Thermal Sterilization Validation

According to the U.S. FDA, process validation is defined as: Establishing by objective evidence that a process consistently produces a result or product meeting its predetermined specifications. In thermal sterilization processes, "time at temperature" is critical to achieve required product sterility.

By Goran Bringert April 1, 2005

According to the U.S. FDA, process validation is defined as: Establishing by objective evidence that a process consistently produces a result or product meeting its predetermined specifications.

In thermal sterilization processes, “time at temperature” is critical to achieve required product sterility. Therefore temperature and time are the primary parameters used for validation of the sterilization process. The mathematical relation between “time-at-temperature” and the achieved level of sterility is exponential and makes it critical to have very accurate temperature measurements.

The lethality rate L is calculated as follows:

At a base temperature T b =121 °C and z =10 °C, the effect of 1 °C error in measured temperature T at 121 °C results in approximately 26% error in the lethality calculation.1Regarding the criticality of accurate temperature measurements, the FDA says: “For both validation and routine process control, the reliability of the data generated by sterilization cycle monitoring devices should be considered to be of the utmost importance. Devices that measure cycle parameters should be routinely calibrated. Written procedures should be established to ensure that these devices are maintained in a calibrated state. For example: Temperature monitoring devices for heat sterilization should be calibrated at suitable intervals, as well as before and after validation runs.”2

The required process temperature uniformity in the chamber, according to regulations3and industry standards, should be better than or equal to 1 °C or 0.5 °C depending on the application. The validation instrument, including temperature sensors, used for validation measurements should be at least three times as accurate as the process variable measured.4This means that the overall validation system accuracy should be better than or equal to 60.33 °C or 60.17 °C, respectively. All components involved in the measurement are referred to as the “measuring chain”—from the sensor via the measuring instrument to the display, as shown in the “accuracy required” graphic.

Random and systematic error sources in the measuring chain and the overall system accuracy.

Calibration compensates for systematic errors only. Random errors are not compensated for and can affect the results of a thermal validation study. The “Accuracy required” graphic shows the areas (in red) where random errors may occur in the measuring chain. Control and management of these variables require a good understanding of the underlying factors and are the responsibility of those who perform the validation study.

Thermocouple type T (copper/constantan) is the most commonly used sensor for temperature measurements in validation applications, due to its high accuracy and low cost. Accurate temperature measurements with thermocouples require proper design and installation of the thermocouple circuit.

Sensor design: A sensor designed for measuring the temperature in a large volume parenteral bag cannot be used for measuring the temperature in a 10 ml vial. The thermocouple reports only what it “feels.” This may or may not be the temperature of interest. Factors to consider when specifying the design of a temperature sensor are: size, shape, response time, heat conduction, and sensor position.

In a small volume for penetration studies, a twisted thermocouple could generate significant errors.

When using twisted, bare wires, the instrument measures the temperature at the first point of contact, which happens to be the furthest point from the tip. Using a twisted thermocouple to measure heat distribution in a steam sterilizer would not significantly affect accuracy. However, in a small vial, the error could be significant (see “Small volume study” graphic). Avoid this problem by reducing the junction to the smallest practical size (see “Welded thermocouple” graphic).

Inhomogeneous regions: An inhomogeneous region5located in an area with large temperature gradient, such as the sterilizer wall, will generate an error that cannot be eliminated by calibration. Random errors caused by inhomogeneous regions can exceed 4 °C.6The principle of errors generated by inhomogeneous regions in combination with a temperature gradient is illustrated in the inhomogenous region during calibration and validation graphics.

A thermocouple with welded tip provides secure contact at a single point, allowing it to be used in many different applications.

The inhomogeneous region does not generate an error during calibration as T 1 =T 2 .

During the validation study a temperature gradient exists across the wall of the sterilizer. An error is generated due to the difference in Seebeck coefficient between the thermocouple wire and the connector. Calibration cannot eliminate this error.

The gradient T 2 – T 1 = 75 °C in this example generates an error of 3.8 °C during the validation study.

Measuring system errors: The cold junction reference temperature will only be correct for all thermocouples under steady-state ambient temperature conditions with a system completely warmed up. A sudden change in ambient temperature will cause thermal scatter at the cold junction terminal. Keep the cold junction terminals covered and maintain the ambient temperature conditions during the validation study.


The error attributable to calibration should be no more than one-third and preferably one-tenth of the permissible error of the confirmed equipment when in use.7Regulations require calibration of the overall measuring system. Each thermocouple must be connected to the same channel in calibration as in the validation study. The validation system should be calibrated under the same ambient temperature and conditions as during operation.

No single measurement should be accepted as being correct unless it is verified by other results. Be patient. A frequent mistake is to take measurements and make adjustments before conditions have stabilized. Also, remember that the calibration of the temperature measurement standard must be traceable to accepted primary standards. Significant random calibration errors that can occur in a dry block temperature reference include: transfer calibration, stem conduction, uneven heat transfer, variable immersion depth, well inserts not used, stability, and time needed for stabilization.

Reference error: When calibrating a thermocouple T 1 , against a resistance temperature detector temperature standard T 2 , a key contribution to error is the difference in temperature between these devices when placed in the reference. This difference is called transfer calibration error and is potentially the largest contribution to calibration errors in dry block references.

Transfer calibration error contains two components:

Stem conduction error, which cools the thermocouple tip, and

Uniformity of the reference wells relative to the standard well.

Parameters influencing stem conduction can be inherent in the reference design while the user can reduce others, such as:

Sensor depth in the well;

Lateral thermal conductivity between medium and sensor wiring;

Wire gage and material of thermocouple wire; and

Temperature difference between medium and ambient.

A view of a dry block with inserts is designed to minimize air space around sensors.

Reducing stem conduction errors

To minimize stem conduction errors in a dry block temperature reference:

Use thin thermocouple wire as it contains less thermal cross section to draw heat from the tip.

Tips should be at bottom of well. Well depth should be sufficient to keep tip away from ambient.

Use inserts to bring sensor wires closer to the block for better lateral heat transfer. Thermal conductivity to the well sides improves with the decrease in air space around the sensor wires.

A dry block designed for maximum transfer calibration accuracy has small diameter wells with inserts that fit the size of the sensors under calibration (see “Dry block with inserts” graphic).

Calibration should be performed prior to a validation study and a calibration check should be performed at the conclusion of the validation.

Pre-study calibration should include a two-point calibration with calibration points bracketing the sterilization temperature for the process under validation; for example, 100 °C and 130 °C. Calibration checkpoint should be somewhere between the two calibration points to verify the calibration.

Post-study verification should feature a two-point comparison between the temperature standard and the temperature sensors to verify that the calibration of the measuring chain is intact. The calibration procedure should be documented and include data on the deviation between the temperature standard and each temperature sensor before and after calibration. To ensure traceability, the documentation must list the calibration parameters and equipment including serial numbers and last calibration dates.


PDA Technical Monograph 1

Draft Guidance Document “Sterile Drug Products Produced by Aseptic Processing—Current Good Manufacturing Practice, Section IX C 2. Equipment Controls and Instrument Calibration.” August 2003.

cGMP, EN 285, EN 554

EN 554

R.P. Reed, 1998, Thermoelectric Inhomogeneity—Obscure Obstacle to Quality, in National Conference of Standards Laboratories, Albuquerque, NM 1998.

Clarence A. Kemper, Design, Installation, and Calibration of Thermocouple Measuring Systems, in Validation of Aseptic Pharmaceutical Processes, Chapter 5. Marcel Dekker, Inc., 1987

ISO 10012—:1992 Metrological confirmation System for measuring equipment, section 4.3

Author Information

Göran Bringert, is director of pharmaceutical and biotech markets at GE Infrastructure Sensing;