Selected Highlights of the Labs21 2010 Annual Conference

Challenging Air Change Rate Guidelines in Good Manufacturing Practice Facilities

Jason Rifkin, Equilibrium      
Patrick Goetz, JennErik Engineering

Abstract

As research and development of biotechnology progresses to marketable products, those products must first be manufactured within a controlled environment. This environment is regulated by the U.S. Food and Drug Administration as well as the European Medicines Agency (for drugs intended for European sale), and the attributes of the facility design are codified in regulations and guidelines. The focus of these regulations is to ensure a controlled, clean, and (where applicable) sterile environment to mitigate the risk of potential contaminates to biologically derived products.

Different levels of clean production space are classified by the maximum particulate counts allowable within a volume of air (measured as particulates greater than 0.5 micrometer [µm] within a cubic foot or meter of air), as well as other factors. Associated with these classifications are air change rates to obtain the various levels of particle-free air. Our premise is to challenge these air change rates as necessary to achieve the associated classifications of "clean rooms" and to propose active controls to modulate the air change rates based on activity within a production space at any given time. A key question is raised by this review. Do the air change rates drive the particle counts or are they merely arbitrary guidelines?

When designing and building a clean room, a great deal of emphasis is placed on reducing particle counts and, when necessary, ensuring sterility. It is important to note that a clean facility is not necessarily sterile and a sterile facility is not necessarily clean. Depending on what is being manufactured, one or both criteria must be met.

In our experience, and as empirical data will show, the air change rates set forth by government regulations and industry standards, coupled with current room construction practices, results in much "cleaner" air than the room classifications would dictate. For example, Figure 1 shows a graphical representation of particle counts within a Class 10,000 (less than 10,000 particle counts greater than 0.5 µm) clean room. This clean room was designed to 40 air changes per hour (ACH), which is the typical design specification for this classification of room in a sterile biotechnology production facility. What we see are particle counts generally less than 1,000, with a rare excursion above the 1,000 count level.

Figure 1
Figure 1: Sterile, Class 10,000 Non-Viable Particulate Counts within an Operational Facility


In addition, Figure 2 describes conditions in an ISO 8 (Class 100,000) facility. This facility was designed to 20 ACH and is a non-sterile biotechnology production suite. A similar pattern of particle counts below 1,000 is observed with the highest particle count reaching 2,241. Remember, the goal of this facility is to achieve particle counts below 100,000.  

Figure 2
Figure 2: Non-sterile, Class 100,000 Non-Viable Particulate Counts and Colony Forming Units, During Clearance Studies within a Validated Facility


Air change rates drive first costs in equipment selection and continued operating costs associated with elevated utility bills from the fan energy and cooling required to move that much air through these regulated facilities. By reducing the air change rates with actual data in conjunction with modulating the air change rates down, when the facility is not in production, first costs and continued operating expenses in production facilities can be greatly reduced.

In conclusion, we would recommend a design approach based more on actual data and scientific principles, not on arbitrary guidelines. We would recommend integrating the design of facility finishes and heating, ventilation, and air conditioning selection into the mix, instead of working in a "vacuum." Finally, we would challenge the particle count guidelines themselves, since designing clean rooms along a logarithmic scale of various particle counts seems too "neat" to be based on actual scientifically obtained data.

Biographies

Jason Rifkin, has worked in the life science industry for the past 17 years and brings extensive experience in the biotechnology and life science sector previously working for Celera Genomics as a quality control supervisor and at NeuralStem as a researcher and laboratory manager. Mr. Rifkin has worked with Turner Construction Company as a life sciences construction market consultant and for Scheer Partners as a senior vice president in charge of development and construction of life science facilities.

Mr. Rifkin holds a Bachelor of Science in biology from the University of Maryland at Baltimore County, a Masters of Science in neurobiology from Montana State University, and a Masters of Business Administration from the University of Baltimore.

Patrick Goetz has 22 years of experience in the life sciences industry where he has prepared master plans, energy analyses, and contract documents, and estimated and provided construction, commissioning, and project management services for single buildings as well as campus facilities. His projects have ranged from research and development to manufacturing facilities for research, biotechnology, and pharmaceutical clients. His experience includes both the design and design-build management for mechanical, electrical, and plumbing systems for multiple clean room and biosafety level 3 facilities.