Design and Performance Qualification, of Pharmaceutical Water Purification System for GMP Compliance.

*Chemist :Soad Yacout Q.C. Manager of Microbiological Control affair European Egyptian pharmaceutical industries (EEPI)

Abstract
This article describes water purification system in pharmaceutical industries operating in compliance with cGMP and also to meet the USP requirements for purified water. Detail of the design consideration and selection ,system description & system validation are presented. Water test results from the system performance qualification are detailed. System deviations for TOC and microbial monitoring are noted and action taken to resolve them are presented. The European Egyptian pharmaceutical industries (EEPI) water purification system components were chosen to control running cost, maximize validation performance, and meet USP requirements for purified water.

Introduction
Purified water is one of the key components in most pharmaceutical manufacturing facilites ,and it is one of the most critical utility systems in a plant operating in compliance with Good Manufacturing Practices (GMPs).The United States Pharmacopeia (USP), which sets standards for different water qualities, states that Water for Injection (WFI) is intended for use in the preparation of parenteral solutions.1 However, for other pharmaceutical applications, the guidance is more general and intended to ensure that the user designs a water system fit for the intended purpose. In many applications, a suitable water system can be defined as providing water meeting the current USP monograph for purified water. This definition allows the system designer to consider alternatives to the typical stainless steel WFI system, while achieving the desired water quality in a more affordable and manageable manner. 1

Water Purification System Design Considerations and Selection

The plant began with an assessment of an existing reverse osmosis/ Continuous deionization (RO/CDI) water system for expansion and validation. The purified water is used for product preparations equipment (indirect contact with the product) and facility cleaning purposes . The water quality specification was set as described in USP 24 . The U.S. Pharmacopeia (USP) dictates the purity of water used for manufacture of pharmaceuticals . The compendial grades of water must meet a specification for conductivity and TOC . In order to meet these specifications it is often necessary to use a membrane process such as RO in combination

with an ion exchange process such as CDI .Using double-pass RO/CDI system .and a location for the system was identified. The design take in consideration the selection of the materials of construction for the water stainless steel piping systems . The major components of the selected system are described below. The purification process is illustrated schematically in Figure 1. The actual system and a point of use are shown in Figures 2 and 3, respectively. Fig. 1Water System Schematic storage vessels and distribution loop, therefore

The feed water from the city supply first passes through pretreatment including dosing with coagulant and pass through sand and carbon filter cartridges and five micron prefilter and passes through UV lamp before entering the Reverse Osmosis ( R.O.) station. The RO system includes an additional integrated pretreatment with antiscalent for calcium hardness and a 5 micron prefilter. This combination protect the reverse osmosis membrane from damage due to fouling from particulates, chlorine oxidation, and formation of mineral scale on the membrane surface. The antiscalent agent is a solid, long chain polyphosphate that weakly binds calcium ions and minimizes calcium carbonate precipitation (scale formation ). The use of the coagulant and antiscalent agent within the system pretreatment were particularly important in the system design as this eliminated the need to use softening for removing calcium hardness from the feed water supply; therefore, saving considerable space. 1-Reverse Osmosis Stage : Pretreated water flushes through a high pressure pump that boosts water pressure before entering the reverse osmosis membrane cartridge. The high pressured water through the reverse osmosis membrane with contaminants being rejected by the membrane between 95 and 99%. Water is rinsed continually along the upstream side to continually flush contaminants to drain. Approximately 70% of the feed water entering the system is processed through the membrane as product water at a rate of 7 m3 per hour, providing

the required ~ 80m3 per day. Reverse osmosis system performance is monitored including feed and produced water conductivity and the calculated % of ionic rejection. The water leaving the reverse osmosis system requires additional purification to meet the USP purified water quality requirements in term of conductivity. 2-Continuous Deionization Stage : The process of continuous deionization (CDI) uses ion-exchange membrane, ionexchange resin, and a DC electrical potential to ionizable materials from water. One of the main advantages of the process is that it does not require chemicals to regenerate the on exchange resins since the DC field regenerates the resins electrochemically. Another advantage is that the electric field helps minimize bacterial growth in resin bed. The CDI installed downstream of reverse osmosis system to remove ions that have not been (or can not be) removed by the RO.

Fig.2 Purified water system ,Major component

Fig.3 Point of use .

3- Storage and Distribution systems : The purpose of the storage tank is to buffer the fluctuating demands of up to 8 m3. The goal when designing and operating the storage and distribution system is to keep the water at these purity levels preventing any of three parameters listed below : a- Prevention of ionic contamination ( Increase in water conductivity ) The storage tank, piping pumps and other components of the system in contact with the purified water is made of mirror polished 316L stainless steel . b- Prevention of contaminants like foreign particles and microorganisms To prevent foreign particles and microorganisms contaminant from Storage and Distribution systems Sterilizing grade (0.22 micron) vent filter on the storage tank. The distribution piping & storage tank maintained with positive pressure. The system pumps used double mechanical seals ,using purified water as a seal flush fluid. Heat exchangers made of double tube sheets.

c- Prevention of microbial growth Microorganisms in purified water system usually form a biofilm on the internal surface of the storage tank and piping2. For this reason several ways to prevent microbial growth in purified water system. 1- Periodically sanitize the system by heating up the water ( 85-90oC) ,and keep it in the distribution system and piping circulating. 2- There are no "dead legs" (stagnant zones such as branch lines) in the piping long enough to allow standing water to cool below 85 oC . 3- Circulation pumps are designed so that all parts in contact with water remain hot. 4- Utilize an UV light installed in distribution loop to continuously sanitize the water stream. 5- The vent filter on the storage tank is necessary both for microbial control and prevent moisture condensation. 6- The top head and wall of the storage tank will continuously flushed with circulating hot water to remain clean. 7- Reduce temperature in piping and tank to (18-22 oC ) . The purified water in ambient storage system is maintained at room temperature. The growth of microorganisms and accumulation of endotoxines is prevented by periodic sanitization of the heat water 8- The ambient temperature is not optimal for growth of most microorganisms sanitizing a system (1-2 a week). 9- Piping sloped is carred out every of 1/8`` per foot (10mml /ml)to assure complete drainage of the system . 10- Liquid velocity of >5ft/fec. ( 1.5 m/sec) in circulation loop >3 ft/sec( 0.9 m/sec) in return section of the circulation loop during peak usage .

Material Selection
Selection of proper materials of constructions for purified water system components is very important issue.(2) The material used are stainless steel ( typically 316 L) for all storage and distribution system and pretreatment stage carbon filter cartridge wall and filtered water tank. The thermoplastic material such as polypropylene and PVDF used in pretreatment stages such as feed water buffer tank , sand filter cartridge wall. On the other hand ,thermoplastic in pretreatment system are resistance to corrosion ,no potential for metallic contamination of fluids and chemicals required for sanitation and extremely smooth internal surfaces without polishing.

Water System Qualification The typical three-stage qualification approach was followed, starting with the Installation Qualification (IQ), transitioning to an Operational Qualification (OQ), and finishing with the Performance Qualification (PQ). For the IQ and OQ . As USP 25 specifies, the IQ stage should consist of instrument calibrations, inspections to verify that the drawings accurately depict the as-built configuration of the water system, and where necessary, special tests to verify that the installation meets the design requirements.4 The vendor-provided IQ protocol package for each of the individual primary components (reverse osmosis, continuous deionized system, and storage reservoir systems) was used to provide verification of the hydraulic and electrical connections as well as the system drawings. An internally generated IQ protocol collected the details of all reference documentation, instrument and utilities verifications, spare parts verification, and drawing verification for the entire system as a whole. The vendor supplied OQ protocol was used to test the primary components to prove that they were operating according to the design specifications. A validation master plan for a water system typically includes an OQ stage consisting of tests and inspections to verify that the equipment, system alerts, and controls are operating reliably.4 This included testing of the equipment's controls and operation with both liquid path hydraulic and electronic tests. Specific testing of the system operating alerts was performed by challenges the system by exceeding the system limits using a calibrated instrument from the manufacturer. The internally generated OQ protocol covered the overall system OQ, which verified system operation including the distribution loop (point of use pressures, temperatures, and flow rates), water system generation, storage system operation, and alarms. The purpose of the Performance Qualification (PQ)was to demonstrate that the system produced and maintained re-circulating water that meets the compendial requirements of USP purified water over a suitable time period. The qualification period was chosen to strike a balance between time and testing burden , the need to demonstrate a robust system (reliability), as well as the knowledge that the system was intended and would continue to be monitored after completion of the qualification testing . After considering these requirements, a ten-week qualification period was approved. A test schedule was prepared with samples drawn from all available point of use (23 POU). All samples were tested for Total Organic Carbon (TOC), conductivity, and microbial analysis (bioburden). The city feed water to the system also was tested for bioburden and coliform bacteria. Acceptance criteria for the system were based on USP 25. All ports in the system required a

TOC result 500 ppb. Conductivity specifications with a range of 1.3 and 2.1 S/cm corresponding to a "normal" room temperature range of 20 to 25C. The USP bioburden requirement of <100 CFU/mL was used.4 The city feed water was monitored for bioburden .Samples for chemical analysis were collected in a 500 mL glass stopper bottle . Microbial samples were collected in 250 mL sterile glass bottles. Chemical analysis was performed using an Sievers TOC 820 Analyzer, which reported TOC , Mettler Toledo meter,model MC126 to measure the conductivity. All microbial testing was determined by the pour plate method3 where 1 mL of water was mixed with molten plate count agar (agar below 46C), plated, and incubated at 30-35C for 48-72 hr and 20-25C for 5-7 days. Samples were tested in duplicate, and the average of the two plates was reported. Total coliform was tested using the Total Coliform procedure3 where 100 mL of water was tested on MacConky broth , and incubated at 30-35C for 48 hours. Performance Qualification testing was performed during July to October of 2002. Conductivity at all points in the system behaved similarly with values ranging from 0.4 S/cm to 1.4 S/cm with most values well below 1.0 S/cm (fig.4 ). All results met the process qualification conductivity requirements. TOC testing during this period at all points in the system behaved similarly with values ranging from 50 250 ppb and one point 450ppb( fig. 5) .Investigation showed that. the deviation was attributed to storage condition.
Trend Analysis Plot for E.C. in CDI station
Linear Trend Model Yt = 0.702036 + 0.00343421*t 1.6 1.4 1.2
Variable A ctual Fits A ccuracy Measures MAPE 23.2663 MAD 0.1727 MSD 0.0445

Trend Analysis Plot for TOC

Linear Trend Model Yt = 90.1410 + 0.173741*t 500 400 TOC ( ppb) 300
Variable Actual Fits Acc uracy Measures MA PE 50.46 MA D 42.48 MS D 3885.78

E.C. ( us/cm)

1.0 0.8 0.6 0.4 0.2 1 14 28 42 56 70 Date 84 98 112 126

200 100

0 1 8 16 24 32 40 48 Date 56 64 72 80

Fig 4 E.C. reading during PQ

Fig.5 TOC. reading during PQ

System bioburden during the test period was consistently very low. More than 100 samples were taken over the course of the qualification with bioburden results values ranging from zero to 40 cfu/ml for all but four samples, which yielded 50 to90 cfu/ml( fig 6 , 7). These isolates were identified to the genus level FIO. As this isolate was not obtained in any other sample, and was never isolated from this or any other location during the remaining weeks of the qualification, it was treated as an isolated event and the system was accepted with continuing monitoring.

Time Series Plot of point of used ( count at 30-35 oC)

90 Total bacterial count at 30-35 oC 80 70 60 50 40 30 20 10 0 1 6 12 18 24 30 date 36 42 48 54
Variable 3-25C 6-25C 8-25C 11-25C 13-25C 14-25C 15-25C 16-25C 17-25C 18-25C 20-25C 21-25C 23-25C

Time Series Plot of point of used ( count at 20-25 oC)

100
Variable 3-35C 6-35C 8-35C 11-35C 13-35C 14-35C 15-35C 16-35C 17-35C 18-35C 20-35C 21-35C 23-35C

Total bacterial count at 20-25 oC

80

60

40

20

0 1 6 12 18 24 30 Date 36 42 48 54

Fig 6 Bioburden results for Point of use at 30-35oC 25 C

o

Fig 7 Bioburden results for Point of use at 20-

Monitoring of using high or low incubation temperature or shorter or longer incubation time were determined during the qualification period ( fig. 8 , 9)
Fig 8 Bioburden results at 30-35oC
CDI water at 30-35 oC
Total bacterial count 50 40 30 20 10 0 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 2 days 3 days Linear (3 days) Linear (2 days)

Date

CDI water at 22-25 oC

Total bacterial count

Fig 9 Bioburden results at 20-25oC

5 days 7 days Linear (7 days) Linear (5 days)

time study Over the course of the next year, no further isolates of any nature were obtained from this location, which underscored this event as isolated. Feed water monitoring results were as expected for potable city water. Bioburden typically
Date Final evaluation of the system demonstrated consistently low TOC, conductivity, and
1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77 81 85 89 93 97 101 105 109

50 45 40 35 30 25 20 15 10 5 0

bioburden in the system. Following approval of the validation report in February 2003, the system was considered acceptable for the manufacturing pharmaceutical products. Water System Performance Following the process validation, ongoing system monitoring and release of tested water is a standard requirement for GMP water systems. Water was sampled and released on a daily basis. The 10 points considered to be most critical, Test returned water ( the last point of use ). The other point used for development and for equipment cleaning were sampled on a rotating

basis every month. The city water also was tested on the weekly schedule to verify the absence of coliform bacteria.The incubation temperature and time 30-35 oC for 72 hr and 20-25 oC for 7 day was selected. Test results were compiled by the quality control and release certificate was attached to the lot files for any drug product batches manufactured. As part of batch release, acceptable release results were required from the sampling date prior to and after the used date of the water in the drug product batch. This testing system was performed from the close of the validation in February 2003 to 2007 with the resulting data set covering more years of system operation. Bioburden results were, with two exceptions, always acceptable - Figure 6,7. The system limit was set at < 100 CFU/mL based on USP criteria. Most daily results were negative for bioburden, with occasional single isolates and rare samples up to 10 CFU/mL. Conclusions The European Egyptian pharmaceutical water system is designed ,installed, qualified and monitored in accordance to USP purified water specification. . Use of standard components also expedited the validation process as the vendor-supplied protocols covered the component details and allowed simplification of the internally generated IQ/OQ protocols to focus primarily on the system details. The PQ was completed successfully and established all ports met the required water purity requirements. Post-validation system monitoring was performed for more than a year. Ongoing bioburden monitoring which typically exhibited zero bioburden in the daily test samples . The results all results support the robustness of the purified water generation and distribution system presented. References 1- Joseph T., George K., Jeffrey D., and Sean M.2006. Design, Qualification, and Performance of a Cost-Effective Water Purification System for a GMP Pilot Plant. . Pharmaceutical Engineering July/August 2006, Vol. 26 No. 4. 2- Leoind S. 2001. Pharmaceutical purified water storage and distribution systems an Engineering perspective. Pharmaceutical Engineering November /December 2001 pp.66-72. 3. Reference Part 9215.B, Standard Methods for the Examination of Water and Wastewater, 20th Edition, American Public Health Association, 1998. 4. USP 25; General Chapter 645 <Water Conductivity>.& General Chapter 1231 <Water for Pharmaceutical Purposes . *Chemist :Soad Yacout Q.C. Manager of Microbiological Control affair European Egyptian pharmaceutical industries (EEPI)