Professional Documents
Culture Documents
By Jeff Hargroves
Alza Corporation
V
alidation of process gas sys-
tems involve documenting ❝This articles validation protocols neces-
the expected system behav- sary?
ior, and verifying that the system
covers the
performs as expected. This article pertinent aspects There is no pat answer. Gen-
covers the pertinent aspects of IQ, erally, we must demonstrate that
OQ, and PQ related to process gas of IQ, OQ, and the process gas delivered at the
systems and many of the potential point-of-use meets the predeter-
problem areas. The validation of PQ related to mined user requirements. As long
nitrogen and compressed air sys- process gas as we demonstrate this, it does not
tems, including breathing air sys- matter what we call the demonstra-
tems, is used as an example which systems and many tion documents.
can be extended to the validation of
most other process gas systems. of the potential Design Considerations
Why Validate Process problem areas.❞ As with any other cGMP system
Gases? or equipment, we must design
nitrogen and compressed air systems so that they can
Process gas systems may include compressed air, be qualified. The design process begins and ends
nitrogen, oxygen, helium, or other inert gases. If the with documentation of the point-of-use require-
gas is used to operate product related system(s), or ments for the system.
directly affects the manufacture of drug products, we A typical compressed air system consists of the
must demonstrate the system can continuously oper- air compressor(s), driers (desiccant or refrigerated),
ate in a state of control. distribution piping, and filtration systems. A typical
Inspection agencies require that we demonstrate nitrogen system consists of either a liquid nitrogen
control over utilities that can potentially impact a prod- storage tank and vaporizer or nitrogen bottles, distri-
uct. However, the methods we use to document and bution piping and filtration systems.
demonstrate control of utilities are currently a source of Breathing air systems are becoming more com-
debate in the field of validation. For example, monplace as our industry increases the research
and production of potent and toxic drugs. The qual-
■ Will contractor start-up documents suffice? ification of breathing air systems is generally simi-
■ Is commissioning, using “Good Engineering lar to that of other process gases. However, there
Practices” adequate? are some specific, generally accepted requirements
Figure 1
Minimum Requirements of a Compressed Air or Nitrogen System
Characteristic Typical When to be concerned about
Acceptance Criteria this characteristic
Pressure, min. & max. 90-110 psig Usually to meet equipment needs; max. is important
if equipment is not capable of throttling pressure.
Flow, min. ≥ 10 scfm Usually to meet equipment needs; max. is important
if equipment is not capable of throttling flow.
Temperature, max. ≤ 90ºF Seldom important, except for breathing air, unless
there is a specific process requirement.
Purity Meets USP Monograph Use only if required.
Particulate Meets particulate class level as Same requirement as the room in which the gas is
defined by Federal Std. 209E introduced, tighter if product contact issues dictate.
Microbial Meets microbial limits, as Same requirement as the room in which the gas is
defined by your company, for introduced, tighter if product contact issues dictate.
given room classifications
Dewpoint ≤ -40ºF Process driven. (Don’t claim -40ºF if your process
doesn’t need it).
Hydrocarbon Non-detectable, (eg ≤ 25ppm Process driven, generally “non-detectable” for process
as measured with Draeger 10a/P applications. Specify the lower “non-detectable” limit.
hydrocarbon tubes)
Characteristic Typical Breathing Air When to be concerned about
Acceptance Criteria this characteristic
Pressure, min. & max. 20-25 psig At points-of-use (e.g., hookups to air hoods)
Dewpoint 0-45ºF NFPA 99 guideline
Carbon Monoxide < 10 ppm NFPA 99 guideline
Carbon Dioxide < 500 ppm NFPA 99 guideline
for breathing air systems. sider the particular process application before
Figure 1 lists the minimum requirements that assuming that conformance to USP specifications is
should be considered during design of a compressed required. Some of these methods are difficult to exe-
air or nitrogen system. cute in the field (e.g., Nitrogen-99%), and should not
be attempted unless they are required.
Process Gas Standards For microbial and particulate monitoring, the spe-
cific criteria developed in our respective facilities for
Currently, there are no universally recognized class 100,000, 10,000, 100 and so on, should be con-
standards for the validation of process gases. sulted. In general, the process gas must not nega-
However, a group within ASTM subcommittee tively impact the room into which it is exhausted.
48.06 is developing validation standards for the pro- For breathing air systems, the NFPA 99 specifi-
cess gases, and for the methods used to test the cations for breathing air is used as the basis for
gases. There are several places to go for direction on acceptance criteria. Additional guidance can be
the requirements of process gas systems. Generally, found in the journals of the Compressed Gas
process gas systems must meet the chemical, micro- Association.
bial, and purity requirements of the products they
will potentially contact, and the requirements of the Use of Final Filters
room into which they are exhausted, if applicable.
The USP has developed test monographs for pro- The use of the air should always be considered
cess gases, such as medical air, nitrogen-99%, nitro- when deciding what type of filter, if any, is required.
gen-75%, and oxygen. Care should be taken to con- Compressed air and nitrogen are often used to power
equipment and motors in the process areas. In clas- personnel can test the alarms much faster than some-
sified (such as 100,000, 10,000, and 100) areas, one who is unfamiliar with, and could possibly dam-
where process gases directly contact the product, age the equipment. Plus, the tests usually take only a
final filters are advisable. few hours with someone familiar with the equip-
Point-of-use filters should be in place, or at least ment, versus a few days for someone who does not
considered during PQ of the system. Many compa- work regularly with the equipment.
nies qualify the system without the point-of-use fil- For high purity gas systems, requirements for the
ters in place to ensure system integrity even if the fil- material in product contact closely mirror those of a
ter develops a leak during its use. high purity water system. All new high purity sys-
Filter model numbers should be documented, and tems should be pressure tested, cleaned, and flushed
controls should be established to ensure the same fil- according to preapproved procedures. For stainless
ters are used over the life of the system. The manufac- steel lines serving aseptic process areas (down-
turer may vary, but critical characteristics, such as fil- stream of final filters), the weld maps should be
ter materials, flow rate, and particulate filtration levels matched to the weld logs and to the material certifi-
must be maintained. This information can be docu- cations. If passivation has been specified, its proper
mented in the system or equipment SOPs, or in the execution and flushing should be verified.
maintenance management system. The filters found in
the system two years after the initial qualification must Operational Qualification Issues
have the same critical characteristics as the filters that
were originally designed, specified, and qualified. As with any system, all critical instruments should
Final filters serving class 100, or cleaner areas, be calibrated prior to the performance of operational
should be integrity tested. The frequency of testing tests. Critical instruments on a process gas system are
should be commensurate with their use. For this rea- those instruments used to measure the parameters
son, it is often advisable to locate the filters outside listed in Figure 1. However, an instrument need not
the process area. A common design approach is be permanently installed for each characteristic. For
using medical grade copper in the distribution sys- example, if diversity testing is done well, the perma-
tem, transitioning to stainless before entering into nent installation of flowmeter(s) should not be
process areas, with dielectric couplings at the transi- required. But a pressure switch used in maintaining
tion. This transition point is usually a good place to the minimum system pressure by turning on the lag
locate the system’s final filters. compressor, should be calibrated.
The instruments used to monitor the critical char-
Installation Qualification Issues acteristics of a breathing air system should also be
calibrated. These include on-line carbon monoxide
As with any cGMP system or equipment, every (CO), carbon dioxide (CO2), and dewpoint monitors.
inch of the system should be checked to verify con- CO and CO2 monitors can be easily calibrated by
formance with as-built drawings, construction mate- using standardized test gas canisters which trigger
rials, valves, cross-connection to other systems, and the alarms at the appropriate levels. The manufac-
unused portions of the system. In any large, unused turer of the respective monitor can usually provide
portion of the system, adequate protection prevent- the certified gas canisters. The gases are typically
ing fluid buildup during system shutdown, which provided in concentrations that correspond to the
could compromise the microbial purity of other alert and action alarm levels. They can be easily input
parts of the system, must be ensured. to the monitor, and then flushed from the system.
All alarms must be tested, including those on the
compressors and desiccant driers for the compressed Sequence of Operation Testing
air system, on the storage tanks and vaporizer for
nitrogen systems, as well as on the distribution sys- For complex systems, such as multiple air com-
tem itself. It is often useful to hire the service repre- pressors or multiple liquid nitrogen vaporizers, care
sentative to conduct these tests. These experienced should be taken to test, or at least bracket, all oper-
ational scenarios. This includes testing each com- all the points of use. Qualification should not be
pressor in a lead and lag position. destructive. If the gas lines are already connected to
Backup compressors or gas cylinders should be process equipment, the reading from the instruments
tested with the rest of the system. All potential oper- on the equipment may be obtained, and the line need
ating scenarios should be explored during the opera- not be broken.
tional qualification process. If an operating scenario Nothing should be taken for granted. If there is a
affects the quality of the gas produced, it should be pressure or flow specification at a point-of-use, that
incorporated into the hydrocarbon, dewpoint, and point-of-use should be tested. One cannot assume
microbial testing. that because the correct pressure appears at one drop
in a room, it applies to all drops in the room.
Testing of System Characteristics Pressure considerations during diversity testing
are discussed in the following section.
The following sections review specific measure-
ment techniques and issues for each system charac- Flow
teristic listed in Figure 1.
Each measurement technique should be carried out The flow rate of each point-of-use should be
with calibrated instrumentation. Although Draeger measured to verify that user requirements are met.
tubes and some accessories will not come with NIST However, for many drops, there may be no prede-
certificates, the flowmeter and timer used to capture the termined user requirement. Typically, a baseline
sample should be calibrated. The results obtained from flow measurement is taken for each drop, whether
accessories, such as Draeger tubes, should be reported or not it has a predetermined specification. By
with a corresponding error based on the manufacturer’s obtaining a flow rate measurement for each use
statement of accuracy. Test results are only as good as point, a comprehensive document is established,
the sum of the errors introduced during measurement. which can be used in the future to help make deci-
Testing must be performed so that additional sions about whether the system can support a new
error is not introduced into the system. For example, piece of equipment. For example, if on initial test,
during microbial sampling, an uninformed protocol only 10 scfm could be obtained, it is clear that the
executor might put the flowmeter between the point- line size or supply pressure will need to be increased
of-use and the agar plate, but this introduces poten- to support equipment that requires 25 scfm.
tial additional microbial contamination from the Performing the flow test at each drop also pro-
flowmeter. Whenever possible, valves, flowmeters, vides a visual check for large pockets of stagnate
flow restriction devices and pressure gauges should water in the pipeline. This can be important because
be placed downstream of the variable being tested. dewpoint measurements may not be performed at
By minimizing introduced errors, the certainty of the all locations. Condensate may form in the lines
final test result is supported. during the initial installation of a system, or after an
Test methods should be thoroughly documented, so old portion of an existing system has not been used
that qualification can be repeated years after the original for an extended period. If water is found in the line,
tests. For a small start-up company, this may mean sim- the system may not have been adequately cleaned
ply writing down the test procedure in the comments and flushed.
section of the protocol. For more mature companies that Flow rate diversity tests should also be performed
will be routinely performing these tests, the methods to identify how many (and which) points-of-use can
should be codified in a company guideline or SOP. be operated simultaneously. For a new system,
diversity values should be predefined in the design
Pressure documents. For existing systems, a few interviews
with the equipment users should provide sufficient
Although pressure is probably the simplest char- information for educated assumptions about simul-
acteristic to measure for a large system, a bucket full taneous use of equipment. Simultaneous recording
of fittings may be needed to provide connections to of flow rate and supply pressure at critical points-of-
use provides very useful information about the abil- ageable level. For example, in a room with three
ity of the system to perform as designed. compressed air drops, a sample from location one on
Flowmeters can be found in most process equip- day one, from location two on day two, and location
ment and instrumentation catalogues. A good con- three on day three should be adequate to ensure that
tract calibration company should be able to calibrate portion of the line is hydrocarbon free.
the flowmeter, and provide a standardization table
for each process gas. Dewpoint
Post Validation