by Ken Goldstein, Ph.D.
 |
The cleanrooms industry has made the case that we can use minienvironments to downgrade the cleanliness of our cleanrooms without sacrificing contamination control in the critical places. But this in no way proves that we should do this. The decision to actually do so must invariably follow from a detailed economic analysis—in other words, the benefits must outweigh the costs.
In my last column (CleanRooms, August 2001), we discussed some of the benefits of utilizing minienvironments. Essentially, this boiled down to “building less facility.” We would install fewer filters driven by fewer or smaller fans as well as air-handling units that required fewer or smaller chill water coils and fewer or smaller pumps. With a smaller fan heat load, we would need fewer or smaller chillers, and with fewer or smaller air-handling units, we would need to devote less space and less steel and concrete structure to support these reduced loads. And because we were building “less” facility, we could surely design it and build it faster and less expensively.
Note that these are all capital items. Similar savings would be expected for expense items such as energy and maintenance—so far, so good.
The preliminary financial analyses for this approach indicated savings of 15 percent to 30 percent for the cost of the facility—a significant sum. But this was not to be. A number of significant “additional costs” had not been included in our original thinking: the costs of the minienvironments themselves; the cost of wafer boxes, cassettes, pods and similar containers; the costs of hardware associated with the critical I/O (input/output) functions at the interface between the tightly controlled interior of the minienvironment and the less-controlled cleanroom ambient; and the costs of the automation hardware and software required to move materials.
And as luck would have it, these “additional costs” exceeded the previously identified cost savings. In other words, the total project cost of the newer (dirtier) cleanrooms using minienvironments turned out to be more expensive than standard (cleaner) cleanrooms without minienvironments.
While this did not appear to be a good beginning, people kept experimenting and a few brave souls—mostly in Asia—actually constructed new facilities using these techniques. And they tried to look at the big picture. Specifically, everyone had been looking at the facilities-related cost savings and weighing those against the tool-related cost additions.
The predictions of faster design and construction schedules turned out to be correct. While important, this was not enough by itself to sell users on the concepts of minienvironment or barrier/isolation technology. But at this time, people began to consider the process advantages of this new technology. Here is where they discovered the true rewards that more than made up for the increased costs.
It turned out that the physical barriers really did work and performed excellently at their intended function—they effectively isolated the sensitive product and process from aerosol contaminants. Numerous studies demonstrated that with proper design and construction, minienvironments could supply clean air that was essentially particle free. This improved contamination control led directly to superior yields; and in the microelectronics industries, this alone guaranteed great interest.
It was then that users began to notice the interactions among the facility, the tool set and minienvironments themselves. During the start-up of a new manufacturing plant, individual tools are installed and started up, often one at a time. As soon as a tool is installed, engineers begin testing it; and as soon as testing is complete, operators begin using it for production. At the same time, other nearby tools are being moved into place and installed and started up. This process continues throughout the life of the facility as older tools are replaced with newer ones.
The problem is that tool installation is, by nature, very dirty because it is a construction activity.
Almost invariably, these tasks involve cutting, drilling and grinding as well as other particle-generating activities. And here too, the isolation provided by the minienvironments performed superbly. It was quickly discovered that the minienvironments effectively isolated the production activities from nearby construction tasks involved in the installation of other tools. It was now possible to keep turning out product with little regard to nearby activities that would normally be considered to be highly contaminating. The result was faster production ramps, higher tool uptime numbers and fewer tool shutdowns to accommodate tool installations.
With higher yields, faster production ramps and fewer shutdowns, it seemed that the wave of the future had arrived. And to be sure, the manufacturing firms that opted for these types of cleanrooms were convinced—and kept building them.
But again, most readers understand that while common, minienvironments have not completely taken over—quite to the contrary. It has been a long slow process in America. We will look at why and offer a few final comments on the subject in our next installment.
Dr. Ken Goldstein is a principal with Cleanroom Consultants Inc. (Scottsdale, AZ) and is a recognized expert in planning and designing of cleanrooms and ultrahigh purity systems. He has been associated with the cleanrooms industry for 20 years, and is a senior member of the IEST. He is active in WG-012 (Cleanroom Design) and WG-028 (Minienvironments).