Building Automation Systems in a PAT Framework

MANUFACTURING PROCESS ENVIRONMENTAL CONTROL

Geoff sits at his office PC reviewing data and adjusting setpoints as he monitors the manufacturing process. "Everything looks good here," he says as Julie walks in. "What do the chart recorders show?"

Jonathan T. Whitney

Julie hands him a pile of charts, and Geoff frowns as he scans the wandering lines on one chart and then on the next. "What's going on here?" he says. "I tweaked the system after the recordings you gave me yesterday, and now I have even more rooms out of tolerance."

Trying to be helpful, Julie offers, "I think Bob also tried to make some adjustments."

With growing concern, Geoff looks up. "What did Bob do? This isn't getting better — it's getting worse! Beth in QA is going to have my head for this!"

Geoff's concern about room conditions is well-founded, as production environments in many cases directly affect manufacturing processes and product quality, and as such, require calibration and validation. The relevant question is not whether space conditions are within specification at a given moment, but whether the process is under continuous control and the conditions constantly documented.

Building automation systems (BAS) can deliver that measure of control and documentation. A well-designed and well-implemented BAS essentially functions as a sub-process of a current good manufacturing practice (cGMP) process. It uses inline acquisition of critical quality data parameters to provide closed-loop feedback control and analysis of the process of manufacturing a quality environment with optimum temperature, humidity, pressure, and particle count.

THE PAT APPROACH

The use of a BAS is fully consistent with the Food and Drug Administration new Process Analytical Technology (PAT) Initiative,¹ which provides a framework and regulatory guidance for an innovative approach to pharmaceutical manufacturing and quality assurance. In its simplest terms, PAT is a methodology for designing, analyzing, and controlling manufacturing by using timely measurements of critical quality and performance attributes of raw and in-process materials. Crucial to a successful implementation of PAT is a good scientific knowledge of the manufacturing process in which all critical sources of variability are well-understood and explained, variability is managed by the process, and product quality attributes can be accurately and reliably predicted. PAT provides a process whereby product quality is a function of process design, not of final testing.

BAS Operator

This contrasts with the previous scenario, which is perhaps an extreme example but illustrates a common concern. Geoff's trial-and-error approach shows lack of understanding of the manufacturing process. To make matters worse, others are making ad hoc, uncontrolled, and undocumented changes to the process based on end-of-process measurements provided by chart recorders.

PAT charts a course to better process control and more consistent product quality. Outside the heavily regulated pharmaceutical industry, one can find many settings where PAT is well-accepted. Many industries have evolved from batch processing to continuous processes that use process analytical technology. Industries such as petrochemical, nuclear power, and steel manufacturing apply PAT concepts to improve product quality, operate more safely, reduce waste, increase facility utilization, and increase speed to market.

Building automation systems are another case in which a PAT framework has been used successfully. BAS have been applied with substantial benefit in numerous industries and are already making strong contributions in pharmaceutical production.

THE RISE OF BUILDING AUTOMATION

Until the invention of the thermostat, room-temperature systems were manually controlled. When the occupant in a room felt too cold, coal was added to the boiler or valves and dampers were manually adjusted. This procedure — essentially batch processing with end sampling — was replaced in 1885 with the invention by Professor Warren S. Johnson of the electric thermostat, which responded to space conditions, feeding back information for automatic opening and closing of valves or dampers. The building control business grew from this simple concept and continued to improve closed-loop control, providing more accurate and constant control of the process output — yielding comfortable space conditions. The addition of the mathematical concepts of proportional, integral, and derivative control strategies provided the ability to adjust more rapidly and precisely to setpoint offsets.

With the advent of computing technology, many of these control strategies moved from pneumatic and fluid logics to electric/electronic, improving repeatability and adding design flexibility. Technology increased accuracy in sensing temperature, pressure, humidity, and particulates. In addition, computers allowed data collection, so that what had been a manual process of reading and recording data at prescribed times became an automated process with enhanced reporting capabilities. The building automation industry was born.

Heating, ventilation, and air conditioning (HVAC) systems have also evolved. The HVAC system (ductwork, dampers, valves, motors, and other mechanical components of an environmental-control system) began doing more than providing comfortable space for office workers; it became critical to production, conditioning the manufacturing environment by supplying fresh air (air changes) and controlling temperature, moisture (humidify/dehumidify), airborne particles (high-efficiency particulate air — HEPA — filtration), and space pressurization.

BUILDING AUTOMATION SYSTEMS

Today's pharmaceutical market demands increasingly complex BAS and HVAC systems. There are still many systems with one heating coil, one cooling coil, and multiple, stand-alone single-loop controllers. However, as manufacturing complexity increases, so do the requirements of the manufacturing environment.

This increase has led to more complex BAS/HVAC requirements for sterile and aseptic spaces; potent compound spaces; critical pressure, temperature, and humidity control; and ISO-14644-1–classified cleanroom spaces. Add to these requirements the good engineering practices of having integrated fire and smoke systems, business concerns to provide redundant services, and regulatory concerns for secure electronic records (in compliance with Part 11 of FDA's Title 21 Code of Federal Regulations, 21 CFR Part 11), and it is apparent that BAS and HVAC systems have become highly complex (Figure 1).

Figure 1. BAS Complexity

Before the age of computerized direct digital control in BAS, pharmaceutical companies would have an employee carry a National Institute of Standards and Technology traceable test instrument into a room at specific intervals to take measurements. This person would read temperature and humidity and record the data on paper. Determining how the room had reached the desired parameters was as much art as science to the facilities personnel responsible for HVAC. Questions arose, such as, "If a room is checked after four hours and is found to be out of specification, exactly when did it go out of spec? And what caused the problem?"

The approach is similar to that of a tablet-manufacturing process that verifies a percentage of the tablets at the end of the line and finds that some do not meet specifications. In that event, the manufacturer could lose some or all of the batch. In the case of the room out of specification, everything manufactured or stored in that room for those four hours would be suspect. Furthermore, if the temperature and humidity in a controlled room are measured and are within tolerance, the following question can be raised: "Is the system truly in control?" There are many ways to obtain a room temperature of 70 degrees, and manufacturing personnel should be able to prove the temperature is at that level by design and not by accident.

WHEN TO VALIDATE

There may still be a few pharmaceutical companies performing manual temperature and humidity sampling. Some companies calibrate and validate only the room-sensing instruments and only casually commission the air-handling units with little or no documentation. This methodology disregards entirely the complex computerized mechanism that produces the air and transfers it to the room.

Federal Regulation 21 CFR 211.68 states specifically that if a computer can affect product quality, it must be validated.² The International Society for Pharmaceutical Engineering (ISPE) Baseline Guide on Commissioning and Qualification states that if something can directly affect product quality, it is classified as "direct impact," and therefore, it should be commissioned, qualified, and have enhanced documentation (such as qualification protocols and change control).³

So the question is often asked, "Does the BAS need to be qualified?" The answer depends on several factors. First, the ISPE Commissioning and Qualification Baseline Guide states that an impact assessment must be performed to determine if a device is "direct impact," "indirect impact," or "non-impact."³ Second, a risk analysis should be performed to determine the extent to which the product might be at risk due to the BAS or HVAC. These steps will indicate whether the BAS needs to be qualified. Depending on the product and the process to manufacture it (whether the process is open to the environment), the BAS and HVAC can be direct impact, have critical components, and need to be qualified. Generally speaking, if the product has temperature and humidity limitations during manufacturing or storage, or both, or if room-pressure control is important for cleanliness, and if the BAS controls these parameters, the BAS must be qualified.

Fortunately, most pharmaceutical producers appreciate the importance of their manufacturing environments, consider the BAS as direct impact, and treat it as such. The systems are designed with the overall pharmaceutical process in mind and the correct GMP documentation. User requirement specifications, functional specifications, and a thorough understanding of the process are well-documented.

If the air being produced is considered direct impact — because it could affect the drug product — then the process to manufacture it should be validated. The BAS should be designed and understood completely so that it builds in quality continuously. There should be a functional specification and validation protocols. Critical instruments should be calibrated, and the systems should be maintained under change control. As in any pharmaceutical process, if continuous quality is the intent from day one, the cost and time required to deliver this will be lessened.

A BAS that is well-understood, validated, and correctly maintained reduces the chance of a critical room parameter slipping out of specification. If something does go wrong, it will first be noticed via an alarm or warning on the air handling unit parameters. This will allow time to determine the cause of the problem and correct it before the room parameters are significantly affected. Thorough understanding and testing of the BAS and HVAC system also enables technicians to diagnose problems quickly.

BAS IN THE PAT FRAMEWORK

FDA's Guidance for Industry defines PAT as "a system for designing, analyzing, and controlling manufacturing through timely measurements (i.e., during processing) of critical quality and performance attributes of raw and in-process materials and processes with the goal of ensuring final product quality."¹

A BAS is a sub-process of the overall process of manufacturing a pharmaceutical. In this context:

• The conditioned air is the finished product

• The outside air, return air, and moisture added through humidification are the raw materials

• Computerized control of the heating coils, cooling coil, dehumidification, pressure, and other parameters is the process.

Conventional pharmaceutical manufacturing typically uses batch processing and evaluates quality through laboratory testing on collected samples.¹ In the HVAC world, this is analogous to running an air handler and using a chart recorder to monitor the space temperature, humidity, pressure, and particle counts. In control terminology, this is an open-loop process.

BAS/HVAC systems, by design, are closed-loop feedback systems that provide continuous process control. They do not manufacture controlled, conditioned air in batches, but rather manufacture it continuously from widely varying outside air conditions. This type of process readily lends itself to the PAT tools of:

• Multivariate — for design, data acquisition, and analysis

• Process analyzers

• Process-control

• Continuous-improvement and knowledge-management.¹

Multivariate Tools for Design, Data Acquisition, and Analysis

There are many ways to design an HVAC system and many ways to implement controls to provide specified space conditions. Fortunately, there is a large experience base to draw upon, above and beyond regulated environmental space in the pharmaceutical industry. Much of this experience has been codified in building codes such as those of the Building Officials & Code Administrators and the National Fire Protection Association, presented as consensus standards such as those of the American Society of Heating, Refrigerating, & Air Conditioning Engineers, and tested in the marketplace, building the industry knowledge base.

Properly instrumented BAS networks also enable capture of huge amounts of in-line and at-line critical process data. Full-featured BAS networks can store this data in a secure format with audit trails for full 21 CFR Part 11 compliance. Operator tools provide simple, secure access to raw data, statistical analysis (outlier identification, mean kinetic temperature, and other data aggregation features), and summary reports (Figure 2).

Figure 2. Sample Summary Report

Process Analyzers and Process-Control Tools

In the HVAC environment, temperature, humidity/dewpoint, pressure, and particle counts are the critical quality parameters of the process. Process analyzers installed at- and inline at various stages allow real-time monitoring. The BAS uses these sensor inputs to monitor the process and, through closed-loop process-control strategies, continuously manipulates the process in real time to maintain the desired process output.

Continuous-Improvement and Knowledge-Management Tools

The large volumes of data generated from the process analyzers is gathered by the BAS and stored in a secure database (21 CFR Part 11– compliant) with ready access for review, analysis, and regulatory submission. On-board BAS tools provide for graphical trending, calculation of mean kinetic temperatures, and comparative analysis. Copies of data sets can be exported for further statistical analysis.

CONCLUSION

As Geoff reviews a binder of validation documentation, Julie walks in and asks, "How's the process running?" When Geoff logs on to his secure workstation, a graphical representation of the HVAC system appears, showing sensor readings of the critical input values, control command outputs, motor speeds, valve and damper positions, and process-output parameters with corresponding real-time sensor readings.

"Things look good," Geoff says. "I can see from the trend log that the temperature rose half a degree starting at 13:40 hours, but the BAS control system compensated and brought it back down within 30 minutes. The alarm log is clear."

Julie smiles. "You know, Beth in QA will want to review all this before she releases this run."

"No problem," Geoff says, "I'll generate a report with the data for the last 72 hours from the secure server and e-mail it to her with my electronic signature."

Jonathan T. Whitney works in business development in life sciences in the northeast region of Johnson Controls, Inc., Pine West Plaza, Bldg. 4, Albany, NY, 12205-5515, 518.869.9595, jonathan.t.whitney@jci.com

Mark A. Granger is manager of validation services for the eastern region for Johnson Controls, Inc., 2250 Butler Pike, Plymouth Meeting, PA, 19462, 610.276.3810, mark.a.granger@jci.com

REFERENCES

1. US Food and Drug Administration. Guidance for Industry, PAT — A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance. Rockville, MD: Office of Training and Communication, Division of Drug Information, HFD-240, Center for Drug Evaluation and Research, Food and Drug Administration; September 2004. Available at: http://www.fda.gov/cder/guidance/6419fnl.htm.

2. Code of Federal Regulations. Title 21, vol 4. Revised as of April 1, 2004. 21CFR211.68. Part 211 — Current Good Manufacturing Practice for Finished Pharmaceuticals. Sec 211.68: Automatic, mechanical, and electronic equipment.US Food and Drug Administration web site. Available at: http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=211.68 and http://www.fda.gov/cder/dmpq/cgmpregs.htm.

3. International Society for Pharmaceutical Engineering (ISPE). Commissioning and Qualification. Tampa: ISPE; 2001. ISPE Baseline® Pharmaceutical Engineering Guide; vol 5. Summary available at: http://www.pharmaceuticalonline.com/Content/ProductShowcase/product.asp?DocID={6CA64F7D-CC1D-419B-9F2D-FF417CA20D70}&VNET.