The Incredible Shrinking Foxconn Plant

More details were released by Foxconn regarding the plant it is building in Mount Pleasant, and two things are clear: the project is tiny compared to the original plan announced by the company, and there are still more reasons to doubt this will be an LCD manufacturing plant. In fact, it may not be a manufacturing plant at all. 

The company provided detailed plans to the state Department of Safety and Professional Standards, which in turn released the construction documents to the media last week.

The planned “LCD manufacturing plant,” as Foxconn continues to call it, would have an 881,549-square-foot footprint. In addition, there are plans for some auxiliary buildings related to chemical storage and wastewater treatment. And Foxconn also built a 120,000-square-foot multi-purpose facility last year.

But all this probably won’t total much more than 1.1 million square feet — or about one-eighteenth of the 20-million-square-foot campus Foxconn originally planned to build. And yet state and local governments are spending $1.6 billion to create infrastructure and connections for a project that looks like it will create a tiny fraction of the promised $10 billion in capital investment and 13,000 jobs.

The base for the plant is still projected to be made of crushed compacted gravel, which cannot support a LCD fabrication plant, as Harvard Professor Willy Shih, one of the few U.S. experts on LCD fabrication, has told Urban Milwaukee. 

“A compressed gravel foundation might be fine for a normal industrial building, but it’s probably not an LCD Fab, which has to have a massive steel infrastructure to support a vibration-free environment for equipment that has to do ultra-precision (manufacturing),” Shih said. 

That steel support substructure is no small undertaking and could be up two floors deep in LCD plants — and nothing like that is specified in the documents.  

As a story Shih did for Forbes about LCD plants noted, “an LCD fab typically uses far more structural steel in its construction than a typical skyscraper… I visited a Gen 5 fab in Taiwan in 2003, and the plant manager there told me they used three times as much structural steel as Taipei 101, which was the world’s tallest building from 2004-2010.”

Then there’s the fact that the plant specified in the documents will have just a single floor. As Shih wrote for Forbes: “LCD fabs are multi-story affairs. The main equipment floor is sandwiched between a ground floor that is filled with chemical pipelines, power distribution, and air handling equipment, and a third floor that also has a lot of air handling and other mechanical equipment…. When they bring the manufacturing equipment in, they load it onto a platform and hoist it with a crane on the outside of the building. That’s one way to recognize an LCD fab from the outside – loading docks on high floors that just open to the outdoors.”

The LCD plant in Hwa-Ya Technology Park, Taoyuan, in Taiwan is one story below ground and five stories above ground. And Sharp’s LCD factory in Sakai, in western Japan, is also five floors high.

The typical construction and equipment cost for the latest-generation LCD fabrication plant in Asia was $4 billion in U.S. dollars, a research paper on the industry noted. Shih estimates that a real Gen 6 plant, if it was built in Mount Pleasant, would have a price tag of around $5 billion.

But there is no sign Foxconn will spend anywhere near that amount of money. It has awarded just $13 million in contracts for the new building.

Based on the plans Foxconn has released Shih says the plant being built “could be a clean manufacturing facility for assembling LCD sandwiches into panels, or assembling panels into monitors or TVs. The floor and structure are not sufficient to support a real LCD fab” manufacturing plant.

Bob O’Brien, co-founder and president of Display Supply Chain Consultants, and a long-time observer of the industry, offers a different view. He believes the plant could be a Gen 6 fabrication plant, but on a very small scale, just a fraction of the size of the original proposal. (And far smaller than the kind of plants built in East Asia.)

The original $10 billion plan, he notes, was to manufacture Gen 10.5 glass panel sandwiches (two sheets of glass sandwiching the liquid crystal material) about 10 feet by 11 feet in size. While Foxconn never said how many it would produce per month, the $10 billion figure suggests the plant would be manufacturing 120,000 “substrates” or glass panel sandwiches, he estimates. 

For the downsized plant, O’Brien says, he is hearing from sources in Asia that Foxconn does plan a Gen 6 plant that would manufacture glass panel sandwiches of about five feet by six feet in size, but would manufacture only 7,500 substrates a month. That’s about one-twentieth the output of substrates originally planned, and those substrates are themselves half the size, suggesting the total operation could be 40 times smaller in scale than what was originally proposed. (Foxconn did not respond to any questions from Urban Milwaukee about these or other details.)

This could allow the plant to produce LCD products for the U.S. auto industry and medical industry, something Foxconn officials have talked about. “That makes sense,” O’Brien says. “If you’re going to be doing something in the U.S. you want to make something uniquely suited to the U.S.” 

One thing both O’Brien and Shih agree on is that Foxconn’s revised plan will get nowhere near the $10 billion capital investment and 13,000 employees the company originally promised. Foxconn has promised that this drastically down-scaled plant will be finished by late 2020, which will be more than three-and-a-half years after the original plan was announced.

If you think stories like this are important, become a member of Urban Milwaukee and help support real independent journalism. Plus you get some cool added benefits, all detailed here.