You are on page 1of 7

PEER-R EV IEW ED

Elements to Consider
When Qualifying Control
Environmental Chambers Used
for CGMP Applications
David Vincent and Allan Marinelli

ABSTRACT INSTALLATION MEETING THE DESIGN


It is fundamental to ensure that all elements of thermo SPECIFICATION
mapping of control environmental chambers (e.g., It is recommended that the installation qualification
refrigerators, walk-in cold rooms, freezers, incuba- (IQ) be conducted to meet the design specification as
tors, walk-in incubators, etc.) are properly designed correlated with the user requirement specifications prior
and qualified with respect to user requirement speci- to any operational testing. The key elements to verify
fications compliant with current good manufacturing are as follows:
practice (CGMP) standards. Material of construction
As part of the installation, operational, and perfor- Major component installation
mance qualification (IQ, OQ, PQ) protocol of the con- Utility requirements (i.e., electrical)
trol of environmental chambers, the thermo-mapping Calibrations of critical instruments
elements for testing are indicated as follows: Operation and maintenance standard operating
Installation meeting the design specification procedures (SOPs)
Empty chamber thermo-mapping test Location of unit
Open-door study (for informational purposes Spare parts list
only) Drawings.
Power failure (for informational purposes only)
Full loaded chamber thermo-profile mapping. EMPTY CHAMBER THERMO-MAPPING TEST
After the installation qualification has been success-
This is a technical article on how to perform a spe- fully executed, the next step is to perform the functional
cific validation function. This article aims to give testing (i.e., operational qualification), comprising of a
guidance to the industry on utilizing one methodol- sequence of operation and alarm testing followed by the
ogy for thermo-mapping profiling. The points to empty chamber study. With respect to the emphasis of
consider for each of the aforementioned testing ele- this article, only the empty chamber thermo-mapping
ments shall be briefly discussed in this article. This study is discussed. The purpose of the empty chamber
article discusses the equipment necessary to qualify thermo-mapping study is to show that all geometrically
a control environmental chamber. placed thermocouples (see Figure 1) are capable of main-

For more Author


information,
go to
gxpandjvt.com/bios
[
ABOUT THE AUTHORS
David Vincent is CEO of Validation Technologies, Inc. He can be reached by e-mail at davidv@vali-
dation.org. Allan Marinelli is a validation project manager and senior lead validation engineer at
Validation Technologies, Inc. He may be reached by e-mail at allanm@validation.org.

86 Journal of Validation T echnology [Spring 2009] iv thome.com


D AV I D V I N C E N T A N D A L L A N M A R I N E L L I

Figure 1: Example of thermo-mapping placement locations.


taining a specified temperature range for the entire dura-
tion of the test. For example in this case, a cold room
is expected to maintain a range 2C to 8C during X
amount of time. Depending on the design of the refrig-
erator or freezer and the advancement in the refrigera-
tion systems, the defrost cycle normally has a negligible TC-12
impact (i.e., very little temperature fluctuation-a tiny blip TC-13
TC-15 TC-14
on the graph) on the overall study. This may not neces- TC-11
sarily be the case for older refrigeration systems where
adjustment of the programmable logic controller (PLC) or
equipment analogue functions may be required to reduce
the impact of the defrost cycle in order to ensure accept-
able temperature profiling. Units that have a defrosted
cycle that will have an impact on achieving uniform
temperature profile over specific period of time may need TC-9
to be replaced by a newer refrigerator. A product impact
TC-7
evaluation can be conducted prior to a final decision of TC-10
TC-8
accepting or not accepting the refrigerator or freezer (i.e., TC-6
several thermocouples can be geometrically placed in
product or solution for a 48-hour study, etc.).
The following is an example of how standard tem-
perature mapping should be performed.
TC-2 TC-5 TC-3
Requirements And Materials
Prior to performing any thermo-mapping studies, the TC-1
environmental chamber (e.g., refrigerator) and moni- TC-4
toring testing equipment (e.g., Kaye 2000) must be
calibrated. In addition, all installation and functional
testing of the cold room, such as meeting design specifica-
tions, sequence of operations, alarms, must meet all the as stipulated in Table II (Kaye calibration require-
requirements of the qualification protocols pre-defined ments). For operation of the Kaye 2000, refer to
acceptance criteria. the Kaye Validator 2000 operation manual and/or
SOP XXXXXX.
Procedure Distribute 16 calibrated thermocouples throughout
Before the empty chamber study can be performed, it is the interior of the cold room in accordance with the
important to verify that all aspects of the environmental drawing in Figure 1. Allow the temperature in the
chamber have been met by ensuring that all installa- chamber to stabilize at set point prior to commenc-
tion and operational tests (e.g., sequence of operation, ing the study.
alarms) have been satisfactorily completed. Once this Once the chamber is in a state of equilibrium, initiate
has been achieved the second phase of the protocol the Kaye data logging
execution can commence. Continue recording data for a minimum of XX hours
A procedure should be developed that clearly defines (user defined can be up to 72 hours)
sequential steps necessary for the successful execu- The cold room door is to remain closed during
tion of the empty chamber thermo-mapping test for data collection. A sign indicating not to open the
a cold room. door of the refrigerator is to be posted on the front
An example of a generic empty chamber thermo-map- door of the cold room.
ping test procedure is indicated as follows: At the completion of the study, stop data collection
Prepare and calibrate 16 thermocouples using the Following the completion of the temperature unifor-
setup parameters specified in Table I (Kaye setup mity study and the open door temperature profile and
parameters). The pre-calibration of all thermo- power failure studies, perform a post-study verification
couples must Pass the pre-defined specifications (i.e., post-calibration) of the thermocouples.
gxpandjv t.com Journal of Validation T echnology [Spring 2009] 87
PEER-R EV IEW ED

Table I: Setup parameters. cycle it is acceptable that the temperature (i.e., ther-
Type of study Temperature distribution mocouples measured in air) to go slightly out of
2C or 8C for a very short period of time. When
Number of thermocouples 16 total
the defrost cycle temperature fluctuation occurs it
Type of thermocouple T
is recommended to perform an additional study
Resolution 0.01C by encompassing the geometric placement of the
Validator hardware thermocouples in product or solution in support
Printer intervals N/A of this protocol and to subsequently ensure that the
PCMCIA intervals N/A 2C to 8C are met.
File intervals 10 min
OPEN-DOOR STUDY
This test is done to measure the time that it takes for the
temperature controlling unit to recover from temperature
Table II: Kaye calibration points requirements. fluctuation caused by opening the door.
Low -5C An open-door study can be useful in explaining how
the product can be affected or not affected to an auditor
High 15C
by the constant opening and closing of the door during
Check 4C a typical work schedule. To fully capture the range of
Sensor stability criteria the opening door data, four phases must be tested as
Maximum temperature change 0.5C follows:
Duration 3 minutes Priorto-open-door phase
Open door phase
Sensor deviation criteria
Close door phase
Maximum uncalibrated deviation from standard 1.0C Normal phase.
Maximum calibrated deviation from standard 0.5C
Log corrected results at 30-second intervals for 3 minutes This test is performed in an empty chamber distribu-
tion profiling X amount of hours or the minimum time
required to attain all four phases in condition that at
Data Analysis least one hour in each of the priorto-open-door phase
The following steps should be performed for data and normal phase is attained.
analysis:
Record any thermocouples that are outside the 2C Procedure
to 8C operating range and the duration they are An example of a generic open-door study procedure is
out during defrost cycle indicated as follows:
Graph the data and interpret the results per SOP Prepare and calibrate 16 thermocouples using the
XXXX setup parameters specified in Table III. The pre-cali-
Verify that no more than 10% of the thermocouples bration of all thermocouples must Pass the pre-
malfunctioned during the study or failed post-study defined specifications as stipulated in Table IV. For
verification. operation of the Kaye 2000, refer to the Kaye Valida-
tor 2000 operation manual and/or SOP XXXXXX.
Acceptance Criteria Distribute 16 calibrated thermocouples throughout
The following criteria should be obtained: the interior of the cold room in accordance with
Calibration verification readings (post-calibration Figure 1
of thermocouples) must be within 0.5C of the Allow the temperature in the chamber to stabilize
reference temperature standard. More than 85% of at set point
all thermocouples must be within 0.5C during Initiate data logging and ensure data are captured in
post-use calibration verification. each of the identified phases specified as follows:
The temperature in every location during empty Prior to open door phase
chamber distribution must be controlled within Open door phase
2C to 8C. Attach all supporting data such as Close door phase
excel graphs, etc. (see Figure 2). During every defrost Normal phase.
88 Journal of Validation T echnology [Spring 2009] iv thome.com
D AV I D V I N C E N T A N D A L L A N M A R I N E L L I

Figure 2: Empty chamber thermo-mapping test.


9.00

8.00
TC 01
TC 02
7.00
TC 03
TC 04
Temperature ( C)

6.00 TC 05
TC 06
5.00 TC 07
TC 08
TC 09
4.00 TC 10
TC 11
3.00 TC 12
TC 13
TC 14
2.00 TC 15
TC 16
1.00

0.00
13:26:22
14:59:00
16:32:00
18:05:00
19:38:00
21:11:00
22:44:00

11:08:00
12:41:00
14:14:00
15:47:00
17:20:00
18:53:00
20:26:00
21:59:00
23:32:00

10:23:00
11:56:00
13:29:00
15:02:00
0:17:00
1:50:00
3:23:00
4:56:00
6:29:00
8:02:00
9:35:00

1:05:00
2:38:00
4:11:00
5:44:00
7:17:00
8:50:00
Time

Data Analysis Table III: Setup parameters.


The following steps should be performed for data Type of study Temperature distribution
analysis: Number of thermocouples 16 total
P rint the Kaye data logger report and the ther-
Type of thermocouple T
mocouple calibration report. Attach all reports
to the protocol. Resolution 0.01C
Interpret the results per SOP XXXX Validator hardware
Verify that no more than 10% of the thermo- Printer intervals N/A
couples malfunctioned during the study or failed PCMCIA intervals N/A
post-study verification.
File intervals 5 seconds

Acceptance Criteria
The following criteria should be obtained:
Calibration verification readings (post-calibra- Table IV: Kaye calibration points requirements.
tion of thermocouples) must be within 0.5C Low -5C
of the reference temperature standard. More than
High 15C
85% of all thermocouples must be within 0.5C
Check 4C
during post-use calibration verification.
A ll four phases are captured as specified in the Sensor stability criteria
procedure section. Attach all supporting data Maximum temperature change 0.5C
such as Excel graphs. For example, refer to Fig- Duration 3 minutes
ure 3. Note: The door was opened for a longer
Sensor deviation criteria
duration on the left side of Figure 3 compared
Maximum uncalibrated deviation from standard 1.0C
to the right side (the second time the door was
opened). Maximum calibrated deviation from standard 0.5C
Log corrected results at 30-second intervals for 30 seconds

gxpandjv t.com Journal of Validation T echnology [Spring 2009] 89


PEER-R EV IEW ED

Figure 3: Open-door studies.


Close Door Phase
Open Door Phase Close Door
Phase
Open Door
Prior to Open Door Normal Phase
Phase Normal Phase
Phase
14.00

12.00

10.00
Temperature ( C)

8.00
TC 04
TC 12
TC 14

6.00 TC 16

4.00

2.00

0.00
10:32:01

13:03:30
11:22:30

14:44:30
12:13:00

13:54:00

21:28:30

0:00:00
0:50:30
1:41:00
2:31:30
3:22:00
4:12:30
5:03:00
5:53:30
6:44:00
7:34:30
8:25:00
9:15:30

10:56:30
11:47:00
15:35:00
16:25:30
17:16:00
18:06:30
18:57:00
19:47:30
20:38:00

22:19:00
23:09:30

10:06:00

13:28:00
Time 12:37:30

POWER FAILURE STUDY Procedure


This test is done in order to measure the time that it takes An example of a generic power failure study procedure is
for the temperature-controlling unit to recover from a indicated as follows:
power failure. Prepare and calibrate 16 thermocouples using the setup
A power failure study can be useful in explaining how parameters specified in Table V. The pre-calibration of
the product can remain within acceptable range, how all thermocouples must Pass the pre-defined speci-
long the product was out of range, and how much time fications as stipulated in Table VI. For operation of the
it takes to recover within specification during an actual Kaye 2000, refer to the Kaye Validator 2000 operation
power failure. To fully capture the range of the power manual and/or SOP XXXXXX.
failure data, the following four phases must be tested: Distribute 16 calibrated thermocouples throughout
Prior-to-power-failure phase the interior of the cold room (see Figure 1)
Power failure phase Allow the temperature in the chamber to stabilize
Power reactivation phase at set point
Normal phase. Initiate data logging and ensure data are captured in
each of the identified phases specified as follows:
This test is performed in an empty chamber distribu- Prior-to-power-failure phase
tion profiling at least 24 hours with the condition that Power failure phase
data for at least one hour in both the prior-to-power- Power reactivation phase
failure phase and the normal phase are attained. Normal phase.
90 Journal of Validation T echnology [Spring 2009] iv thome.com
D AV I D V I N C E N T A N D A L L A N M A R I N E L L I

Figure 4: Power failure study.


Normal Phase
Power Failure Phase
Power Reactivation
Prior to Power
Phase
Failure
25.00

20.00
Temperature ( C)

15.00
TC 04
TC 07
TC 14
TC 16
10.00

5.00

0.00
11:16:17

14:01:00

18:36:00

23:11:00

12:56:00
12:11:00

15:51:00

0:06:00
1:01:00
1:56:00
2:51:00
3:46:00
4:41:00
5:36:00
6:31:00
7:26:00
8:21:00
9:16:00

11:06:00
12:01:00
13:06:00

14:56:00

15:41:00
16:46:00
17:41:00

19:31:00
20:26:00
21:21:00
22:16:00

10:11:00

13:51:00
14:46:00

Time

Following the completion of the power failure tem- reference temperature standard. More than 85% of
perature profile study, perform a post-study verifica- all thermocouples must be within 0.5C during
tion of the thermocouples. post-use calibration verification.
All four phases are captured as specified in the pro-
Data Analysis cedure section. Attach all supporting data such as
The following steps should be performed for data excel graphs (see Figure 4).
analysis: Note: This test is to be used For information only.
Print the Kaye data logger report and the thermo-
couple calibration report. Attach all reports to the FULL-LOADED CHAMBER THERMO-
protocol. PROFILE MAPPING (PQ)
Interpret the results per SOP XXXX. The instruction for the empty chamber test is also
Verify that no more than 10% of the thermocouples applicable to the loaded chamber thermo-profile. The
malfunctioned during the study or failed post-study full-loaded chamber is defined as having at least 80%
verification. of surface area loaded with samples and containers in
order to determine the effects and impact on the prod-
Acceptance Criteria uct. This will ensure both thermo profiling extremes are
The following criteria should be obtained: tested, empty versus loaded conditions. Consequently,
Calibration verification readings (post-calibration a product impact assessment can be conducted and
of thermocouples) must be within 0.5C of the easily rationalized during an audit. Optionally, an
gxpandjv t.com Journal of Validation T echnology [Spring 2009] 91
PEER-R EV IEW ED

Table V: Setup parameters. open-door/power failure study can be conducted as


Type of study Temperature distribution part of the full load testing in order to simulate the
Number of thermocouples 16 total effects on the product and to subsequently evaluate
the overall product impact.
Type of thermocouple T
Resolution 0.01C CONCLUSION
Validator hardware The process presented in this article is one method
Printer intervals N/A that can be used to qualify a control environmental
PCMCIA intervals N/A chamber. It can be concluded that all elements of
thermo mapping of control units such as refrigerators,
File intervals 5 minutes
walk-in cold rooms, freezers, incubators, and walk-in
incubators may include the following tests:
Installation meeting the design specification
Table VI: Kaye calibration points requirements. Empty chamber thermo-mapping test
Low -5C Open-door study (for informational purposes
only)
High 15C
Power failure (for informational purposes only)
Check 4C Full-loaded chamber thermo-profile mapping.
Sensor stability criteria
Maximum temperature change 0.5C This article discusses just one methodology that can
Duration 3 minutes be used for performing temperature mapping. The
aforementioned tests as stipulated in this paper can
Sensor deviation criteria
provide a roadmap leading to successful audits that
Maximum uncalibrated deviation from standard 1.0C can be defendable to US Food and Drug Administra-
Maximum calibrated deviation from standard 0.5C tion inspections. JVT
Log corrected results at 30-second intervals for 30 seconds
ARTICLE ACRONYM LISTING
CGMP Current Good Manufacturing Practice
IQ Installation Qualification
OQ Operational Qualification
PLC Programmable Logic Controller
PQ Performance Qualification
SOPs Standard Operating Procedures

92 Journal of Validation T echnology [Spring 2009] iv thome.com