Comparing Facility Layout Options for Managing Business and Operating Risks

Efficiently and reliably manufacturing biopharmaceuticals requires controlling business and operating risks using enterprise control strategies (ECSs). ECSs are built using the three manufacturing enterprise elements-process, facility, and infrastructure (1). The facility element includes the facility’s layout strategy. The layout can be used to decrease the likelihood of realizing both business risks associated with product development and manufacturing, and risks associated with operating sequences of process unit operations (UO) grouped into logical operating units (LOUs) necessary for manufacturing products. After defining the goals of all control strategies (CSs) and briefly describing an enterprise’s control elements, this article compares the impact of several common facility layout strategies, including the multi-purpose facility (MPF) (2, 3), on managing important business and operating risks. 

Manufacturing facility designs

Facility design layout strategies used within the biopharmaceutical industry generally fall into three categories: 

• Purpose built facility (PBF)-Layout is designed around the process equipment required to implement a well-defined UO/LOU sequence for a process or set of processes. Large-scale manufacturing facilities designed around fixed stainless-steel systems are PBFs. The PBF may also use some or all single-use technologies. However, when a PBF is designed, the process implementation is simultaneously integrated into the facility layout to achieve the desired efficiency, segregation, and operating flows. PBFs may have more than one production train. Multiproduct manufacturing is usually executed on a campaign basis within a single process train. Flexibility is limited to process formats included in the initial design. Adapting a PBF to new processes can be expensive or not feasible because of ongoing manufacturing requirements. 

• Shared flexible space facility (SFSF)-Layout design is based on using large open spaces for either fixed or moveable single-use technology or stainless-steel equipment required for making one or more products. The layout commingles a large number of UOs for one or more processes and products in a flexible open ballroom configuration. Segregation may be limited to large LOUs such as upstream and downstream, or specialize activities such as bulk filling.  Operations within large spaces are conducted by common personnel to achieve labor efficiencies. Multiproduct operation may or may not occur within shared spaces. The simple layout of the SFSF reduces capital investment requirements.  Amgen’s Singapore facility is an excellent example of the SFSF layout concepts (4). 

• Multi-purpose facility (MPF)-Facility layout, as shown in Figure 1, is based on a matrix of small non-dedicated, multifunctional operating areas accessed by primary and out corridors allowing both bi and unidirectional flow of materials, equipment, and personnel. Multiproduct manufacturing is accomplished by placing product dedicated LOUs using movable single-use technology or stainless-steel equipment within a variety of possible room configurations capable of operating the process (2, 3). Advantages of MPF are similar to the SFSF except the MPF is more complex due to the increase in the number of rooms, corridors, and other facility systems. 

Figure 1  Matrix layout. Facility layout shown is a matrix of individual non-dedicated multi-purpose rooms in three arms (A–C) each having six rooms (1–6) that can be independently configured to house a variety of process and support function LOUs necessary to achieve the require manufacturing capabilities.  All equipment, primarily single use, is movable for placement or relocation within the matrix as required to operate a wide variety of process implementations. The facility can be expanded by increasing the number of arms and the number of rooms in each arm in the initial design or adding arms later above the out corridor at the top of the figure (2, 3). [All figures courtesy of the authors]

Each facility layout strategy has strengths and weakness. The PBF has been the classic layout strategy used to design biopharmaceutical facilities, particularly those for operating large stainless-steel systems for a well-defined process. The SFSF was developed using single-use technology to provide additional operating flexibility and reduce facility complexity to lower capital investment (4). The MPF was proposed to speed up the launching of new products by providing sufficient scale and process implementation flexibility to operate a wide variety of process formats and simultaneously support pre-clinical, clinical, and commercial manufacturing with minimal tech transfer time and effort (2, 3). 

Control strategy goals

All manufacturing control strategies, including enterprise-wide control systems (ECSs) have three goals: 

• State of control-Provide a qualified robust control strategy that reliably controls all the process’s operating steps to achieve predefined product material attributes and process behavior quality metrics that assures each step can be released for executing the following operating step, including release of final product. 

• Proof of control-Provide documented release based on product and process quality metrics at control points for each operating step to allow release for executing the next step. The information should be sufficient to prove to an unbiased external reviewer that the step was completed as planned and defined. Proof of control can be, by far, the hardest goal because it sometimes requires “proving a negative” associated with establishing that a failure of an external operation within the same manufacturing enterprise did not impact UO steps for other processes or products. 

• Return to control-Should an operating step not pass its release criteria, assure sufficient process information is collected to determine the failure’s impact on product quality and rapidly return the step to a state of control using an investigation, including a root cause analysis. 

Achieving all three goals is a difficult task, particularly when dealing with multiproduct manufacturing of a wide range of upstream and downstream UOs combined into a variety of LOUs to achieve important operational and process segregation requirements (e.g., pre- vs. post-viral, etc.). Control strategies designed to achieve all three goals must be constructed using an efficient combination of the following ECS elements.

Elements of enterprise control strategies

A pharmaceutical manufacturing enterprise can best be described by three ­elements that are combined to complement each other to achieve the above control strategy’s goals. Each element provides important tools for building effective control strategies.  The most efficient and effective ECSs use an appropriate balance of the following ­elements (1): 

• Process-UOs, grouping of UOs into LOUs, process equipment, components, instruments, automated process control systems, input and in-process materials, and products. Process systems can include stainless-steel and single-use systems. 

• Facility-Buildings, environmental systems, layouts, operational flows, logistical support, utility systems, and other building control systems such as the building management system.

• Infrastructure-Practices, procedures, people (training, discipline, and qualification), maintenance systems, and automated procedural control systems (MES, EBR, etc.) that control the facility and process elements.

The enterprise can be summarized as “the process operating inside the facility under the control of the infrastructure” (1).

With the control strategy’s goals and the enterprise elements for building ECSs defined, the foundation of the risk management method for understanding the layout’s impact on controlling various risks can be described. 

Risk analysis approach

The risk analysis approach is a system risk structure (SRS) methodology (5). SRS is based on a threat-process-risk consequence model shown in Figure 2 for business risks and Figure 3 for operating risks. SRS describes the likelihood that one or more threats, such as an input failure or change in the risk process, will result in a negative risk consequence (5). The risk process can be designed or redesigned to control the risk by decreasing the likelihood that the threat will successfully pass through the risk process to result in the realized risk. 

Figure 2. Business risks (BR) and threats (BT) that may adversely impact the manufacturing enterprise.  The discussion focuses on the relative likelihood of three facility layout strategies (PBF, SFSF, MPF) controlling the threat’s ability to produce the risk consequences. PBF is purpose built facility. SFSF is shared flexible space facility. MPF is multi-purpose facility.

A SRS’s risk process can be anything from an entire manufacturing facility, in this case the facility’s layout, to a simple piece of equipment or procedure depending on the scope of the risk analysis. Based on the processes to be evaluated, the risk processes can be combined into sequences similar to a process flow diagram (PFD) to form a SRS for analyzing complex risk problems (5, 6). Because all risks are assumed to be output consequences from a risk process caused by an input threat or trigger, the only difference between a threat and risk is the process it comes from in the PFD or SRS.

In this article, the risk process is limited to one of the facility layout strategies (PBF, SFSF, or MPF). If the facility layout does not control the risk by adequately mitigating the likelihood of the threat’s impact, then the risk must be either accepted, or other ECS elements, such as scheduling, procedures, and facility modifications, must be added or enhanced as necessary to provide sufficient control of the threat. In the past, some of the business risks described in the following have frequently been accepted as a “fact of life” (e.g., product delays, shortages, additional capital costs, etc.) with great negative impact on patients and the business’s bottom lines.

The following risk analysis subjectively rates the severity and likelihood of various threats and risk consequences relatively to each other. The subjective ratings are based on the author’s 33 years of engineering, operations, and risk analysis experience in the biopharmaceutical industry (5, 6). When evaluating different layout options, each company should make the same ratings based on their specific situations, experience, and expertise. 

Figure 3. Operational threats and risks that may adversely impact the performance of the manufacturing enterprise. The discussion focuses on the relative likelihood of three facility layout strategies (PBF, SFSF, MPF) controlling the threat’s ability to produce the risk consequences. UO is unit operations. PBF is purpose built facility. SFSF is shared flexible space facility. MPF is multipurpose facility.

Significant business threats and risks

The following discussion is limited to a few important representative threats and risks that could possibly occur to the manufacturing enterprise. Figure 2 summarizes a few of the business threats (BT) and business risks (BR). 

The threats are risk consequences that result from threat processes (e.g., from a prior or secondary threat) not described in this analysis. The ­likelihood of the secondary threats occurring is assumed to be independent of the layout selected and thus outside of the scope of this discussion. 

• Changes in product demand (BT1)-Market forecasts have not accurately anticipated product demand. 

• Product failures during development (BT2)-Product fails during clinical testing and no longer needs to be manufactured. 

• Uncertainty in process design and format (BT3)-During process development, the UO/LOU sequence or format for a new product is significantly different than previous processes (e.g., from batch, used to design the facility’s layout, to an intensified or continuous process).

• Insufficient or excessive support resources (BT4)-Capacity limitations resulting from and inability to supply the process with sufficient media or buffers.

• Inefficient facility layout (BT5)-The layout increases operating labor or capital investment. 

Although the business threats vary widely between companies, their impact severity on patients and economic sustainability can be ranked qualitatively relatively to each other (> greater than, >> much greater than) in the following order:  (BT1>BT2 >> BT3>BT4 > BT5). 

Each of the layouts could be impacted by any of the threats individually or collectively to result in the risks described as follows. Each risk will be briefly discussed in terms of the layout’s overall ability to decrease the likelihood of the threat’s impact on producing the risk. A detail analysis is left to each company to understand the impact of the threats and risks on selecting a layout strategy. 

The following BRs shown in Figure 2 are subjectively evaluated without considering the interactions with the operating risks. Realization of operating risks can also have a significant impact on any or all of the following BRs:

1. Insufficient manufacturing capacity (BR1)-Demand for clinical or commercial material cannot be supplied because manufacturing capacity is unavailable or existing manufacturing capacity cannot run the required process. 

Relative likelihood rating:  BT1 >> BT3 > BT4.

Unless initially constructed (an exposure to BR3, BR4, and BR5), expansion of the PBF’s capacity requires the time and capital for building a new facility or extensively modifying the existing layout. 

The SFSF can prioritize product campaigns, but rearranging processes in different configurations and formats (BT3) may be difficult to achieve without interfering with on-going manufacturing and increasing operating risks.

Because of its high flexibility, the MPF can quickly prioritize products to increase capacity or be quickly expanded by adding additional suite rows at the top of Figure 1 with minimal impact to ongoing production. 

Relative overall threat mitigation rating:  MPF > SFSF >> PBF

2. Delayed product launch (BR2)-An inability to launch a product due to the unavailability of manufacturing capacity. 

Relative likelihood rating:  BT1 > BT3 > BT4.

Having capacity creation on the critical path can have a significant impact on the approval process, patient health, and business revenue. The impact of prebuilding capacity (BR3, BR4, BR5) is significant and depends on the flexibility of the facility to accommodate different processes associated with other products in the pipeline. 

PBFs are typically not designed for scale flexibility or efficiently adaptable for both early clinic and commercial-scale process campaigns. Including options for scale and process flexibility significantly increases capital investment.

SFSFs may be designed with scale flexibility, but extensive ­infrastructure control strategies are required to manage commingled multiproduct pre-clinical, early clinical, and commercial ­production.

The segregated matrix of the MPF provides a great deal of process scale and format flexibility.

Relative overall risk mitigation rating:  MPF > SFSF >> PBF

3. Excess or unused manufacturing capacity (BR3)-Unused ­manufacturing capacity because of BT1, BT2, BT3, and BT4. BR3 is concomitant with BR4 and BR5. 

Relative likelihood rating:   BT2 > BT1 >> BT3, BT4.

The PBF is the most likely to be oversized because it may have to accommodate future capacity growth associated with possible market expansion. Other products produced in the PBF would have to have similar processes.

The SFSF has some flexibility to manipulate its equipment arrangement to adapt to other product’s processes, but product change-overs are more difficult during production of other products. 

The MPF can quickly adapt to other products for clinical or commercial manufacturing regardless of scale and process format. 

Relative overall threat mitigation rating:  MPF > SFSF >> PBF

4. Higher capital investment (BR4)-Excess capital investment might result from building unneeded or unusable ­manufacturing ­capacity, including an inability to manufacture pre-clinical or early clinical products in the pipeline; concomitant with BR3. The risk of spending too much capital is largely dependent on the flexibility of the facility to adapt to the threats. 

Relative likelihood rating:  BT2 > BT1 > BT3.

If the PBF is designed for process and capacity flexibility to handle capacity uncertainty, capital costs increase significantly.  

The SFSF’s flexibility may be limited by ongoing manufacturing, and unanticipated process formats may be difficult to incorporate. 

MPFs have significant flexibility and thus are more likely to be usable for manufacturing new products over their entire lifecycle. 

Relative overall threat mitigation rating:  MPF > SFSF >> PBF

5. Higher cost of goods (COG) (BR5)-Facility is inefficient and results in a higher cost of goods for making clinical and commercial products. This risk can be caused by threats BT1 and BT2; concomitant with BR3 and BR4. 

Relative likelihood rating:  BT2 > BT1 >> BT3, BT4 > BT5.

For a well-defined process with known capacity requirements, PBFs often can provide the lowest COG. However, COG may increase significantly as the uncertainty of the process definition and capacity requirements increase. 

The SFSF can control costs by sharing operating labor within common areas allowing staff to work on multiple LOUs at the same time. However, COG may increase and operating flexibility may decrease due to control strategies additions required for operating commingled processes. 

The COG advantage of the MPF relies on its flexibility to achieve a higher utilization rate under high uncertainty from being able to manufacture many different clinical and commercial products using a wide range of different processes. 

Relative overall threat mitigation rating: MPF > SFSF >> PBF 

Operating threats and risks

Operating risks are important secondary threats to business threats that may ultimately produce business and patient supply risks. The analysis of how operating risks threaten business risks is outside the scope of this analysis. The layout can have a significant impact on mitigating operating threats to prevent the likelihood of realizing operating risks shown in Figure 3. 

The discussion here will be limited to the operating risks caused by the representative operating threats shown in Figure 3. To understand how the layout might control the threats to minimize or prevent the operating risks, the risk process is again the facility layout. For layouts that provide minimal threat control, other ECS elements such as closed single-use systems, procedural, and time-based sequencing controls must be used to minimize the likelihood of the threats producing one or more operating risks. 

The operating threats (OTs) are the result of secondary threats to operating threat processes from equipment, components, procedures, and human operator errors used to operate the process and facility systems. In most enterprises, the largest source of secondary threats are mistakes by operating personnel (7, 8). In this analysis, the secondary threats are assumed to be independent of the layout. The following are the representative operating threats:

1. Facility contamination–other process (OT1)-Facility is contaminated from an external operation such as a second product in a multiproduct operation unrelated to the immediate process steps being evaluated. 

2. Failure of UO changeover (OT2)-Failure to properly execute a lot or product changeout requiring clean-up and removal of single-use components or cleaning of a contaminated stainless-steel system. 

3. Failure to operating UO (OT3)-A failure that occurs ­during normal operation (e.g., a leaking bag or coupling or other equipment failure). 

4. Failure to set-up UO (OT4)-A failure to properly setup a piece of equipment and prepare it for operation (e.g., incorrect connection or damaged single-use bag, etc.). 

These threats, if realized, can result in the operating risks listed in Figure 3. Because of space limitation and incomplete knowledge of the processes, only the likelihood that the threat will result in a significant risk in the context of the facility’s layout is discussed. For this analysis, secondary threats are assumed to be independent of the layout strategy. The operating risks (ORs) are:

1. Product cross contamination (OR1)-The risk consequence is an undetected or undetectable cross contamination of one product or lot with another product or lot. 

Relative likelihood rating:  OT2 >> OT3 > OT1.

The most important control objective is proof of control. The less segregated the facility’s layout, the more reliance on closed single-use systems (a process ECS element) is required for preventing the risk. In general, single-use technologies are threatened by difficult to control human factors during setup, operation, and changeover. Controlling these secondary threats is difficult, but achievable by using both process and infrastructure (timing, procedures, training, etc.) ECS elements. The more UOs are commingled in a single space, the more complex the SRS making the ECSs to prevent both perceived and actual cross contamination more complex.

Relative threat impact likelihood rating:  MPF > PBF >> SFSF

2. Facility contamination (OR2)-An equipment, procedural ­execution, or component failure, such as a leaking single-use bag that results in a contamination of the surrounding facility. 

Relative likelihood rating:  OT2 > OT3 >> OT4.

The extent of the process UO segregation plays a significant role in limiting the extent of the contamination’s impact on other processes. The complex SRS of the SFSF makes controlling the impact of a facility contamination difficult. The complexity of the SRS is a key factor in designing a PBF and an intrinsic advantage of the MPF.

Relative threat impact likelihood rating:  MPF >> PBF >> SFSF.

3. Personnel exposure (OR3)-Extent of exposure and ability to limit, isolate, and remove personnel from operating areas during a contamination event.

Relative likelihood rating:  OT2 > OT3.

The high process segregation and unidirectional flow capability of the MPF is a significant advantage for isolating and mitigating personnel exposure in the event of OT2 and OT3. PBF and SFSF layouts may include additional controls in their design, particularly personal protection equipment.

Relative threat impact likelihood rating:  MPF >> PBF>SFSF.

4. Process contamination (OR4)-Process UO becomes contaminated resulting in the loss of a lot (e.g., bioreactor contamination). 

Relative likelihood rating:  OT2 > OT1 >> OT3>OT4.

The likelihood of a process ­contamination is impacted by the number of threat interactions described by the SRS surrounding each UO. The more threat inputs to the UO, the more likely a contamination could occur (5). 

Relative threat impact likelihood rating:  MPF > PBF >> SFSF

5. Operating schedule delays (OR5)-Failure of one process causing delay of or interference with the operation of another process LOU. OR5 is typically not significant unless it causes a train wreck that impacts multiple processes and products.

Relative likelihood rating:  OT2 > OT3 > OT1 >> OT4.

The more unit operations within a given space, the more procedural and scheduling controls will be required to prevent operational interference. The SFSF has the highest exposure to schedule “train wrecking” if operating threats occur at the wrong time in a large space executing numerous closely scheduled operations. 

Relative threat impact likelihood rating:  MPF > PBF >> SFSF.

Conclusion

The bottom line on manufacturing facility layout risks is the uncertainty factors that impact reliability and utilization. The above risk analysis evaluates the impact of various business and operating risks on the three facility layout strategies. If the manufacturing enterprise knows exactly what will be made over the facility’s lifespan, the PBF with appropriate control strategies tailored to provide the defined process and capacity requirements is probably the most effective and efficient. 

As the process and product uncertainties increase, the SFSF becomes more effective from its capital and operating cost advantages for ­multi-process operation. The higher operating risks of the SFSF must be controlled by comprehensive infrastructure elements (scheduling, procedures, training, etc.) to mitigate a more complex SRS associated with the commingled processes. 

When the process and product uncertainty become significant, the MPF has the advantage of the high process segregation that minimizes the SRS of the individual processes along with the MPF’s ability to efficiently adapt its resources to a variety of different process formats and scales. The MPF has the additional advantage of controlling operating risks that allow the facility to support the entire product manufacturing lifecycle. 

The choice between the PBF, SFSF, and MPF’s layouts should be made by each company to optimize utilization and reliability depending on the anticipated threat and risk uncertainties of the product portfolio and the processes it anticipates needing to operate. 

References

1. M. F. Witcher, “Impact of Facility Layout on Developing and Validating Segregation Strategies in the Next Generation of Multi-product, Multi-phase Biopharmaceutical Manufacturing Facilities;” Supplement to Pharm. Engr. (Nov./Dec. 2013).
2. M. Witcher and H. Silver, Pharmaceutical Technology 42 (9) (2018), www.pharmtech.com/multi-purpose-biopharmaceutical-manufacturing-facilities-part-1-product-pipeline-manufacturing
3. M. Witcher and H. Silver, Pharmaceutical Technology 42 (11) (2018), www.pharmtech.com/multi-purpose-biopharmaceutical-manufacturing-facilities-part-ii-large-scale-production
4. B. Bader,  “Multiproduct Madness–Flexible Manufacturing Facility, Case Study Amgen MOF,” presentation at ISPE Biopharmaceutical Manufacturing Conference, San Francisco, CA, Dec. 4, 2017.
5. M.F. Witcher, BioProcess J, 16 (2017), https://doi.org/10.12665/J16OA.Witcher
6. Witcher, M. F., “Understanding and Analyzing the Uncertainty of Pharmaceutical Development and Manufacturing Execution Risks using a Prospective Causal Risk Model”, Submitted to BioProcess J. 2019.
7. T. Muschara, Risk Based Thinking–Managing the Uncertainty of Human Error in Operations, Routledge (Taylor & Francis Group, 2018).
8. G. Peters and B. Peters, Human Error–Causes and Control (CRC Taylor & Francis, 2006).

Article Details

BioPharm International
Vol. 32, No. 6
June 2019
Pages: 26–32

Citation

When referring to this article, please cite it as M. Witcher and H. Silver, "Comparing Facility Layout Options for Managing Business and Operating Risks," BioPharm International 32 (6) 2017.

Peer-reviewed

Article submitted: Feb. 13, 2019
Article accepted: March 19, 2019

About the Authors

Mark F. Witcher, PhD, is a consultant, witchermf@aol.com. Harry Silver is senior process/facility designer at Harry Silver Designs.