Look for some more sites on Recycling>>>>
Google

Recycle- with EDI - virtually all semiconductor process water!

If you want to give your company a big cost reduction project, help the community it's in, and define your image as a successful DI or Process Engineer, this page is for you.

If you're responsible for a cleaning operation for semiconductors, electronic assemblies, flat panels, optical components, or anything that uses or -could- use DI or UPDI water; or maybe you're a DI technician, a facility manage or engineer, an EH&S environmental liaison--now is the time!

Or maybe you have a problem with building new facilities because the local government says you'll use too much water?

For most 'high-tech' businesses, where cleaning parts with softened (or better) water occurs, this web site offers a comprehensive answer. You won't believe how easy it is, and how much money it will save.

Full Disclosure: I'm not selling anything! I don't get a penny from any vendor, spam, city, water company, or relatives. I just think this recycling thing is very, very hot. You'll see why. If you agree, you can get some more info from these friendly people. Even if it's not EDI or water related, it'll enrich your life!

Electrodeionization --very high recoveries, return on investment!

Recycling water to the extreme in our semiconductor fab has had very strong benefits. Here are the numerous reaons why we did it Here's the environmentalists' view.

But that's only one viewpoint, and not the best one. More later.

First, we think EDI can help recycle water better than any other technology. We couldn't find any published info on EDI in a recycle system, so we designed our own system with only little help from the equipment vendors.

Recycling---Study up! It's On Its Way!

Some references first--What's this all about? This is some background on recycling in industrial applications. There are obvious winners and not so obvious losers. The ERC website at the University of Arizona course (you might not be able to view it--ERC copyrights--so you'll have to trust me) has the leading edge thought on recycling. The next generation of engineers will know why and how to recycle as much as possible. In the meantime, we have to wait for equipment suppliers to catch the drift. Even the latest article (12/04, way down in the table) -totally- misses the importance of recycling. The first reference in the following table gives seven good reasons to recycle and some examples of recycling in various industries. They unfortunately end with examples of where they think recycling is a -bad- idea. Hard to believe. I think they should have spent just a little more time on an ROI study. It's a good study of the recycling advocate's current thinking, though.

Seven reasons to recycle...or not. I'm confused. This is actually a good case study for recycling, but misses the uniqueness of semiconductor water use. The discussions would be clearer without the emphasis on sludge, which is somewhat irrelevant to water recycling. Do you think the last article is rather dated? Dilution is the solution?

SEVEN GOOD REASONS TO RECIRCULATE YOUR RINSE WATER

Oops--they pulled the article from the website. Wait! Google cached it, but no pictures. I grabbed a copy. Wait again! It's back! Oops! Gone again. cache

The Semiconductor Environmental, Safety & Health Association is, of course, a big fan of recycling. John DeGenova, PhD, presented some of his history of success with recycling at Texas Instruments. There's also discussion of software used to model water system (including recycle) design. This is a great reference.

The UPW Recycle Strategy presentation, here.

  Here's a detailed discussion of a semiconductor application for recycle. It's unfortunate the author's feelings toward recycling ("The last (i.e., 'least') potential use for the recycled water is in the UPW system.") is contrary to what is happening, what makes the most economic sense. You read it:

"As much as possible of the wastewater collected should be utilized in reclaim applications, such as in cooling towers, boilers, and scrubbers. Another high-priority use of the water, which should be quantified, is in non-wafer-contact processing, such as tool cleaning, quartz sputtering, and glass cleaning. Use in fab humidification systems is also a possibility. "

Well, -I- think wafer-contact is where the savings and performance are. Time to advance the science, not retard it.

 Designing practical DI-water recycling systems for use in semiconductor fabs

The article in Micro Magazine.

  Here's yet another detailed discussion of a semiconductor application for water efficiency. Once again, the difficulty of recycling back to the Fab clouds the accurate portrayal of the real gains good recycling can do. To quote,

"If possible, return the final portion of rinsewater flows to the circulating deionized (DI) water return loop. It may be feasible to re-treat this water at the DI water generating plant less expensively than incoming municipal water."

Now, does that get you excited about recycling? I doubt it. 'Return to the fab' is last of 14 bullets that offer rinse (10 bullets) and recycle options. From there, the dozens more bullets hilite cooler, restroom, and landscaping uses. I'd like to see this otherwise comprehensive article updated to include EDI results.

 Energy and Water Efficiency for Semiconductor Manufacturing

The article in the Pacific Northwest Pollution Prevention Center web site.

 Oki has done some interesting work on a completely closed system UPW operation at one of their semiconductor plants. It's complex and encompassing--I think so much so that you'd never convince management to do it. It appears the water is reused, not recycled. Un-reused water is evaporated.

 Introduction of Ultra Pure Water Close System into Semiconductor Plant

The article at the Oki website.

  Sandia National Labs (SNL) started a project to evaluate EDI's from several vendors. All finally backed out but one. Despite problems with the unit, there were some good numbers coming out of this work.

Recycle Spent Rinse Waters With EDI Careful--they move the article around on the SI website.

Here's some more info on the same project, using CDI:

Performance of a Treatment Loop for Recycling Spent Rinse Waters

 

 Recycling CMP water--this looks like a great integrated system for removing particulates & recovering the water. It uses an E-Cell EDI after the microfilter. >95% recovery of water is the claim. Great! Kinetico Engineered Systems Division.

Click the link.

 

  Sandia National Labs "co-sponsored a conservation workshop for industrial, commercial, and institutional water users on March 29, 2000 in conjunction with the New Mexico Water Conservation Alliance and Intel. The topic of the conference was water conservation in semiconductor fabrication plants." Good, recent work. Check out Kurt's recycle system--including the EDI in the loop!

Kurt's 'Reducing Water Usage in Semiconductor Manufacturing'

The Workshop

 

  This PowerPoint is a comparison of water savings estimates made by (John) Rosenblum Environmental Engeneering in year 2000 at Fairchild with actual 2002 results achieved by the process engineers. In hindsight, the estimates seem REALLY conservative, John! The 'real work' was done by Greg Klusewitz and Jim McVeigh. These are great savings in both water and money.

John's 'Improving Fab Water Efficiency: Results from Fairchild Semiconductor, Mountaintop, PA'

  The article in the current (Nov. 2004) UltraPure Water by Thomas Fulde, "Part 1: Meeting Water Requirements for 300-mm Fabs" is apparently the Americanized version of the original. Mr. Fulde has experience in Shanghai and Singapore, and I would have thought that recycling water would be a top priority for him. Indeed, the UPW diagram in the article shows the MAIN source of water for the system is from a box called "Reclaim System Appr. 80% of Consumption" (should be Recycle, not Reclaim). But if you look at the water analyses before and after the water injection point, they are identical! Is the Reclaim system turned off? It must be. There's NO mention of the system in the entire article. The whole article can be summed up by a sentence he used to describe the use of UF: "this technology and the used products are relatively standard and typically perform within their design parameters". The only new part is the hint of an awesome recycle system, and it's ignored in this installment, and judging by the description of Part 2, apparently that article as well. When are We going to focus on the inevitable?

There's no link to click. If I find a non-subscription version, I'll connect to it.

  The current (Dec. 2004) issue of UltraPure Water has the second part of the article by Thomas Fulde, "Part 2: Meeting Water Requirements for 300-mm Fabs". As promised, Mr Fulde failed to talk about the recycle system that supplies most of the water to his treatment system. He does mention it, though, after describing the difficulty of getting the TOC level below 0.5ppb. There were many issues with the ACF (activated carbon filters), the UV lamp types, membrane degassifier systems, and RO rejection. He got lucky with higher-than-anticipated THM rejection in some of these areas. The VERY understated punch line--"The system constantly shows numbers between 0.35 and 0.4 ppb (TOC) under full load conditions and of course much lower in recirculation mode". How much lower? And, how about that Boron problem--how much did THAT improve? Ever think about EDI in the indescribable Recycle system? In the closing paragraph, there's mention that the overall performance included "a very well defined Reclaim (Recycle!) plant that could be discussed further." Please Discuss It Further!!! It's more important than UV lamp pressure and degassifier flow rates!!!!

There's still no link to click. If I find a non-subscription version, I'll connect to it.

-----------

Back to EDI: What's it for?

You'll see it's common in a primary loop set-up: RO, EDI, polish resins. The links below will show you equipment that can fill this need. But you won't find it recycling any water--an unfortunate lack of data.

Back to my project: After reviewing the contenders, it was time to select a vendor of EDI to supply us one to evaluate in recycling. The only real choice was Ionics EDI. Their specs met our needs, and right off-the-shelf. It looked safe to say that this experiment could lead to a breakthrough for the next level of recycle recovery.

Management, hindered by outdated accounting fears, had trouble accepting the free-for-a-month EDI, with soft, friendly terms like 'if it meets your specs you can buy it...or not!'! But in October 1998 we were finally underway with the Ionics EDI project. Here is the project as it unfolded, including some rather abrupt slowdowns right from the start.

  1. First, could the technology do what we'd like? That is, recover conductive water and make it look better than RO product water? Our goal was to get 500 Kohm water from 250, 500, or 1000 uS (that's only 1000 ohms!) water. Our resin-based recycle system at Labs had a conductivity setpoint around 25K. The average of recovered water was around 100K. Some vendor data suggested we could get about 1 meg water out of the EDI with a 4K (250 uS) input conductivity. Thus, with proper storage and mixing, we would be able to handle water that was about 25 times more contaminated than we currently recovered.
  2. Early results from Sandia National Labs' recycle program indicate that their water was at a contamination level that the Ionics EDI could handle. Their average conductivity was in the 3K-5K range. Look for the blue diamonds. 8/3/98
  3. To be fair, I looked at other vendors. It wasn't pretty. Other EDI vendors took a negative stand, and in rather bold statements, advised against this strange idea. This email exchange , from the R&D Manager of a manufacturer of EDI, sees nothing but trouble. 8/17/98.
  4. This negativity actually sweetened the pot. Here's a simple plan of attack for screening tests of the EDI. 10/14/98.
  5. Got the unit, but still more questions. The manual didn't help much, but the email did. Here's the dialog.

A little time passes, say, a month or two.....

December 1, 1998. EDI had been helping us recycle water for several weeks. Time to write a paper. By March 99, we had lots of info. Here's a paper summarizing the results up 'till then. Click to read. There would be more performance data as it arrived.

September 17, 1999. Everything (our fab) was shut down. But there's more information to read about. Organics and some extreme overload (test only!) conditions. Insults to the system were easily detected, averted, diluted beyond recognition.. Read the last installment here.

I was invited to speak to the 1999 Workshop on Disaster Prevention/Management & Green Technology, sponsored by the National Science Council of the Republic of China. Fortunately, it was held in Foster City, CA, and not way the heck on the other side of the world. Here's the slide presentation--if you found this site by searching for my keywords, it'll all make sense.

Mail me whatever at (fix the address first!) this address

Page Counter