Clean Steam Quality in GMP: Dryness, Non-Condensable Gases, Superheat and Condensate Testing

A steam-quality result answers a specific question

A compliant condensate result does not prove that steam arrives with suitable physical properties. A successful dryness test does not prove acceptable chemical quality. Neither result, alone or together, validates a sterilization cycle. These distinctions matter when a utility investigation crosses the boundary between the clean steam network and equipment that uses steam for a validated process.

The correct question is whether the supplied steam meets the approved requirements at the relevant location and operating condition. This article explains how to build that evidence through physical tests, condensate testing and investigation. The decision tables and examples are GuideGxP recommendations. Numerical test limits must come from the applicable requirement, adopted standard, validated process interface and site specification; they are not universal values for every pharmaceutical steam application.

Identify the regulatory and technical scope

EU GMP Annex 1, in its current 2022 revision, addresses steam used as a direct sterilizing agent. Paragraphs 6.16–6.17 connect appropriate generation with defined quality. For the stated direct-sterilization applications, condensate is related to the relevant current WFI monograph; microbial testing of steam condensate is not mandatory under that provision. Physical attributes include non-condensable gases, dryness and superheat, unless otherwise justified. Apply the complete clause within its intended scope.

EN 285:2015+A1:2021 is a current technical standard for large steam sterilizers, primarily for medical devices in healthcare. ISO 17665:2024 concerns development, validation and routine control of medical-device moist-heat sterilization. These are useful references where applicable, but their scope does not automatically make their test methods or limits universal for pharmaceutical SIP networks. Obtain the controlled standard when adopting a method; a public catalogue cannot replace its detailed procedure.

Separate the three evidence layers

The first layer concerns chemical and relevant endotoxin quality, assessed through a suitably collected condensate sample. The second concerns physical delivery: water carried with steam, non-condensable gases and superheat under defined conditions. The third concerns the receiving process: air removal, exposure, load or equipment configuration, heat distribution, penetration and the established sterilization acceptance strategy. Each layer has its own measurement and decision boundary.

A utility can support a process without proving the process's effectiveness. Conversely, an acceptable chamber record may not answer whether upstream contamination occurred. Keep utility release and cycle acceptance coordinated through an interface specification. The same event may require separate investigations: one establishes whether the supply is suitable; the other assesses the equipment, loads and batches that could have been affected.

Understand dryness without turning it into a label

Steam carrying liquid droplets behaves differently from dry saturated steam. Excess entrained water can wet loads or surfaces, disturb heat transfer and indicate inadequate separation or drainage. The relevant test evaluates a defined sample using a specified arrangement. A reported dryness value should be interpreted according to that method, not treated as a perfect direct measurement of every part of a changing distribution network.

Investigate plausible sources of wetness: generator carryover, rapid demand changes, inadequate warm-up, failed drainage, uninsulated sections, low points and downstream backpressure. Check whether liquid was already present in the sample line or introduced by an unsuitable arrangement. The pattern matters. Failure only during simultaneous start-up suggests a different mechanism from persistent failure at one branch after steady operation.

The corrective action should follow the cause. Increasing generator pressure without analysing the network may change other operating conditions without correcting condensate retention. Replacing a trap without checking its discharge route may leave the same problem. Verify the repair under the demand condition that originally exposed the weakness, while confirming that other users and operating modes remain acceptable.

Assess non-condensable gases as a system behaviour

Non-condensable gases can interfere with the intended steam-contact conditions because they do not condense like water vapour. Sources can include gases in feed water, inadequate removal during generation, air retained during start-up or ingress associated with faults and operating transitions. Their significance depends on the receiving application and where they collect. The network and the process must each provide appropriate removal paths.

A utility sample evaluates the gas behaviour represented by its location, demand and test method. It does not establish that a complex SIP circuit or porous load is free from air pockets. Investigations should therefore compare generator conditions, branch preparation, venting and the receiving equipment's air-removal sequence. Treat a local result as evidence about a stated boundary, not a blanket certificate for every connected user.

Record feed-water conditions and operating transitions during testing. If results change after a long shutdown, inspect the start-up sequence and rarely used branches. If they vary with generator load, review the relevant separation and gas-removal design. Method execution is also important: collection, temperature control and calculations must follow the adopted procedure. Uncontrolled modifications can make results incomparable with the acceptance criterion.

Interpret superheat using the correct pressure basis

Superheated steam is above the saturation temperature corresponding to its pressure. Before it can behave as condensing saturated steam at that pressure, it must lose that excess thermal energy. Pressure reduction can contribute to a superheated condition, depending on the initial state and system behaviour. The engineering question is whether the condition at the receiving interface meets its validated need.

The saturation relationship requires an appropriate absolute pressure basis and measurements representative of the same location and time. Comparing a remote pressure with a local temperature can produce a misleading conclusion. Sensor installation, calibration, response and heat transfer to the probe all affect interpretation. A temperature-pressure consistency check can support diagnosis, but does not replace the specific adopted tests for dryness or non-condensable gases.

When investigating superheat, review pressure-reduction stations, upstream conditions, demand and the sample arrangement. Do not simply cool a test stream until a favourable reading appears; that would change the object being assessed. Any conditioning required by the method must be defined and reproducible. Test acceptance applies to the specified arrangement and its justified relationship to the utility interface.

Make condensate testing representative and contamination-controlled

Condensing a representative steam sample allows relevant chemical and endotoxin testing. Select a suitable cooler, contact materials, connections and collection containers. The sampling system should not introduce the analyte being measured or remove it through uncontrolled adsorption. Define cleaning, conditioning, cooling-water separation and safe handling. A clean sample bottle cannot compensate for a contaminated cooler or an unrepresentative dead section of sample pipe.

Choose analytes from the applicable condensate requirement and site risk assessment. Conductivity and total organic carbon answer particular chemical questions; neither independently characterizes all possible contaminants. Endotoxin testing, where required, needs suitable collection and method controls. Steam condensate should not be casually described as bottled WFI for administration: its testing role, collection process and regulatory application are different.

Document temperature and handling requirements, holding conditions and laboratory instructions for the adopted methods. Investigate unexpected results using both the utility and sampling system. A cooler blank or controlled comparison may help identify a sampling contribution, but invalidating an initial result requires documented evidence. A later clean sample does not by itself demonstrate that the earlier contamination was an artefact.

Select locations and operating conditions together

Generator-outlet testing evaluates generation. Header and branch tests examine distribution. Testing near or at the receiving interface addresses what the process receives. Select representative and challenging locations using demand, distance, geometry, pressure reduction, drainage and use frequency. The furthest point is not always the worst case, and a convenient point beside the generator cannot represent every downstream configuration.

Specify whether testing occurs after defined preparation, during steady use, at peak overlap, at minimum demand or following a changeover. Record other users operating at the same time. If the goal is to challenge routine performance, the test should not create an artificially favourable network condition that never occurs in production. Safe engineered arrangements and trained personnel are necessary because steam sampling involves stored pressure and thermal energy.

Qualification and periodic testing have different purposes

Qualification establishes that the installed system can meet its requirements over the approved operating envelope. Periodic testing checks continued performance and can reveal developing problems. Select routine frequency from risk, duty, history, maintenance, variability and the requirements applicable to the specific use. Do not prescribe the same interval for every generator, outlet and physical test without considering what failure it is intended to detect.

Online pressure and temperature records are valuable for operating context and some failure modes. They do not automatically replace physical quality tests or laboratory analysis. Combine continuous or frequent operating measurements with justified periodic tests. Review trends by location and operating condition, because averaging across unlike users can hide a deterioration. Define when maintenance, extended shutdown or a distribution change requires additional testing before use.

Investigate failure without testing into compliance

Start by protecting the affected process and preserving evidence. Record the result, method execution, instruments, sampling arrangement, generator state, demand and time. Identify which users could share the same failure mechanism. Determine the period of potential impact from operating data and knowledge of the fault; the last passing sample is evidence, but not an automatic proof that all later use was acceptable.

Investigate test validity and system performance in parallel. Confirm calibration and execution, then inspect relevant generation, drainage, regulation and distribution conditions. A documented assignable sampling error may justify invalidation under the quality procedure. An unexplained failure should not be discarded because a repeat passes. Additional samples should answer hypotheses and define the scope of the problem, rather than serve as a search for a favourable result.

Worked example: dryness fails while chemistry passes

At a fictional site, condensate chemistry remains acceptable but a physical dryness test fails during concurrent equipment start-up. The team keeps these observations separate: chemical compliance does not neutralize a physical failure. It compares a stable-load test with the start-up scenario, reviews local drainage and confirms the sampling arrangement. A restricted condensate discharge is identified and corrected through controlled maintenance.

Return-to-service evidence includes the original challenging scenario, representative affected points and confirmation that the intervention did not introduce another quality issue. The receiving-process owner separately assesses affected cycles and loads. A new passing steam test supports current utility suitability; it does not retrospectively validate a potentially affected cycle. The final investigation links the cause, product-impact reasoning, corrective action and recurrence prevention.

Build an approval package that remains usable

The specification should identify each test, its rationale, method edition, location, operating condition, acceptance criterion and responsible reviewer. Reports should retain raw observations, calculations, calibration status and deviations. Preserve links to the current distribution drawing and the process interfaces covered. This allows another competent reviewer to understand what the result means without reconstructing the test from memory.

• Separate physical steam quality, condensate chemistry and process validation.
• Confirm the applicability of every regulatory or standard-based criterion.
• Use controlled test arrangements with representative sampling boundaries.
• Record demand, pressure basis, timing and preparation with results.
• Investigate failed results before deciding whether they are invalid or indicate a real defect.
• Assess affected users, cycles and batches through their own quality responsibilities.
• Verify repairs under the conditions that exposed the problem.
• Update monitoring, maintenance and change controls using the investigation findings.

Interfaces and source references

The Critical Utilities Systems hub connects steam design and quality strategy. Feed-water questions belong with Pharmaceutical Water & WFI Systems. Process steam use and SIP belong with Cleaning, CIP & SIP Systems; sterile interfaces also connect to Aseptic Fill-Finish & Barrier Systems.

• [GUIDANCE] EU GMP Annex 1, 2022, paragraphs 6.16–6.17: current official GMP guidance for the stated sterile-manufacturing steam applications.
• [TECHNICAL STANDARD] EN 285:2015+A1:2021: current technical standard; large sterilizer scope verified in the BSI catalogue.
• [TECHNICAL STANDARD] ISO 17665:2024: published technical standard for medical-device moist-heat process validation. Catalogue scope only; protected procedures and limits are not reproduced.