December 1, 2005 (Vol. 25, No. 21)

Choosing the Right System and Maintenance Plan

The “land of sky blue water” is a myth perpetuated by advertising companies. The reality for biopharmaceutical companies and other manufacturers is water purified from municipal water systems that may “vary like the Himalayas,” in the words of T.C. Soli, Ph.D., consultant, SoliPharma Solutions (

As companies begin designing, or tweaking, their pharmaceutical water systems, Dr. Soli recommends getting several years of actual municipal water data (not averages) to identify exactly how bad the water has been. Then, companies can design systems that can deal with the worst. A good working relationship with the municipal water office also helps.

In designing a high purity water system, companies have several options, including deionization, which removes chemicals but doesn’t sanitize the water; reverse osmosis, which removes chemicals and, if heated, also effectively sanitizes water; and distillation, which purifies and sanitizes water.

No one agrees which method is best. Reverse osmosis, for example, is allowed in Japan and the U.S., but not in the EU, as a means of making Water For Injection (WFI), which traditionally is made by distillation.

Redefining WFI

Earlier this year, the U.S. Pharmacopeia redefined WFI as “water purified by distillation or a purification process that is equivalent or superior to distillation in the removal of chemicals and microorganisms.” That opened the door to other technologies, but caused some to question whether reverse osmosis (listed in 2004) was still acceptable.

The distinctions among purification systems, however, are academic.

“Unless it’s put into a sealed, sterile container, water won’t remain sterile,” points out Joseph Manfredi, president and CEO, GMP Systems ( The problem is that, when installed, system components, are not sterile, so the high purity water system must be sanitized using heat, ozone, untraviolet (UV) light, or chemicals, alone or in combination.

No Sanitization Required

Arion Water ( challenges “conventional wisdom” in the biopharma industry. Bob Livingston, CTO, says Arion’s purified water system requires no sanitization at all, yet produces WFI-grade water from deionized systems with microbial activity levels below 10 cfu per 100 mL, endotoxin levels below 0.01 endotoxin units per milliliter, and minimal maintenance.

Livingston says Arion’s system costs one-tenth to one-third the cost of heat-based WFI systems. Competitors challenge that, saying savings are cancelled by water pre-treatment costs and the eventual cost of sanitizing a system not designed for sanitization.

After more than one year of operation using that system, Etex and Oncogene Science, a division of Bayer HealthCare Diagnostics, both have gone on record praising Arion’s system. After 21 months of operation, Pamela W. Adams, senior vp and COO at Etex, said in a letter that it had operated with no sanitizations and no changes of the deionization mixed beds.

Oncogene reports more than four years of operations with no contamination problems and only one sanitization, which occurred when point of use valves were updated.

The key to Arion’s proprietary design is high purity water. Although Livingston wouldn’t divulge the details, he did say it relies upon “processing equipment in the distribution loop to economically provide very high purity water.”

Arion’s system has fewer than 10 parts per billion of total organic carbon (TOC). In such a low nutrient environment, bacteria do not thrive.

The fallacies Arion is combating, Livingston says, include the perceived need for sanitary design, the use of stainless steel, single-stage centrifugal pumps, sloped pipes, continuous reverse osmosis, and microbial pretreatment.

“None of these are used in the semiconductor industry,” he says, “yet it’s well known that the semiconductor industry’s level of purity far surpasses the purity in the pharmaceutical industry.”

The semiconductor industry, he says, uses reverse osmosis, mixed bed deionization, and UV-TOC control, but never uses heat because of its expense.

“Control the microbiology and you’ll control the purification process,” Livingston says. Even in this system, Dr. Soli comments, “You won’t starve the bacteria, but you can limit their nutrient supply so they grow very, very slowly. They look like they’re not present in water samples because they grow in a tenacious biofilm that doesn’t readily shear off into the circulating water.”

Therefore, he says, “you still need periodic sanitization” to prevent slowly-growing biofilm from eventually becoming problematic.

Livingston counters that “Decades of semiconductor success demonstrates ultra-high purity water does not support microbial proliferation. The semiconductor industry sanitizes nothing and achieves less than 1 cfu/10 liters, dead or alive, routinely.”

Bacteria are a fact of life in ambient temperature water systems. Although our bodies can tolerate these bacteria when ingested, the endotoxins they produce may cause fever or other ills when injected intravenously.

“In ambient temperature water systems, nutrients accumulate on hard surfaces, so that’s where the bacteria go,” Dr. Soli explains. They proliferate slowly, forming a protective, ionically-charged biofilm that attracts more nutrients. “Once formed, it’s very difficult to completely remove,” he says.

Biofilm can’t be starved to death, but it can be killed by heat, ozone, oxidizing chemicals, and UV light. However, each of those solutions comes with caveats.

UV light kills only what it shines on. So, although the bacteria in a water stream may be killed, the biofilm and underlying bacteria remain in the pipes throughout the system. In WFI systems, which are heated, biofilm is found in the cooler hoses and taps that access the water system.

Oxidizing chemicals, like chlorine or hydrogen peroxide, are effective, Dr. Soli says, but are difficult to completely remove.

Ozone, with a half life of less than 20 minutes, is most effective as a continuous sanitant. Because it can’t be generated in high concentrations, compared to other sanitants, it doesn’t penetrate and kill established biofilm well. Therefore, its effectiveness as an occasional sanitant is less than when used continuously.

Manfredi agrees, but notes that with near continuous contact, ozone will minimize biofilm formation. “The water in systems I design typically is deozonated for only about 90 seconds, based only flow velocity,” he says. In those 90 or so seconds, the water is exposed to UV light to eradicate ozone, piped through the system, and returned to the ozonation tank.

“If you approach ozonation that way,” Manfredi says, “you won’t have a major problem with biofilm as long as you continue using the ozone in the same manner.”

To improve ozone operations, companies need to pay close attention to maintenance, Manfredi says. For corona discharge type systems, molecular sieves are one potential failure point because their resins eventually degrade and can allow air-borne contaminants into the ozone generation system.

Manfredi recommends investing in high quality dissolved ozone monitors, which may cost in the range of $10,000. Less expensive models can be inconsistent and less reliable, he says.

Manfredi also recommends an ambient ozone monitor and interlocked and isolated equipment. Despite few recorded injuries, accidental leaks are possible.

Whichever periodic sanitization method a company routinely uses, it must be used frequently. Sanitization at three-to-four-month intervals is generally ineffective because, although the sanitization kills the bacteria in the circulating water, it doesn’t necessarily kill and remove all the biofilmwhich is food for new or surviving bacteria. Therefore, plants often see spikes of microbial activity after sanitization as the new bacteria enters the system and feasts on the dead, but intact, biofilm.

“If you sanitize once per week you probably can stay ahead,” Dr. Soli says, “and then hit the system with another form of sanitization like peracetic acidannually or so,” to remove any residual organic material from accumulated, dead biofilm.

Routine sanitization can take many forms. “If you start right you’ll end right,” Dr. Soli says. That necessitates using the proper materials and controls, and incorporating the ability to use multiple sanitization methods. “Don’t cut corners, or you’ll have problems,” he emphasizes. Install a good system to begin with.

Previous articleAdvancing Viral Vector-Based Technology
Next articleArt Ludwig’s Sound Page