0.77 0.25 1.0 0.125 1.37 16 0.25 0.4 NA 0.65 0.25 1.0 NA 1.25 7 0.25 0.4 NA 0.65 0.25 1.0 NA 1.25 Sub 1 = System w/ Ultra Low PD primaries Sub 2= System with conventional PD primaries

System 1 %Savings Eq. 10 38.7 42.2 43.8 48.0 48.0

In this case where all things are equal except the primary HEPA filter, the use of Ultra Low PD filters results in a relative filter PD energy savings ranging from 38.7% for ISO Class 2 to 48% for ISO Class 7. Table III shows the results when comparing a system (1) designed in accordance to the transient and dilution models above with Ultra Low PD primary HEPA filters to a system (2) designed in accordance to the design chart shown in Table I with no primary HEPAs but only terminal HEPAs (single HEPA or ULPA filters). It is assumed that for system 1 the primary HEPAs and the pre-filters have the same size as above and operate at 2000 scfm. Similary the system 2 pre-filters are the same size and also operate at 2000 scfm each. Obviously in order to achieve at least similar (but somewhat inferior) performance in system 2 ULPA filters would have to be used for ISO Class 1-3 rooms. Further, for System 2 for Class 1000 and above it is assumed that there is only 35% ceiling coverage and for Class 100 there the ceiling coverage is 75%. These values and the average room velocity determine the terminal filter PD. The percent savings, due to system 1, in terms of filtration system pressure drop, for this case are given by: % Savings = [1-V1xPD1/(V2xPD2)] x 100

(11)

The results for this case in Table III shows that although for most classes of cleanrooms the system 1 has slightly higher pressure drop due to the lower flow rates utilized in this case there is a net relative energy savings related to the filtration systems of between 29.6 (for ISO Class 2) to 48% at higher airflow rates. In order to get a more quantitative value of energy savings the following formula may be used. The result is the power savings due to low pressure drop filters compared to conventional filters taking into account the overall or thermodynamic efficiency of converting electric power into work. Power Saved = Flow Rate x (Pressure Drop Savings)/Overall Thermodynamic Efficiency (12) In English units, Power Saved, KW = Flow Rate, scfm x Pressure Drop reduction, “WC/Fractional Overall Efficiency x 0.0001176

(13)

Assuming an overall thermodynamic efficiency (fan plus motor) of 0.38 each ultra low pressure drop filter run at 2000 scfm will save 0.3713 KW or 3252 KWH per year per filter for continuous operation. Table III Savings Due to Ultra Low Pressure Drop Primary HEPAs using Models vs Charts Room Class

Sys. 1 Velocity

V1 fpm 1 10 100 1000 10K

60 45 25 16 7

Prefilter

PD1 “WC 0.25 0.25 0.25 0.25 0.25

Primary Terminal Total HEPA HEPA PD1 PD1”WC PD1”WC “WC 0.4 0.4 0.4 0.4 0.4

0.3 0.22 0.125 NA NA

System 2 Velocity

V2 fpm

0.95 0.85 0.77 0.65 0.65

90 75 50 30 16

Prefilter

PD2 “WC 0.25 0.25 0.25 0.25 0.25

Terminal

Total

System 1

PD 2 % Filter “WC Savings PD2”WC Eq. 11 0.65 0.54 0.33 0.42 0.23

0.9 0.79 0.58 0.67 0.48

29.6 35.4 33.6 48.3 40.7

Summary The methods presented here provide a rational basis for cleanroom airflow design. Energy and operating cost efficiency can only be achieved by utilizing such engineering analysis rather than simply using dubious design charts which do not take into account the variables that affect cleanroom performance. Additionally guidelines regarding air handling and the use of ultra low pressure drop filters have also been presented. These methods along with these guidelines and ultra low pressure drop filters have been extensively and successfully used by Technovation in the design of about 75 cleanrooms. Figure 7 shows the design velocities used in some of these cleanrooms versus the velocities recommended by the Table I chart. Clearly in these cases the chart velocities are unnecessarily high. Although these methods are rational they are only a start; more research is needed in order to develop more sophisticated design methodology.

E x a m p le s o f A irflo w C u s to m e rT y p e B u r lin g to n F a b r ic s D o w C h e m - la u n d r y M e d P h a rm E n c e lle B io te c h

IS O C la s s (2 0 9 E ) 2 (< 1 )

6

D e s ig n

D e s ig n fp m 5 5

V e l

C h a r t V e l. fp m 7 5 -1 0 0

4

(1 0 )

3 5

7 0 -9 0

4

(1 0 )

4 0

7 0 -9 0

(1 0 0 0 )

1 6

2 5 -4 0

Figure 7 – Actual Design Velocities vs. Table I Chart Velocities in Select Projects

References 1. 2. 3. 4. 5.

6. 7. 8. 9. 10. 11.

Lawrence Berkley Laboratories, (2000) “Lawrence Berkeley National Laboratories High Tech Buildings Program – Cleanroom Benchmarking Plan”, Internal Report, LBL, Berkeley, CA. Hansz, T., (1996), “Cleanroom Programming and Planning”, Proc. CleanRooms 96 East, 11th Int. Conference On Advanced Technology and Practices for Contamination Control, p 233. Fitzpatrick, M. and K. Goldstein, (2002), “Cleanroom airflows Part II: The Messy Details”, Cleanrooms, Vol 16, #7. Jaisinghani, R.A., (2000)," New Ways of Thinking About Cleanroom Airflow Design", A2C2, Vol 3, # 11. Jaisinghani R.A., T.J. Inzana and G. Glindemann, (1996), “High Filtration/Biocidal Performance Cleanroom System”, Proc. CleanRooms ’96 West, The Conference on Advanced Microcontamination Control and Ultrapure Manufacturing, October 28-30, 1996, Santa Clara Convention Center, Santa Clara, California International Standards Organization, (2001), Tannous, G and K. Compton, (1998), “Studies Conclude Low Air Velocity Increases Effectiveness of Minienvironment Design”, Cleanrooms, March issue. Vazquez, M. (1999), “The Study of Altering Air Velocities in Operational Cleanrooms”, Proc. Cleanrooms West ’99, Santa Clara, CA. Dhaniyala, S. and B.Y.H. Liu, (1999), “Investigations of Particle Penetration in Fibrous Filters”, J. IEST, v 42, #1. Jaisinghani, R. A. (1995) U.S. patent 543,383. Jaisinghani, R.A. (2002), Patent pending.