In 2002, Abbott Laboratories was having issues with the pretreatment of purified water at its Abbott Park, Ill. facilities. The company has six vapor-compression stills at the site, providing qualified purified water to its pharmaceutical finished goods and medical device operations, as well as R&D labs. Abbott couldnt consistently meet microbial requirements at the stills sample feed points, and it couldnt discover the root cause of the problem.
The pretreatment system for each still was comprised of duplex water softeners, carbon filters and a thermal treatment unit (TTU) that pasteurized Lake Michigan feedwater at 250º F through steam and chilled water heat exchangers. The elaborate process required multiple sanitizations per week and added complexity to operations, and contamination issues persisted.
We needed to do something, says Jose Rodriguez, P.E., Abbotts energy manager. The system was a clear operational and maintenance burden, and presented safety concerns as well, since water had to be maintained at 250º F and 55 Psig.
Abbott tried several alternatives, each of which had limitations or created new problems:
- Heating dead leg excursions to sanitize piping locally. This approach controlled microbial activity in the dead leg piping, but not throughout the system.
- Cooling the recirculating pretreatment water, using existing TTU heat exchangers. At cooler temperatures, this practice led to gasket leaks.
- Injecting sodium bisulfite, which led to concerns about added substances and resulted in the need to control within the sodium bisulfite itself.
Abbotts team next looked into buying new equipment. A reverse-osmosis (RO) system was considered, but would have been very expensive.
Instead, the company decided to install 5-micron particulate filters for contamination control, and to rely on an out-of-favor technology, ultraviolet (UV) lamps, for dechlorination. Previous attempts to use UV had been unsuccessful, primarily because the lamps were not built with dechlorination in mind, and vendors had little experience in this area. However, over time the technology had advanced considerably, and better solutions were available.
To ensure that microbes are kept out of the pretreatment systems, Abbott conducts an exercise in which it opens up all valves, shuts off power to the UV equipment, and flushes the systems with free chlorine. Initially, the company performed this exercise every month, but after five or six months, it still experienced microbial activity.
The flush is now done weekly, and so far has worked well. We thought we would have more trouble than we did, says Rodriguez.
The process has also met the approval of Abbotts quality personnel, although the team has had to convince some skeptical visiting auditors, explaining that the UV lamp is not for sanitization, but for dechlorination.
Rodriguez and his team share lessons learned from the experimentation with UV:
- Select a UV equipment vendor with experience in dechlorination.
- Understand the complexity of the equipment and how to use it. For example, wavelengths at 185 nanometers are optimal for dechlorination, unlike the 254 nanometers that has been used for microbial control.
- UV can be used for low-flow applications (roughly 300 to 800 gallons per hour), but is inadequate to handle large flow rates.