The complexity of the Frederick Manufacturing Center (FMC)
Expansion construction project undertaken by MedImmune made being on time and
on budget a challenge from the very beginning. Aggressive timelines and phase
overlaps meant that the project team had to think outside the box in order to
meet the company’s goals - developing extraordinary solutions from ordinary
tools was one of the project themes.
The project team drew on the military background of some of
its members to develop thorough and efficient operator training and
risk-management techniques. Additionally, MedImmune called on experience gained
from previous projects in order to fine tune project management methodologies.
But it was the unwavering attention to safety that really impressed the
Facility of the Year Award (FOYA) judges, helping the FMC Expansion project get
over the line in the Project Execution category.
Safety First
To have no lost-time accidents during 2,300,000 man-hours is
an amazing feat, and impossible if project managers don’t take ownership of
safety.
“An owner has to drive safety. There are a lot of firms that
somehow believe that it is the responsibility of the prime construction
manager; we simply have a position that a lost time incident is an unacceptable
event,” explains Andy Skibo, Senior Vice President of Global Engineering and
Facilities at MedImmune.
“You have to engage a [construction] firm that can manage
that way. They will perform better for an owner who has that drive than they
will for an owner that says ‘you guys go mange that,’” adds Skibo.
Progress meetings provided Skibo with a platform to
reemphasize the priorities to team members.
“When we report project progress, we don’t report cost or
schedule first, we report construction safety first. Only when that data is
available do we go into cost and schedule - it’s known that there are no gold
stars handed out for good cost and schedule performance if the construction
safety is not up to snuff.”
But the real key to such an outstanding safety record,
according to Skibo, is the workers themselves.
“We don’t hand out construction safety incentives to our
prime contractors, but we do hand out incentives to the actual workers. If the
workers themselves don’t want to work in a safe environment, there’s no way of
achieving two million work hours without a lost-time incident.”
Skibo also had a deeper, more ideological incentive for
safety on this project: “It’s inconsistent to deliver pharmaceutical products
for human health and not worry about the people that work for you.”
The Problem: Training & Risk Management
MedImmune had over 100 staff members that needed detailed
training on the use of the new process control system (PCS). Switching from a
small-scale, semi-automatic facility to a fully-automated, multi-product
facility meant that the job functions for a large number of employees would be
altered significantly.
Skibo was acutely aware of the problems that can come from a
lack of staff training and transition planning.
“MedImmune had experience with a pilot plant facility - a
brilliant physical project, on time and on budget, and a brilliant technical
facility when it was done, but there was a difficult transition for our
development team moving from a laboratory, bench-scale environment to a
fully-integrated, two-story development facility.
“So for the FMC expansion project, when we were still a year
away from mechanical completion, we wanted to take advantage of that experience
and focus on training of the entire team to ensure that the baton wasn’t
dropped between the physical completion of the project and the start up of the
manufacturing environment,” says Skibo.
The Solution: A Military Approach
Working with Skibo was Patricia Rader, Executive Vice
President at Raland Technologies, and Brent Hill, Director of Automation Global
Engineering. Rader and Hill had military backgrounds and saw several
opportunities to use this experience to improve project execution.
The project called for a “vertical cliff of startup” between
construction and manufacturing operations, not a smooth transition over many
months. Figuring out how to design the startup, according to Skibo, is only
half of the battle. The more important question was how to undertake a
failure-mode analysis and failure response when things do not go as planned.
Rader and Hill immediately recognized this situation as a
civilian version of advanced mission planning.
“Much of mission planning is figuring out how to workaround
problems and not just ‘how do I do it if everything goes perfectly.’ That’s
where the military training came into it,” explains Skibo.
The military training methodology consisted of a four-tiered
approach:
• Concept training.
• Review of operational standard operating procedures
(SOPs).
• Hands-on, instructor-led training.
• On the job training.
During the third phase, trainees used a complete, isolated
replica of the PCS to not only familiarize themselves with the software, but
also to prepare for problem scenarios that may arise. The system was built
using Rockwell’s SoftLogix software package and allowed training and validation
activities to be performed simultaneously.
“We knew, given the complexity of the plant, that once
you’ve qualified an area, if you have something in the area that needs to be changed,
the process is very demanding. You have to manage the change. It doesn’t leave
you any room to run through 15 ‘what-ifs’ to figure out what needs to be
changed. This is where many startups like this run into time delays. The
solution isn’t obvious, and the only way to figure it out is to play with the
equipment. If you’re in a validated state, that can set you back a month or
two.
“From day one we said we were going to have a training room
with two full PCS simulators that look, to the operator, just like the real
thing. It was a completely off-line computer system backed up with training
modules that could run a synthetic batch or problem - the operator could put in
a problem and try various solutions.” says Skibo.
Mastering Phase Overlaps
If there was any room for improvement on this project, Skibo
sees the potential for a deeper understanding of the non-sequential integrated
commissioning and qualification (ICQ) process.
“The area that we think could be improved upon is making
sure that we have, throughout the company, a clear understanding of what the
ICQ process is. It’s much different to the sequential process where we
commission first, then start installation qualification (IQ), then operational
qualification (OQ) and finally performance qualification (PQ).”
“ICQ overlaps a lot of commissioning and the IQ stage in
particular, and it overlaps some of the OQ. The goal is not only to overlap the
schedule, but to not do the same task twice. That has to be understood top to
bottom by the engineering, construction, manufacturing, quality and
qualification groups,” explains Skibo.
Recognizing the need for continuous improvement is the first
step towards another best-in-class facility construction project for Skibo and
MedImmune.