Predicted Contamination Levels in Cleanrooms When Cleanroom-Dressed People Are the Contamination Source

May 1, 2006
Berit Reinmuller, PhD

Berit Reinmuller is a senior researcher in building services engineering at Kungl Tekniska Hogskolan, SE 10044 Stockholm, Sweden, tel. +46 8 790 7537, fax +46 8 411 8432, berit.reinmuller@byv.kth.se. She is also a consultant for the cleanroom industry.

Pharmaceutical Technology, Pharmaceutical Technology-05-01-2006, Volume 2006 Supplement, Issue 2

Calculations show the predicted contamination levels in cleanrooms with turbulent mixing air and with vertical unidirectional airflow when people are dressed in modern cleanroom clothing systems. Comparisons are made between operation theatres and cleanrooms.

Particulate and microbiological tests conducted on selected clothing systems in a dispersal chamber installed at KTH (1–3) have shown that the particle levels reach higher values after 25 washing and sterilizing cycles than after 50 cycles (see Table I). This result might be explained by the fact that after a certain number of washing and sterilizing cycles, the fabric releases particles but with time, the released particles seem to be washed away from the fabrics. With this information from performed studies of the source strength, it is possible to calculate predicted contamination levels from people dressed in clothing systems washed several times while in cleanrooms with turbulent mixing ventilation and in cleanrooms with vertical unidirectional airflow.

Turbulent mixing air

If the contamination sources and the design of the ventilation system are known, a mathematical model can be built of the level of airborne contaminants in a cleanroom having completely turbulent mixing air. (Descriptions of such ventilation systems are provided in various textbooks.) With the assumptions of no leakage into the room and of close to 100% efficiency from the final filters (e.g., HEPA-filters), the simplest possible expression describing the concentration c in the room is

in which qs is source strength—outward particle flow (numbers/s) and bacteria-carrying particles (colony-forming units [cfu]/s)—and Q is the total airflow (m3/s). This equation provides estimations of contamination levels in the following examples.

Example 1. In an aseptic filling room of 90 m3 and 20 air changes per hour, the only contamination sources are the cleanroom-dressed operators. How many operators are allowed in the filling room if FDA 2004 (5) requirements for US Customary Class 10,000 (ISO Class 7) shall be fulfilled?

To find the answer to this question, use the following limit values of FDA 2004 (5), US Customary Class 10,000 (ISO Class 7) in a dynamic state:

  • number of particles ≥0.5 µm = 352,000/m3;

  • number of airborne cfu = 10/m3.

Using the data provided in the example, the total airflow Q = 0.5 m3/s. The calculation procedure is as follows:

  • Use Equation 1 to calculate the concentration in the cleanroom when one person is present: find the source strength for one person (data from Table I) and divide by the total air volume flow.

  • The maximum allowed number of people in the cleanroom will be the limit value (or recommended value) divided with the above calculated concentration for one person.

Table I: Comparison of the source strength (mean values) for people dressed in various clothing systems after various wash and sterilization cycles (4).

According to Table I, the maximum particle-source strength for a high-quality cleanroom clothing system occurs around 25 washing and sterilizing cycles. For particles ≥0.5 µm, the value of qs is 3950 particles/s. The concentration becomes:

Concentration (particles ≥0.5 μm) = 7900 particles/m3

The maximum allowed number of people in the filling room depending on particle size will be the limit value of the FDA Class 10,000 divided by the calculated concentration. Therefore, for particles ≥0.5 μm, the maximum allowed number of operators within the filling room is 44.

The source strength for cfu increases with the number of washing and sterilizing cycles (see Table I). Following the same calculation steps and using the initial data from Table I for cfu and high-quality cleanroom clothing systems, results have been summarized in Table II.

Table II: Source strengths for high-quality cleanroom clothing systems and estimated number of operators allowed.

Table II shows—with FDA US Customary Class 10,000 requirements—that after 50 washing and sterilizing cycles, the allowed maximum number of operators is four. Should the FDA 2004 (5) with Clean Area Classification US Customary Class 1000 be used instead, the microbiological limit is set to 7 cfu/m3 , and the maximum number of operators allowed would be only three.

This example shows that when the operators are the only contamination source, the value for cfu presents a stricter requirement than that of particles. On the other hand, various production machinery also generate particles, which could be a reason for the larger ranges for particles ≥0.5 μm.

If unidirectional flow units with HEPA-filtered air are used, the situation improves. The airflow through the units should be added to the room airflow. For example, if in the filling room a unidirectional flow unit with HEPA-filtered air is installed with the same airflow as the room airflow, the total airflow is 1.0 m3/s. Therefore, the maximum number of people allowed in the filling room is about twice the value estimated in the above examples.

Example 2. This example calculates the number of people dressed in surgical clothing systems that can stay in an operating room with turbulent mixing air when the total airflow is 2000 m3/h during conventional operation and orthopedic operation (e.g., hip joint replacement) when recommended values of airborne cfu/m3 are as follows:

  • <100 cfu/m3 for conventional operation;

  • ~5 cfu/m3 for orthopedic operation.

The airflow is given to 2000 m3/h, which is approximately 0.56 m3/s. Table I shows the source strength for cfu and surgical clothing system washed once, 25 times, and 50 times. Tables III and IV show the results after using the same calculation procedure as in the first example and the initial data provided.

Table III: Source strengths for surgical clothing systems and estimated number of people allowed during conventional operation.*

Results indicate that surgical clothing systems for conventional operations should not be used after more than 50 washing cycles. For orthopedic operations, the surgical clothing system discussed here should not be used at all. It should be mentioned that hip joint replacement operations often are performed in an HEPA-filtered, unidirectional airflow environments with the surgical team dressed in high-quality clothing systems.

Table IV: Source strengths for surgical clothing systems and estimated number of people allowed during an orthopedic operation.*

Downward (vertical) unidirectional airflow

A mathematical model describing the dispersion of airborne contaminants when a cleanroom-dressed operator is standing in a vertical unidirectional airflow has been reported (6).

Measurements have been performed on an operator inside the dispersal chamber with an air velocity of 0.45 m/s (7). The operator had a height of 1.86 m and an estimated radius of 0.15 m. While in the chamber, the operator's arms moved in a calm manner in standard cycles (one hand from hip shoulder and back, then the same movement with the other hand) (7). The source strength for particles ≥0.5 μm was calculated to 500/s per meter. This value might be somewhat small, so an additional value of the source strength chosen of 20,000 particles/s per meter (particles ≥0.5 μm) was used in the theoretical calculation. Data from Reinmüller show that this value for the source strength is comparable with "lab coat–disposable coat" (1). Figure 1 shows both the theoretical and measured contamination region edges for an operator. The contamination region edge is set to 35 particles/m3 (particles ≥0.5 μm) or 1 particle/ft3 ).

Figure 1 also shows that the measured results for an active operator (i.e., with arm movements) deviate from results based on the mathematical model. For a moving operator, the distances from the center of the cylinder to the contamination region edge are ~50% larger than those theoretically calculated.

Despite the large difference in source strengths between the two theoretically calculated curves—predicted for an operator standing still—in Figure 1, the contamination regions differ less than 0.1 m.

For surgical clothing systems, calculations using data from Table I show that the contamination-region edges are positioned between the two theoretical edges (see Figure 1). When estimating values for an operation in a vertical unidirectional airflow, use the distance from the surgeon's head to the operating table or the operator's head to the plane of the exposed product.

Figure 1: Contamination region edge (c(ρ,x) = 35 particles (≥0.5 µm) per m3) at an air velocity of 0.45 m/s. The thick line represents the edge for measured values. The thinner lines show the position of the calculated edges according to the theoretical model and the source strengths of 500 and 20,000 particles/m s (7).

Studies previously conducted have shown the relationship between the number of particles ≥0.5 μm per volume unit of air and the number of aerobic airborne cfu per volume unit of air (1, 3). In a typical cleanroom environment, with people dressed in new modern cleanroom clothing systems (washed and sterilized once) as the main contamination source, it may be possible to establish relationship at a ratio of ~1500:1 (particles to cfu). When the cleanroom clothing systems are washed and sterilized several times, this relationship will increase, and the values of the ratio are estimated to be <10,000:1 (compare Table I). This means that the contamination region edges will have theoretical cfu values much lower than 1 cfu/m3.

In Europe, filter ceiling systems with HEPA-filtered air and with flows about 10,000 m3/h (~2.8 m3/s) are commonly used in operation rooms. The air velocity just below the filter ceiling is below 0.3 m/s. According to Nordenadler, who has performed smoke-visualization tests, air movements are rather sensitive to disturbances (8). People, the radiation of heat, and objects (e.g., lamps, filter fixtures) have a decisive influence on the flow pattern and even the transport of contaminants, which has been shown with particle challenge tests and a particle counter.

In the region around the operation table, the air movements are not unidirectional but irregular. Therefore, the calculation procedure described for turbulent mixing air should be used to estimate the maximum number of persons that are allowed to be in this room. Because the airflow is five times higher for the filter ceiling system than that of a conventional operation room, the number of people allowed in the room with filter ceiling system are about five times those values given in Tables III and IV.

In conclusion, a unidirectional airflow system with velocities <0.3 m/s will never fulfill the demands of a Class 100 (ISO Class 5) environment during operational conditions.

Discussion

The predominant sources of contaminants within a cleanroom are people and machinery. Potential risk situations are created by the interaction among people, air patterns, and the dispersion of airborne contaminants, and are difficult to predict. Calculations show that it is possible to make a first estimation of the number of people allowed in a classified cleanroom during activity when people are the main source of airborne contaminants. These estimations are based on data from source strengths when people are dressed in various clothing systems.

When people are the main source of airborne contaminants, the microbiological requirements constitute stricter demands than those for particles (5), likely because the machinery of production lines also generates particles. During the design and construction phase, source-strength data can be used in a first estimation of the expected air cleanliness during activity. In a dynamic state (e.g., in an aseptic filling room), environmental monitoring must always be performed in a conventional way.

Operating rooms are seldom classified as cleanrooms and regulatory recommendations do not exist in the same way as for pharmaceutical cleanrooms. It should, however, be noted that the tested surgical clothing systems for conventional operations have a limited life and should not be used after 50 washing cycles. For orthopedic operation rooms where higher levels of cleanliness are desirable, this kind of clothing system should preferably be exchanged for high-quality cleanroom clothing systems.

Bengt Ljungqvist, PhD, is a professor of safety validation and head of the department of civil and architectural engineering, and Berit Reinmüller, PhD,* is a senior researcher in safety ventilation, Building Services Engineering, in the department of civil and architectural engineering, KTH, The Royal Institute of Technology, Brinellvägen 34, SE-100 44 Stockholm, Sweden, tel. +46 8 790 75 37, fax +46 8 411 84 32, Berit.Reinmüller@byv.kth.se.

*To whom all correspondence should be addressed.

References

1. B. Reinmüller, "Dispersion and Risk Assessment of Airborne Contaminants in Pharmaceutical Clean Rooms," PhD thesis, Bulletin No. 56, Building Services Engineering, KTH, Stockholm (2001).

2. B. Reinmüller and B. Ljungqvist, "Modern Cleanroom Clothing Systems: People as a Contamination Source," PDA J. Pharma. Sci. Technol. 57 (3), 114–125 (2003).

3. B. Ljungqvist and B. Reinmüller, "People as a Contamination Source: Cleanroom Clothing after 1, 25, and 50 Washing/Sterilizing Cycles," Eur. J. Paren. Pharm. Sci. 8 (3), 75–80 (2003).

4. B. Ljungqvist and B. Reinmüller, Cleanroom Clothing Systems; People as a Contamination Source (PDA/DHI Publishing, LLC, River Grove, IL, 2004).

5. US Food and Drug Administration, Guidance for Industry, Sterile Drug Products Produced by Aseptic Processing—Current Good Manufacturing Practice (FDA, Rockville, MD, 2004).

6. B. Ljungqvist, B. Reinmüller, and O. Söderström, "Cleanroom-Dressed Operator in Unidirectional Airflow: A Mathematical Model of Contamination Risks," Eur. J. Paren. Pharm. Sci. 8 (1), 11–14. (2003).

7. R. Jonsson, "Particle Dispersal from a Cylindrical Source in a Clean Zone with Unidirectional Airflow," MSc thesis, Report No. 85, Building Services Engineering, KTH, Stockholm (in Swedish) (2002).

8. L. Nordenadler, thesis in progress, Building Services Engineering, KTH, Stockholm (2006).