“The key to forensics is getting evidence; analyzing that evidence very carefully; and, most important for an academic situation, coming up with an impartial understanding of what the problem is,” says Jay Shoemaker, principal and director of science and technology for Symmes, Maini & McKee Associates (SMMA) in Cambridge, Mass. “Then all of the parties—users, facilities personnel, architects, and engineers—can weigh in, set aside their own assumptions of how the building worked or what the issues are, and identify exactly what problems have to be solved.”

SMMA was hired by the Massachusetts Institute of Technology (MIT) to renovate Building 39, a 40-year-old structural concrete building that already had gone through several iterations, from data processing center in 1967 to research building with cleanrooms following a 1984 renovation. The building has operated at full bore since then, while maintenance crews came and went, maintenance records and building plans faded from memory, and systems began to deteriorate.

“It was very interesting to come back to the building in 2005 and 2006 and figure out what the original requirements were,” says John Engle, MIT’s project manager. “The past 20 years took a tremendous toll. The building had declined to the point that the researchers would be out of luck with temperature and humidity excursions for many days during the year.”

“This building is a 79,000-sf machine that has been run flat out for 20 years, and it’s tired,” says Shoemaker. “What you really want to do is gut it and rebuild it, but you can’t because it’s too expensive, and because you can’t interrupt the research going on inside.”

Several years ago, MIT identified a need for more cleanrooms and retained a firm whose narrow tasks were to determine how many more cleanrooms the building could accommodate, identify the mechanical/electrical system needs and architectural requirements, and provide a cost estimate for the first phase of adding cleanrooms. That report was completed in 2003, with a cost estimate of $6.4 million. But, by the time funding became available in 2005, the focus changed.

“By then, some of the faculty had changed and the major push became making the building operate as it should,” says Engle. “Creating more research space, although very important, became secondary.”

Creating the Team

Before the process could move forward, before SMMA was even on board, Engle had to gather an internal team to work together toward a common goal. The problem was agreeing on exactly what that goal was.

“It was difficult from the standpoint of gathering research people, maintenance people, and facilities people, getting them all on the same playing field, and understanding what the requirements were,” says Engle. “That was accomplished over the course of about two months of very intensive work.”

That work required dispelling some commonly held beliefs.

“One of the things we learned early in the process was that the users felt that the tighter you made the humidity and temperature requirements, the more likely the building was to operate to those requirements,” says Engle. “From a design or facilities operation standpoint, it was understood that making constraints tighter doesn’t necessarily mean the systems will provide an acceptable cleanroom environment. As a matter of fact, the risk of not achieving them increases.”

That discrepancy was explained to the team by an outside consultant, who remained involved through the entire process because the faculty respected and trusted his point of view.

At the onset of the project a third party cleanroom expert was included in the team.

“Our cleanroom consultant was very good at explaining the reality of what was behind the wall, and how it operated,” says Engle. “In the five to six months we were working on this, we achieved the ability to communicate with the users and operations personnel that if allowed to relax those standards, you end up with a $10 million budget, not a $25 million project. We have done a lot of work to understand why the building doesn’t work today and what as a team we need to design for the future.”

Mapping the Systems

Before you can determine what systems need to be enhanced, you need to know what you already have, and how it currently operates compared to how it was designed to operate. None of this is as obvious as it might seem.

“A building is not in stasis,” explains Shoemaker. “Over time, variation of maintenance routines, new people, new research programs, and new tools changed how the building operated. So, the building began to drift away from the original specifications it was designed to achieve. For example, the air balance between spaces change; one thing affected another.”

Building 39, like most academic science buildings, is a very complex structure: almost 50 percent of the gsf is consumed by mechanical systems; two of the six stories are filled by air-handling equipment alone. System mapping can help get a grasp of the project scope.

“System mapping is the process of making very simple diagrams of systems that have a high risk for the project,” explains Chad Wisler, principal and director of engineering for SMMA.

It’s not necessary, or advisable, to map every system; Building 39, for example, has over 30 major systems.

“If you have a system that is not clearly understood by the whole project team, or you do not clearly understand how the scope of that system changes the project, it is a good candidate for a system map,” says Wisler. “Conversely, if you already are doing a major renovation for one of the systems, there is no need to do a system map because the scope is clearly defined.”

The HVAC, for example, was highly suspect because users reported consistent problems with temperature and humidity in the cleanrooms. The maintenance people knew the air handling units had degraded to such an extent that air was bypassing the coils. The cooling coil condensate was not draining properly and was leaking through cracks in the floor, damaging some of the lab equipment on the floors below.

To map a system, you first need to determine whether it ever operated properly to begin with because many systems in older buildings were not designed, constructed, or maintained properly.

Next, review the existing documentation, including building drawings, submittals, control diagrams, sequence of operations, previous studies, and projects that were designed but not necessarily installed. Follow that by interviewing the people who have worked with the system, from the building users and maintenance people to the original design engineer.

“One thing we found very beneficial was talking with the vendors,” says Wisler. “We often take the vendors for granted, but they have a lot of experience with and knowledge of the equipment that is installed out in the field.

“And nothing beats actually getting out in the field yourself,” he says. “Don’t trust what someone tells you, or what you read on the drawings. There is a good chance that what you see in the field will be different.”

The goal in the end is to create very simple diagrams of the system that everyone on the team can understand and buy into.

Forensic Testing and Diagnostics

The mapping process shows how the systems are configured, but not how they operate. Forensic testing and diagnostics can generate real data to answer those questions.

“Take the system diagrams and build upon them,” says Wisler. “Identify points that you want to measure, communicate that with the team, get the buy-in, have the points tested, and do the engineering analysis.”

Making sure everyone understands the information and buys into it is critical at every juncture before moving on to the next step, he stresses. For example, SMMA generated forms for the HVAC fans that show how much flow rate each had decreased compared to its design specs; any additional information would have been superfluous, says Wisler.

This information can now be used to make informed design decisions, once you understand the constraints of the physical space and the activities it houses, specifically whether those activities can be suspended during construction. Budget considerations should not be considered a constraint at this stage in the process.

“You want to have those design juices flowing,” says Wisler. “You don’t want to be hindered by thinking of budget as a constraint at this stage of the project.”

For Building 39, the team developed a “good-better-best” strategy for presenting options for the HVAC. The bare minimum was to continue a patch-and-repair method to keep the units operating. The better option was to replace the units in kind in their exact same spot, but that would not have solved the space constraints, which affect capacity and shutdown time. That also would increase the risks for the project, because in an existing building, the renovation activities have more uncertainties.

The best option was to install a brand-new, factory-fabricated, custom-built, pre-piped, pre-tested, pre-wired air handler on top of the roof.

“I think most everyone’s gut reaction was that the custom unit was going to be a more expensive option,” says Wisler. “But when you add in the cost of labor to install or patch the units, and start to apply some cost for the risk for the other two options, the third option really comes out to about the same cost.”

“The process was not just a matter of cutting bells and whistles to trim the budget,” adds Shoemaker. “People did not choose the cheapest options, but the ones that give the best value.”

The next step, developing schematic design documentation, requires the cost manager and cost consultant.

“One of the most important points I can recommend is to use their experience,” advises Wisler. “Use their input for the options and as a sounding board.”

The cost consultants helped define categories of expenses that the team could work from, and helped keep downtime for the users at a minimum. Costs were not attached to any of the options until the team had prioritized them in the course of one four-hour meeting.

“We all knew the budget was not unlimited,” says Engle. “We had all agreed to a set of objectives at the beginning of the process, and we were striving to achieve them.”

Keep Communications Simple

At every step in the process, great effort was made to keep the information simple. This not only helped the project team complete its work, but also helped the team present its findings to senior management at MIT, who had to approve the funding.

“On the engineering side of this industry, I think we need to start getting away from reports and thickness of documents and get more to the core of the subject,” says Wisler.

Making the information easily digestible by everyone on the team fosters greater communication among the members, which in the end, produces better work. That open communication helped solve the mystery of the unreliable temperature and humidity controls in the cleanrooms: It turned out that whenever a user renovated a lab or a cleanroom, they capped the exhaust because it seemed like good practice. In reality, that hurt the temperature and humidity control within the space.

When it came time to ask for funding, the team presented senior management with a single sheet that listed the expenses in order of priority, designating which items were recommended, which were not required, and how much could be covered by the original $6.4 million.

“We found that there was an additional $3.5 million worth of work that needed to be done,” Engle says.

“As a tool to put things in front of senior management, this has ended up working extremely well,” he says. “The recommendations from the team were completely accepted.”

An important goal is to make sure this newly gathered information remains current and accessible once the team is disbanded. There will be the usual turnover of maintenance people and users in the years ahead, and it is essential that they know how the building was designed and how it should be maintained and operated. Engle says he plans to conduct classes for the current maintenance staff and create books of flow diagrams that will be kept on site. The documentation also should be maintained electronically in a format that can be easily updated as circumstances change.

By Lisa Wesel

We welcome your Questions and Comments

Copyright 2008 Tradeline Inc.
All Rights Reserved
ISSN: 1096-4894