By Azita Yazdani, P.E. | Founder and CEO, Exergy Systems, Inc.
In recent months, we saw much talk about the demand for chips and the chip shortages experienced due to COVID-19 by the various segments such as the automotive industry. The demand is more significant than ever and expects to grow further due to new AI/IoT and other trends, requiring pressures on the semiconductor industry to produce more chips for various new services and devices.
Furthermore, this translates to exponential growth in more production pressures on existing fabs and their capacities and capabilities. Many companies are planning new production capacity – Intel, TSMC, and others and this trend for expansion is growing across the sector. Intel’s acquisition of GlobalFoundries and/or planning new fabs and its plans to build two new fabs in Arizona and a new fab in New Mexico is a response to such demands. At the same time, TSMC is planning a new fab in Arizona. These global expansion plans only means one thing: even more demand for resources such as water and materials to satisfy the planned production growth.
Amid this activity, the pressures on climate change hover over the entire industry. The semiconductor industry is now believed to be one of the most significant contributors to global warming gasses, which is now surpassing the auto industry (Bloomberg, April 2021). The demand for resources, such as water and energy, puts severe pressures and exposes the industry to severe vulnerabilities that can impact production and competitiveness. Meanwhile, many major players claim they have this under control – by signing on to treaties and other climate commitments, thus signaling the realization that they are aware of these issues and have them under control.
However, after a closer look, much of the fabs and their activities, water, for instance, remains a highly used and wasted material. Explaining why; the fabs use large volumes of water to rinse chemistries and contaminants off parts. A typical fab used from 1-4 million gallons of water a day, like a small city, and while many say that they are “reclaiming” water, their actual water footprint remains unchanged. Reclamation in this context means that fabs reuse costly ultrapure water (UPW) that they produce for production, and after one use, they reutilize them for uses that do not require such clean water – like scrubbers, etc. Moreover, while this is a slight reduction in footprint demand, significant issues are being overlooked. Production of 1-gallons of UPW requires 1.5 gallons of freshwater. The water purification processes currently in use are only 50% efficient, thus resulting in about 50% waste or reject stream from the purification processes. This 50% wasted “fresh water”– which can otherwise be consumed by humans, goes down the drain as waste.
Then we enter the fabs, where 100% of the water used is wasted, continuously – and 24/7. If this water is not “reclaimed” for any secondary uses, which still required UPW production, the waste ends in the environment after treatment in the plant wastewater system.
Some fabs are now talking about how they are recovering some of this water for reuse. However, there lies yet another problem. Central water recovery technologies, commonly used, are also about 50% efficient. This is due to the mixed nature of the centrally collected wastewater, which is created from various operations throughout the fabs containing various chemistry: the “Wastewater Soup.” All fabs are mainly engaged in this way of managing wastewater, a practice that should have long been outdated – and which may include valuable materials, such as metals – copper, cobalt, etc.
The traditional approach is vigorously defended because change is not easy to implement in many existing fabs. The central treatment approach is tied to the current plant/fab design and is engrained in the industry which further enhances the typical use and waste methodologies adopted today, because this is how we built and designed fabs in the past 20-40 years.
While the current state of the fab designs is widely practiced and accepted, there are many opportunities lost for resource efficiency, recovery and reuse. If we do not allow new methodologies and approaches to be embraced, we will set the industry on a path of extreme climate pressures for the future. We not only have to be open to fab redesign, but we should embrace the tool redesign and how they can be modified.
As an example, the current tool designs do not allow technologies, to allow recovery and reuse. Our fabs today are “flow-through systems” – meaning that water and materials are fed and wasted constantly, 24/7 as we believe this is the way to keep the process clean, and impurities at bay. However, with the current state of the planet and availability of resources in many parts of the globe, these practices and tool designs need to be reconsidered. Since the tools in the past were flow-through systems, new tools designs should be redesigned to allow purification and reuse.
The tool production companies are not demanded nor encouraged by their customers; the top-notch fabs to employ the newest and latest technologies to realize this. That allows removal of impurities continuously so that the baths, and rinses are kept contaminant-free, so they are not wasted. The idea that flow-through keeps the process contaminant-free is primitive and is a 40-year-old approach to manufacturing.
So, the questions remain; will we modernize? Or continue to build fabs the same way we have built them for the past 30-40 years?
This is a question many of us in the industry should be asking. Demand for change in the way the industry can manage natural resources better and enhance its ability to be a better global citizen. We cannot ignore the semiconductor industry’s global climate and resource footprint any longer. As member of the industry, WLI and its members should push the industry to adopt more effective and efficient solutions to tackle Climate Change – this largest challenge to the humanity.