Cutting through rocky waters

Atmel Corp., a Colorado Springs manufacturer, knows that maintaining effective water-treatment-system operations is crucial.

By Ronnie Mills, CPMM 

PlantServices.com

Founded in 1984, Atmel Corp. is headquartered in San Jose, Calif., with principal manufacturing facilities in Colorado Springs, Colo.; Nantes and Rousset, France; Heilbronn, Germany; and North Tyneside, England. Atmel designs, develops and manufactures semiconductors for nonvolatile and embedded memory, logic, analog, micro-controller products and system-level integration.

The company employs about 1,800 employees at its Colorado Springs site. With a staff of 77, the site facilities team is responsible for environmental, health, safety, security, engineering, buildings and grounds, operations and maintenance. The primary function of site facilities is to provide sustaining services in support of the site's manufacturing, memory test and administrative functions.

Manufacturing in Colorado Springs requires a continuous supply of large volumes of high-quality reverse osmosis deionized water (RODI). During peak manufacturing, the site easily can process more than 500 million gal. of water in a year. Thus, raw feedwater quality affects the ability to maintain critical water specifications. Managing the effectiveness of water treatment system operations is paramount.


Figure 1. Removing minerals raised the cycles of concentration (Cycles), eliminating scrubber scaling and saving millions of gallons per year (MG/yr).

Damped down bottom line
The manufacturing site faced several water quality issues simultaneously. First, the water treatment process was obsolete and not well managed. The vendor's inability to improve the existing program to accommodate complex and changing water conditions exacerbated an already difficult situation.

Second, the most severe Western drought in three years (and since the dustbowl days of the 1930s) was worsening. Aquifers and reservoirs reached all-time lows, thus degrading water quality.

Third, the City of Colorado Springs Utilities (CSU) was forced to activate Pinello Wells, a regional source of water infrequently used. Although CSU's hydraulic model indicated this "blend" would solve the flow/delivery issues, the reduced water quality was becoming detrimental to Atmel's bottom line.

Because of the implications to manufacturing, Site Facilities formed a team to resolve the water issue as quickly and effectively as possible. The first step was to contact chemical water treatment companies for bids. The vendors' bids detailed the technology that would counteract the water chemistry changes and included an implementation plan that would not jeopardize the manufacturing effort. The major factor in Site Facilities' final selection of Ondeo-Nalco (ON) was the vendor's clear, concise plan that supported implementation well within project deadlines.

The team evaluated an array of factors to integrate ON's plan into the project work process. These included, among other things:

• The current system's level of operation.
• The expected rate of water degradation.
• Support of critical production.
• Return on investment (ROI).
• The complexity and learning curve for new technology.
• The ease of incorporation into existing predictive/preventive maintenance workloads.
• The potential effect on production equipment.
  
  
  Figure 2. Reduced water consumption adds up to a projected annualized water savings of 53.5 million gal. per year, benefiting the water-depleted region and helping the City of Colorado Springs Water Department supply higher quality.
  

Target: scale
Once the Atmel/Ondeo-Nalco (AON) team completed its assessments, the implementation process began. The newer, less congested systems in Building 5 (Fab 5) were the most critical and the easiest to modify. The problem there was calcium-based scaling of the polypropylene spherical fill media (SFM) in the building's acid scrubber system. Scaling decreases system efficiency and SFM life expectancy. The existing chemical treatment system required high concentrations and frequent treatment to minimize deposition of hardened precipitate.

A new water softener eliminated the scaling problem as well as increased the cycles of concentration (COC) (Figure 1), thus increasing the ROI for chemical treatment, resource utilization, fill media replacement cost and overall water consumption. Reverse osmosis was evaluated, but at $8 to $9 per 1,000 gal. of water, was deemed cost prohibitive.

Cycles of concentration can be based on any constituent in a water system. COC is generally calculated either on the basis of specific conductance or some infinitely soluble ion that isn't being added via chemical treatment. Simply put, COC is a measure of the system's ability to reuse water with reduced blowdown frequency. A reduction in COC, on the other hand, is attributable to an increase in scalant concentration (sulfate, calcium or silica). Because these constituents work in concert and it is rare that the concentration of only one constituent changes at a time, we use the ON "Microcape" program to model the solubility of any potential scalant as the water changes.

The second application the team chose was a sulfuric acid injection system (SAIS). This effort proved to be considerably more complex and costly. The SAIS minimizes the effects of high alkalinity/calcium water sources and increases the overall cycles of concentration. The average COC was 2.5 before the SAIS went online but, after commissioning, is now in the 7 range.