Pure water without the regulatory headaches

With water systems under the spotlight like never before, purity in the lab is becoming ever more important

Stringent regulations about the purity of water used in pharmaceutical manufacturing have been in place for many years, and water purification, storage and distribution must to be validated and qualified in order to comply with Good Manufacturing Practices (GMP). Recently, in an attempt to regulate the production process from beginning to end, these controls have been extended to cover the analytical instrumentation and water purification systems of pharmaceutical laboratories. The design, operation and maintenance of water systems are under the regulatory spotlight like never before.

Drug discovery and analytical operations frequently need highly purified water to ensure that analytical detection limits are optimised and results are reproducible. All types of impurities have an effect on results to some extent, and the elimination of ionic, organic, colloidal and bacterial impurities, as well as certain gases like carbon dioxide is essential. Ion chromatography is an example of a technique where impurities can have a dramatic effect on results. Sometimes the presence of contamination can be observed straight away, such as higher blanks, higher background noise and chemical interference. These are simple enough to correct with pure water, but others occur over a longer period of time and lead to slowly deteriorating results (Table 1). Calibration blanks should have very low levels of impurities, as should the water used to prepare standards, samples and reagents. It is equally important to avoid adding contaminants when the system is being cleaned. Figure 1 illustrates the extremely low levels of impurities possible in water from polishers such as the Purelab Ultra. The ultrapure water is contributing little to the peaks even at ppt concentrations of ions. Table 2 shows the impurity parameters for standard and ultra-trace ion chromatography.

  Table 1.

















Table 3.

Designing a water system
When designing a feedwater system from scratch, there are three main ways of supplying the water:
• a pre-purified central supply, with the water polished at point of use
• mains water supplied and purified at point-of-use
• mains water supplied to a local pre-purified loop and then polished at point of use.

With pre-purified central supplies, the mains water is purified by reverse osmosis or deionisation in a central plant room, the water distributed around the laboratories and then polished to the required specification at the point of use. This is the best approach if a loop is already in place to supply production requirements and, as the water is pre-purified, the loop has to be designed to meet good practice standards. This is also a low-cost option if an adequately sized pre-purified water loop is already specified for other uses, although the cost of the additional plumbing has to be taken into account. There are a number of disadvantages though including:
1. If the system fails the entire water supply could run out.
2. Every point-of-use system has to be physically isolated during routine sanitisation to prevent chemicals from damaging or contaminating the polishers or laboratory equipment.
3. The materials and fittings for the plumbing to the laboratories are expensive.
4. The pipework has to be designed very carefully to prevent static areas forming.
5. A lack of local ‘ownership’.
6. Production needs will always come first.

                     Figure 2.                                           Figure 3.

With a mains water supply system, the potable mains water is fed directly to each point of use in each laboratory. As standard plumbing is already in place for sinks and washers, minimal additional pipework is necessary and the water can be purified to the required standard at each point-of-use. This system has several advantages over a pre-purified central supply system, including:
1. The low cost of additional plumbing.
2. Little need for expensive inert pipework.
3. Little need for a design to minimise dead legs.
4. Drop feed and return lines to the point-of-use are not required.
5. As each user’s system is independent, reliance on others to maintain the supply is removed.

However, more point-of-use units will be needed so equipment costs will be higher, although this could be offset against the cost of running a central feed system and installing a dedicated loop. Also, the extra equipment may also take up valuable bench space, but wall mounted and under-bench cupboards could solve this.

The third alternative, the packaged central laboratory water system or local loop, is a hybrid between the two approaches and is becoming increasingly popular. Mains water is fed into a pre-purification system and then into a suite of laboratories or a floor in a facility, where the necessary final polishing is carried out at the point-of-use. It is a much less expensive option because it combines the equipment savings of central pre-treatment with a lower installation cost as the pre-purified loop is local. It is independent of the site’s facilities and simple to incorporate a buffer volume to reduce the impact of planned or emergency system outages. In addition, high volume flowrates can be fed directly to autoclaves, sterilisers and glass washers, and point-of-use bench space is minimised. However, the pre-purifying system itself requires space in a laboratory or associated area and only once the feed system has been decided upon can the location of the laboratory purifiers be considered. User preferences also vary from wanting a point-of-use polishing system on the bench next to the analytical equipment, to being tucked away underneath the bench or on the wall to maximise free bench space.

Design considerations
Attention to detail when designing and defining the system architecture is critical for each specific installation to ensure the maintenance of performance specifications of the entire system. One essential component is dynamic distribution, which keeps the water ‘polished’ and maintained at peak quality by recirculating it through the active technologies. However, too high a flowrate in a continuous recirculation system can cause the water temperature to rise and encourage microbial growth. A compromise is to operate the laboratory’s purifiers in an intermittent or reduced flowrate mode, which will maintain water quality while minimising heat generation and eliminate the need for expensive heat exchangers to cool the water. Using the appropriate materials is also extremely important. Recommended pipework should be multi-layer co-extruded materials and have excellent chemical resistance, low surface leaching properties, allow minimal gas diffusion and be smooth.

With regards to water purification, optimum results can be obtained by using a combination of complementary technologies, and Table 3 indicates the areas of peak effectiveness for each. All have their own strengths and there is some overlap between the different technologies. For example, reverse osmosis may allow some low molecular weight organic impurities to pass through the membrane, but they will be caught by adsorption on activated carbon. Ion exchange resins can also take out some colloidal materials, and even some gases while microfiltration can remove gases like carbon dioxide and oxygen if hydrophobic membranes are used.

Security and validation
Poor quality water can lead to misinterpreted data and wasted time and effort, yet it tends to be taken for granted that the water being used is always up to standard. Effective, clear, consistent monitoring of the supply is paramount. Resistivity and total organic content (TOC) monitors are a good start, as they give an excellent overview of system status, but alone they do not provide complete peace of mind.

Far better are enhanced system designs that minimise the chance of impurities accidentally entering the purified water supply. An example of this is the use of two purification packs in the series. The first pack is used to purify the water and the resistivity is monitored before the water passes through the second pack. As the first pack becomes loaded with impurity ions it begins to become less effective, and the monitor between the two packs detects this so the first pack can be replaced. During this time, any weakly ionised impurities that may elute from the primary pack will be trapped in the second, barely-used pack, so the final purity of the water is always protected. Figure 2 shows how water quality would drop if a single cartridge were used, compared to the maintenance of 18.2MOhm-cm water with two packs and inter-stage monitoring.

Good laboratory practice (GLP) requires access to certain operations to be restricted and consumables and media to be traceable. PIN codes or passkeys can ensure system configuration parameters are secure, as only trained, approved personnel can alter critical operational parameters or initiate procedures such as sanitisation, and different levels of access can be allocated to different operators. Data tags can be used to enable traceability of consumables, with everything from date of manufacture, media type and batch number to the identity of the operator being accessed via a database. The microprocessor management system can also detect if deionisation cartridges are incorrectly fitted or even whether they have been used before and where.

When it comes to validation, some factors like alarm set points and operating regimes can, to a large extent, be pre-validated. However, the final validation process must take place on the installed product. Full system validation is the best way to ensure that a system is performing to the standards set down in the design criteria, and a validation support manual is invaluable in making this process as smooth as possible.

Contamination avoidance
Despite the low levels of nutrients in high purity water, microorganisms can still multiply at an alarming rate and it is far easier to keep a clean system clean than to recover a contaminated one. The first step in minimising microbial growth is careful system design, but all systems will still need routine sanitisation and all surfaces should come into contact with the sanitising agent to prevent rapid regrowth. Using a composite vent filter will help eliminate microorganisms, as well as particles and carbon dioxide.

Figure 3 shows bacterial data collected over a five-month period for three different types of system. The blue graph shows the water quality obtained from a static reservoir that has been filled with high purified water - the bacteria count has risen to over 1000cfu/ml. The green graph shows water from a similar reservoir with recirculation - the results are much better. By far the best results, however, are obtained from the system shown in the purple graph. Water is taken from a dispense tap that is part of the recirculating loop just after the active technologies and just before the return water is fed back into the reservoir. Even without a point-of-use filter, counts of less than 0.1cfu/ml are consistently achieved.

Conclusion
The increasingly stringent requirements in pharmaceutical laboratories pose a challenge when creating an effective, reliable ultrapure water system. But even the most demanding specifications and validation requirements can be met with close attention to product design and installation, as well as operational procedures. A continuous supply of high purity water that consistently meets stringent quality specifications can, with care and attention, be routine in the laboratory.

by Alan Mortimer, Technical Director, ELGA LabWater