After 20 years, the clean-in-place (CIP) operations in Building B at Emergent Biosolution’s Winnipeg, Canada, facility were ready for retirement. Built in the early 1990s, the systems had reached their life expectancy and were now out of date.

Building B manufactures human hyper-immune plasma products such as Anthrasil [anthrax immune globulin intravenous (human)], VIG (vaccinia immune globulin), and several contract products. The facility’s two CIP systems cleaned its process and buffer tanks, flex hoses, and process piping—essentially everything except the UF skids and chromatography columns.

The company’s practice is to operate cleaning and validation cycles to meet government regulations and ensure all product has been removed from the system before next-batch processing. While the systems were working properly, it was time for them to be replaced.

Like most pharmaceutical companies, internal competition for capital expenditures is high and resources are limited, requiring creativity and unconventional thinking to win approval for the new project.

First, it was decided to replace one CIP system while keeping the other original system operational. This would allow for continued manufacturing products while implementing the new CIP technology, plus help with the challenges of sample collection and validation.

Second, the design and development work that went into the original system would be leveraged, while adding the capability to make minor improvements.

Third, the latest automation technology would be installed to reduce cost and ensure flexibility for future expansion.

With that three-part strategy, the project was approval and the work began.

Objectives

Project goals centered on speed of execution, cost, and risk reduction, and included the following:

• Retain the functionality of the original CIP system
• Limit the design, commissioning, and qualification times
• Provide a seamless transition for operators that required little training
• Reduce project costs

With those objectives in mind, the project was launched with two partners. Lakeside Process Controls supported software application in Winnipeg and built the control panels in Toronto. Holland Applied Technologies fabricated the process skids at its facility in Chicago.

Automation Design and Development

The goal was to reuse the original CIP system’s software to save time and cost. To start,  all of the class-based code from the system’s Emerson DeltaV software was leveraged. This included the physical and procedural models defined in standard ISA-S88 batch process control; the physical model consisting of control modules, equipment modules, and units; and the procedural model consisting of phases, operations, unit procedures, and procedures. The original graphics also were leveraged. 

In addition, the piping and instrumentation diagram (PI&D), user requirement specifications (URS), and functional requirement specifications (FRS) were retained with a few minor modifications.

Prior to the project, the DeltaV software on Emergent’s CIP systems had been upgraded to Version 12 with CHARM architecture, enabling the project to incorporate Emerson’s ASCO Numatics 580 Series CHARM node to control the CIP system’s 35 process valves.

By directly linking the 580 CHARM node to the DeltaV system via the CHARM baseplate, the node allowed the pneumatic I/O to function the same as electronic I/O— simplifying pilot valve manifold commissioning. Panel space also was saved, and the number of network interfaces and gateways, wiring, and junction boxes were dramatically reduced.

Instead of installing 35 additional wires between the I/O bank and the pilot valves, the CHARM 580 node did it with one wire, plus its integration with DeltaV Explorer made the programming faster and easier. The node also brought redundant communication and power connections to the pneumatic valve manifolds.

Validation and Commissioning

Once the software work was completed, a Lakeside team brought their qualified technology to Holland Applied Technology in Chicago for factory acceptance testing (FAT). The goal was to qualify the entire CIP system before shipment to Emergent’s facility.

In one week, the skids’ piping, materials, and instrumentation were tested with the automation system to ensure everything was wired and operating correctly. None of the class-based code required testing, because it had already been validated. If the project had involved newly designed skids and software, FAT time would have been tripled.

The qualified CIP system was dismantled in Chicago, shipped to Emergent’s building in Winnipeg, and installed in its final location. A quick site acceptance test (SAT) was conducted that focused only on instances. For example: Were the wiring and voltages correct? Did the I/O check out? Did the pumps have the correct rotation?

The skid manufacturer’s turnover package for the hardware components cut qualification time to one week. The software SAT, which leveraged Lakeside’s work, also required only 1 week. Because we used the original CIP software’s space-based logic, (run, hold, stop, abort, and restart logic), none of it required revalidation or testing.

An existing validated base CIP procedure was employed to create new procedures. While a good cleaning sequence was in place, a few days were required to optimize the system’s operation. Parameters such as return pump delay times, proof of return times, and final conductivity alarm times, required some fine tuning. Product qualification was performed in a way that did not compromise the GMP manufacturing process.

Benefits and Results

With the new CIP system operational, aggressive objectives were achieved.

• All the functionality of the original CIP system was retained, but with significantly more flexibility for future modification and expansion.
• Overall, project design, validation, and commissioning required about 4 weeks—less than half the time of an entirely redesigned system. The speed was attributable to three factors: the use of the original PI&D, URS, FRS, and class-based code; the collaborative relationships and trust forged with Lakeside and the skid manufacturer; and the FAT at the skid manufacturer that qualified the system before final installation in Winnipeg.
• Since the original system’s SOPs and work instructions were reused, the operators required little training. Procedures were simplified. Now, an operator starts the CIP cycle in the DeltaV control software and walks away.
• The Emerson CHARM technology provided design flexibility. For example, having the flexibility to add I/O without having to worry about backplane limitations was a benefit. It enabled Lakeside to do panel design without accounting for all of the I/O. In addition, the ASCO valve bank reduced the number of CHARM cards, allowing the operation of both the existing and new CIP systems during the transition, plus providing for future CIP system distribution.
• Auditors have an easy package to review with system designs, qualification documentation, SATs, and version control in the software.
• About 1,200 hours of internal design time was saved by leveraging the original CIP system’s PI&D, URS, and FRS. This represented a savings of about $60,000.¹
• Automation design required 80 percent less vendor time than if the project had called for all-new URS, FRS, CIP cycle, class-based code, equipment and control modules, and software design programming and testing. The result was $280,000 in cost savings.²

Overall, the strategy to leverage the design and development work from the original system, plus incorporate CHARM technology, eliminated 25 percent of the project’s total cost. In addition, CIP functionality was maintained, operator productivity improved, and system flexibility was enhanced for future growth.
___________________________________________________________________________

References:

¹ Nominal internal hourly rate of $50 per hour

² Nominal consultant hourly rate of $100 per hour