Great Expectations, Wrong Parameters

June 2, 2009
By failing to assess what is truly critical to safety, revised steam quality requirements fly in the face of common sense and a generation of sterilization science.

The publication of HTM-2010 by the U.K.’s Medicines Inspection Agency in 1994 brought about a wholesale re-examination of steam sterilization across the global pharmaceutical industry [1]. While most of the documents in this lengthy tome offer greater insight into an important process, some have created tension and may lead to misunderstandings in a number of areas.

Inspectors, initially from just the U.K. but later from across Europe, began to assert the validity of “steam quality” to practitioners in the U.S. and elsewhere. An entire generation of sterilization scientists found their concepts and assumptions regarding sterilization effectiveness challenged by the precepts embodied in HTM-2010. Inquiries into non-condensable gases, moisture fraction, superheated steam and equilibration time resulted in difficulties when the exacting requirements of HTM-2010 were applied.

Resistance to the tenets of HTM-2010 has led to difficulties; facilities have been disapproved, processes altered and product changes and approvals have been delayed in order to promote conformance to the monograph. Voices raised in opposition have been somewhat placated by statements that the expectations apply only to porous load sterilization [2, 3].
Where sterilization of sealed containers is the concern, the “steam quality” concepts are acknowledged as irrelevant (many of these processes use air-over pressure to aid container-closure integrity during cooling of the load post-process, thus negating several of the key steam quality concerns). This has led to an unfortunate separation of thought. At the core of both schools of thought, processes, parts and terminal sterilization are the very same: moist heat destruction of microorganisms using saturated steam; and while the differences are seemingly substantial, there is commonality that if examined closely leads to clearer perspectives on porous load sterilization.

This article will re-examine revised steam quality expectations with full consideration of how steam sterilization is accomplished in settings other than porous loads.

Non-condensable Gases

A maximum of 3.5 mL of non-condensable gases per 100 mL of condensed steam is considered acceptable [1]. [pullquote]At first glimpse, the notion that steam for sterilization should not contain gases that are not steam, and which could inhibit steam penetration to the surface of the object, seems eminently reasonable.

Consider, however, what occurs in a sealed sterile container of product with an air or nitrogen headspace. When an inoculum is applied to the underside of the stopper, and then subjected to the sterilization process (in the sealed container with heat supplied by steam on the outside of the container), the microorganisms are wholly inactivated provided there is steam contact with the surface where the microorganisms are placed (it is possible to inoculate where the stopper surface cannot be reached by the internally developed steam, in which case a moist heat resistant inoculum may survive). This occurs without any removal of air from the container and is in direct contrast with the expectation for minimal air in the moist heat process.

The situation inside the sealed container is similar to an autoclave without any air trap or condensate drain. The internal pressure in the headspace is higher than what would occur at equilibrium in a pure steam environment as a consequence of the non-condensable gases present in the container above the liquid surface. The absence of any gas removal from the container, accompanied with the destruction of the microorganisms internally, suggests that the presence of a substantial amount of non-condensable gases (as opposed to the minimal amounts tolerated by HTM-2010) has no adverse impact on the destruction of the microorganisms by moist heat. The notions of boundary layers and insulating effects of trace non-condensable gas on the sterilization process seem nonsensical in the face of complete kill without any air removal at all.

Moisture Fraction

The next “steam quality” element of concern is the potential presence of excessive moisture in the steam. The expectation here is that the steam will contain not more than 5% (10% for non-metallic loads) liquid in equilibrium with the gaseous steam [1]. This is somewhat counter-intuitive to begin with: the very first thing that happens with saturated steam when it contacts a cooler object is condensation of the steam vapor to liquid water. Placing a restrictive limit on this makes little sense given what will occur naturally in the autoclave when the steam contacts the load (as well as the walls of the sterilizer).

Returning to the terminal sterilization model, the absurdity of the requirement is even more evident. In terminal sterilization of liquids, the challenge microorganism is placed in the liquid (in addition to the container-closure area noted above that is evaluated separately).  Sterilization of solutions is readily accomplished where the moisture fraction is 100%, there being no steam present at all! The real effect of excessive moisture in saturated steam is to reduce the amount of heat available in the steam for heating the load elements to process temperature.

There is one more aspect of the moisture fraction that bears discussion. Moist heat sterilization derives its name from the presence of moisture in the heating process. Dry steam that does not condense during the sterilization process is no more effective than hot air in the sterilization of materials. Water must be present to attain sterilization in the 120°C to130°C range in the desired time period. Without liquid water, moist heat sterilization is impossible. A misnomer that seems all too common in this area is that of “dry saturated steam”. No such material exists in reality, as the most basic of requirements for saturation is the presence of both phases; liquid and gas.

(Concerns for excessive moisture in the steam also relate to the potential for moisture to blind the surface of porous materials preventing steam penetration into the interior. This is a valid concern for linen sterilization in hospitals where such loads are commonplace. In industrial steam sterilization, linens are never present, and the concern somewhat excessive. Applying this criterion to wrapped goods belies the fact that condensation formation during steam sterilization is inevitable, and mandating a low limit for it in the incoming steam is little more than a make work exercise.)

Superheated Steam

The HTM-2010 expectation is for a minimum amount of superheat in the steam as it approaches the sterilizer. Superheat can be created as saturated steam drops in pressure at constant temperature upon entry into the sterilizer. Its creation for a brief period of time is unavoidable. However, it is certainly the least important of the steam quality concerns, as the presence of a cold load and sterilizer wall means that any superheat developed in the steam upon entry will be present only momentarily. The transfer of heat from the steam to the load items and sterilizer wall will eliminate superheat rapidly. Concern for this inherent phenomenon in steam sterilization serves little purpose.

Equilibration Time

HTM-2010 limits to a maximum of 30 seconds (15 seconds in small sterilizers) after the start of cycle timing for the load to reach the set-point temperature. This is the last and perhaps most difficult of all the “steam quality” concerns to meet. Meeting the expectation ordinarily requires an increased number of pre-vacuums (5 or more appear to be required) and can extend the overall sterilization cycle substantially. The sense of the requirement is that a delay in equilibration of any penetration probe to the set-point temperature is an indication of residual air in the chamber.

The earlier discussion on non-condensable gases indicates that small amounts are likely innocuous, and depending on the configuration even large amounts might be ignored. The presence of residual air in the load can impede effective sterilization; however, the utility of a timing requirement as the definitive factor in determining its presence is unknown. I have personal experience with sterilization processes where the equilibration requirement was easily met, yet effective sterilization was not accomplished because of residual air elsewhere in the load. 

At the same time, it cannot be proven that thermometry will always detect the presence of residual air.  Doing so would require placing a thermocouple in every potential air retention location. It must also be noted that in many instances merely positioning a thermocouple in these difficult locations can aid air removal and steam penetration by altering the way in which the item is wrapped for sterilization. Moreover, there have been numerous reported instances where thermometric data is essentially useless in establishing sterilization effectiveness as the critical surfaces cannot be evaluated by physical means: i.e., needle cannula, membrane filters, and other largely inaccessible areas. These locales can only be assessed by biological indicators. So, while the concern is largely correct, the remedy appears inappropriate. That an arbitrary requirement unsupported by microbial data can be used to categorize a process as unacceptable or unacceptable seems wholly inappropriate [4].

Come-up times are merely an adjunct to sterilization validation and should not be considered definitive proof of effective (or ineffective) sterilization. While they can be associated with inadequate air removal, there are other causes of extended come-up times.  The sheer mass of the item being sterilized can delay attainment of the set point temperature without any residual air. The equilibration time constraint is intolerant of any explanation other than residual air. To offer it as a universal requirement overstates its utility in confirmation of cycle effectiveness.

An in-process draft of PDA’s Technical Report #1 included a similar, but slightly different error [5]. That draft implied that penetration thermocouples placed inside enclosed volumes of porous load items but not in contact with the item surface should be utilized to assess equilibration time. This appears to be erroneous as well. Residual air present inside these items can be compressed by the surrounding steam and thus appear to be at temperatures in the effective sterilization range, resulting in the conclusion that the process is acceptable. If air were present inside these items, its lower heat capacity (even though it is at the same temperature due to the compression) would mean that the surrounding load surfaces would take longer to reach sterilization temperature and might be inadequately sterilized.

Temperature is not a measurement of sterilization effectiveness unless it is known to be measured in a moist heat environment. Since it cannot be certain that a void space thermocouple is in contact with steam and not air, they should not be considered definitive in any way. Penetration temperature measurements must be of component surface temperature, because the presence of steam is proven by the change in temperature at the surface. Fortunately, this “requirement” was eliminated in the final version of the PDA document [6].

Physical vs. Microbiological Data

Sterilization with moist heat has been the subject of numerous texts, many publications and untold validations in our industry. Our understanding of steam sterilization is perhaps greater than any other sterilization method in current use, and it is difficult to reconcile the precepts of HTM-2010 (and its successor documents) with that understanding. One of the core elements of sterilization validation is the establishment of correspondence between the physical and microbial data derived from the process. Considering the interrelationship between the sterilizer performance as confirmed by physical measurements, and the desired microbiological outcome—microbial destruction—it must be recognized that the microbiological data must take preference. Physical projections of biological kill are derived from mathematical approximations fitted to microbial kill rates—lethal rate curves to be exact [7].

The term Fphysical is used by some firms to describe their sterilization process. This has led some practitioners to the erroneous impression that the physical measurements should carry the same weight as the microbiological data. The physical data from the process, and calculations made from that data, namely F0, should never be given the same consideration as microbial results, as they are derivative estimates of lethality, and can provide neither the clarity nor the certainty the biological indicator results can.

So while the relationship between the determined physical and microbiological lethality for a process should always be compared, precedence should always be given to the microbiological results. Any other perspective is incorrect; that the mathematical models have been weighted equally with the original microbiological death curves that they endeavor to emulate is most unfortunate.


If evaluating steam sterilization effectiveness using physical measurements of lethality is as difficult as described above, it is certainly appropriate to ask: What real value do the “steam quality” requirements have?

The origins of the “steam quality” expectations are hospital sterilization, where the adventitious infection rate of the patients was monitored in relation to the hospital’s central supply sterilization process [4]. In an environment where biological indicators were unknown, multipoint data loggers unavailable, and validation as practiced in our industry unheard of, the “steam quality” expectations have proven valuable. Their continued use in hospital and other situations where sterilization science and equipment are primitive is justified.  Within the global healthcare industry, where steam supplies are subjected to rigorous qualification and sterilizer validation has long established the effectiveness of the process using the most direct means—microbiological challenge—“steam quality” expectations have little usefulness. They may do no harm in the pharmaceutical industry, but it is unproven as to whether they do any good. Their imposition belittles the core of what sterilization validation must demonstrate and provides little benefit to the manufacturing firm or patient safety.

About the Author

James Agalloco, BSChE, MSChE, MBA, is president of Agalloco & Associates, a technical service firm to the pharmaceutical and biotechnology industry. He is a past president of the Parenteral Drug Association and served as an officer or director from 1982 to 1993.


1. Hospital Technical Memorandum 2010 (HTM 2010), Part 3, Section 9.0, UK National Health Service, 1994.
2. Agalloco, J. “Steam Sterilization & Steam Quality.” PDA Journal of Pharmaceutical Science and Technology, 54, No.1, 2000.
3. Compton, M.E. “Letter to the Editor, re: Steam Sterilization & Steam Quality.” PDA Journal of Pharmaceutical Science and Technology, 54, No. 6, 2000.
4. Shuttleworth, K. “The Derivation of United Kingdom Steam Quality Test Limits.” PDA Letter, December 1999.
5. PDA, Technical Report #1. “Validation of Moist Heat Sterilization Processes: Cycle Design, Development, Qualification and Ongoing Control.” Draft 18, 2006.
6. PDA, Technical Report #1. “Validation of Moist Heat Sterilization Processes: Cycle Design, Development, Qualification and Ongoing Control.” 2008.
7. Pflug, I. “Microbiology and Engineering of Sterilization Processes.” Environmental Sterilization Laboratory, Minneapolis, 1995.

About the Author

James Agalloco | Agalloco & Associates