Validation & Commissioning (IQ/OQ/PQ)

As-built drawings and P and IDs genuinely reconciled to what was installed, a weld map with a log tying every weld to its location, welder and schedule, current ASME Section IX welder qualifications, coupons matching the production welds, material test reports with heat numbers reconciled to installed locations, slope and drainability verification per line, a passivation record showing procedure, chemistry, concentration, temperature, contact time, coverage and verification result, signed pressure and examination records with their acceptance criteria, and instrument calibration certificates tied to installed tags.

It is a document mapping every item in the turnover package to the IQ checklist item it satisfies. It is the item most often missing from a handover and the one with the highest return, because a document set that exists but cannot be navigated costs the validation engineer the same weeks as a set that is incomplete. Requiring the index up front, before work begins, is the single cheapest thing a buyer can do to protect the qualification schedule.

Rule out the instrument first, because it is the cheapest test and it is routinely skipped. Compare the reading against a calibrated portable at the same point, check whether the sensor sits in representative flow rather than a stagnant pocket or an air-bubble path, verify immersion depth on temperature sensors, and confirm the analyser or probe is within its maintenance interval. Repeat TOC and microbiological samples with a fresh sample line before declaring an excursion, since sample line contamination and carryover are common.

Most often insufficient immersion depth or an air gap in the thermowell. An under-immersed sensor reads the pipe wall rather than the fluid, so it tracks ambient and process conditions in a plausible but wrong way. Thermowell fit, immersion depth and heat transfer paste are installation details that determine whether the reading means anything, and they are far cheaper to get right at installation than to diagnose later.

Because an instrument installed where it was convenient to reach rather than where the process is critical will read plausibly and consistently for years while telling you nothing useful. Placement is also nearly impossible to correct once the system is built and qualified, which is why sensor location belongs in design review rather than being left to the field. In hygienic service there is a second penalty: a probe set back from the bore in a threaded boss is both a poor measurement and a dead leg, which flush-mounted or extended-neck sanitary fittings solve together.

Judge an autoclave validation provider on how they handle heat penetration rather than on a list of clients. Steam sterilisation is validated by demonstrating that the coldest point in the coldest load reaches and holds the required lethality, expressed as F0, so the work turns on load pattern definition, thermocouple placement at genuine worst-case locations, and biological indicator challenge using Geobacillus stearothermophilus. Ask how many load configurations they will map and how those were chosen, since a validation covering only convenient loads leaves the real ones unqualified. Ask whether they perform empty-chamber heat distribution before penetration studies. And ask whether they can correct a failure, because a consultant who can document a cold spot but not fix the drain, trap or air removal causing it leaves you holding the problem.

Ask five specific questions. How will worst-case load configurations be identified and justified, because that decision determines whether the validation means anything. Where will thermocouples go, and does that include the drain, the densest part of the load and any wrapped item most resistant to steam penetration. How is air removal verified, since trapped air is the most common cause of sterilisation failure and a Bowie-Dick style test addresses it directly. What F0 is being demonstrated and how is it calculated. And who fixes it if the cycle fails: a provider tied to a contractor can correct a failing trap, a leaking door gasket or a drain problem in the same visit, while a documentation-only firm will hand you a deviation report and leave.

Validation is the documented demonstration that a process, system or piece of equipment consistently does what it is specified to do. In practice it covers commissioning and qualification of equipment and utilities through IQ, OQ and PQ, process validation for the manufacturing process itself, cleaning validation for shared equipment, computerised system validation under 21 CFR Part 11, and analytical method validation. It matters for two reasons. Legally, cGMP under 21 CFR 210 and 211 requires it, and absent or weak validation is among the most common FDA 483 observation categories. Practically, it is the mechanism by which a plant knows its equipment still performs as it did on the day it was accepted, which is what protects both product and patient.

There are two structurally different offerings and choosing the wrong one is expensive. Pure validation consultancies write and execute protocols but install and repair nothing, so when a test fails they document the deviation and you coordinate a separate contractor to fix it while the schedule slips. Contractors with in-house validation build the system and qualify it, so a failure found on Tuesday can be corrected and re-executed the same week. Consultancies are the right answer when you need independent review, a validation master plan, or capacity for protocol writing. Integrated contractors are the right answer when the qualification is attached to a build or a remediation. Paul Industries delivers installation and the full commissioning and qualification programme under one contract nationwide.

Screen on approach rather than price. Ask whether they will write the user requirement specification or only execute protocols someone else wrote, because a protocol tested against a vague specification proves very little. Ask how they apply risk-based verification under ASTM E2500, since scaling effort to what actually affects product quality is what separates a proportionate programme from an expensive one. Ask how deviations found during execution are handled and who has authority to close them. Ask what leverage of vendor factory and site acceptance testing they will use, since repeating supplier testing wastes money. And ask for a redacted turnover package from a comparable system, because that document shows whether their output is usable evidence or a binder.

Computerised system validation is a distinct discipline and general validation experience does not transfer. The relevant capability is 21 CFR Part 11 and Annex 11: audit trail completeness and review, access control and user privilege management, electronic signature implementation, data integrity across the ALCOA principles, backup and restore, disaster recovery, and change control for software. Ask how they scope validation by GAMP 5 software category, since a configurable off-the-shelf system and a bespoke application need very different effort. Ask how they test audit trails specifically, because that is where inspections concentrate. And ask how they handle supplier assessment, since leveraging a credible vendor quality system legitimately reduces the testing you must repeat yourself.

Six categories cover almost everything. Commissioning and qualification of equipment and utilities, running from design qualification through IQ, OQ and PQ. Process validation for the manufacturing process, now structured in three stages as process design, process qualification and continued process verification. Cleaning validation, demonstrating residue removal to a carryover limit derived from health-based exposure limits. Computerised system validation under 21 CFR Part 11 covering audit trails, access control and data integrity. Analytical method validation, demonstrating that test methods are accurate, precise, specific and robust. And sterilisation or aseptic process validation, including heat penetration studies, media fills and container closure integrity. A programme quoting only IQ, OQ and PQ has scoped one of the six.

The useful screening question is what each stage will actually test on your equipment, because generic protocols are the common failure. Installation qualification should verify the system was built and installed as specified, drawing on material certificates, weld records, calibration status and utility connections. Operational qualification should challenge each critical operating parameter across its range, not merely at setpoint. Performance qualification should demonstrate the system performs in its intended use, which for a water system means a multi-week sampling campaign and for a steriliser means loaded heat penetration studies. Ask to see sample protocols for a system like yours, and check whether acceptance criteria are specific and measurable or vague enough to pass anything.

Commissioning is the engineering activity of bringing a system into working order: checking installation against design, setting it to work, balancing, tuning and demonstrating it functions. It is good engineering practice and applies in any industry. Qualification is the regulated documentation exercise demonstrating, to a standard an inspector will accept, that the system meets predefined requirements and performs consistently. The two overlap heavily, and modern practice under ASTM E2500 deliberately integrates them so that commissioning testing is planned, witnessed and documented well enough to serve as qualification evidence rather than being repeated. Where they remain separate, plants pay twice to test the same thing. The distinction that matters commercially is that commissioning can be redone cheaply and qualification cannot.

Locality matters less than continuity for this work, because commissioning and qualification staff travel and mobilisation is a fixed cost spread across a programme measured in weeks or months. What genuinely matters is whether the same team stays from protocol writing through execution and deviation closure, since handovers between phases are where errors enter. Ask whether the people writing the protocols will execute them. Ask who has authority to close a deviation found during execution. Ask whether they can physically correct a failure or only report it, because that single capability difference determines whether a failed test costs days or weeks. Then ask for references at facilities with comparable regulatory exposure rather than comparable geography.

The sequence begins before the equipment arrives. A user requirement specification defines what the system must do, and everything downstream tests against it. Design qualification confirms the proposed design satisfies those requirements. Factory acceptance testing at the supplier and site acceptance testing on delivery catch problems early and, if properly witnessed and documented, can be leveraged as qualification evidence. Installation qualification verifies the system was installed as designed, covering utilities, materials, calibration and documentation. Operational qualification challenges every critical parameter across its operating range. Performance qualification demonstrates sustained performance in intended use. Then the system enters a maintained validated state, held there by change control, calibration, preventive maintenance and periodic review, which is where most compliance failures actually occur.

Consultancies are genuinely the right choice for some scopes and the wrong one for others, so match them to the need. They are strong at writing validation master plans, performing gap assessments, providing surge capacity for protocol writing and execution, and giving independent review before an inspection. They are weak where the finding requires physical correction, because a consultant who identifies a dead leg, a non-drainable branch or a failing trap can document it and nothing more. The expensive pattern is engaging a consultancy for a remediation programme, then discovering that every finding needs a contractor who must be procured, scheduled and brought up to speed. For remediation and build work, integrated delivery is faster and usually cheaper.

Autoclave and steriliser validation is a specialised corner of validation and the questions are specific. Ask how load patterns are defined and justified, how thermocouples are placed and how many, how air removal is verified, and how F0 is calculated and against what target. Ask whether biological indicators are used and how they are placed relative to the thermocouples. Ask whether empty-chamber heat distribution precedes loaded penetration studies. And ask what happens when a cold spot is found: the common causes are inadequate air removal, a failing steam trap, a partially blocked drain, a door gasket leak or an unrealistic load pattern, and a consultant cannot correct any of them. Paul Industries delivers installation, correction and the qualification programme under one contract.

The market splits into equipment suppliers who qualify their own machines, independent validation firms, and contractors who install and qualify. Supplier-led qualification is efficient for standalone equipment but stops at the machine boundary, which leaves utilities and tie-ins unqualified. Independent firms give impartiality but no corrective capability. Contractors give correction capability but should be asked how they maintain objectivity when qualifying their own installation, and the honest answer is a documented deviation process with quality oversight rather than a claim of independence. Choose based on where the risk sits in your project: standalone machine, whole-system integration, or remediation. Paul Industries delivers installation, tie-ins, passivation and the full qualification programme under one contract nationwide.

Aseptic process validation is the most demanding category and the capability is narrower than general validation. It covers media fills, also called process simulations, which substitute growth medium for product to demonstrate the aseptic process can be performed without contamination, typically at defined intervals and with defined acceptance criteria. It covers airflow visualisation smoke studies demonstrating unidirectional flow is not disrupted by equipment or operators. It covers container closure integrity testing, environmental monitoring programme design, gowning qualification and sterilisation validation for every component entering the aseptic core. The revised EU GMP Annex 1 adds a formal Contamination Control Strategy tying these together. Ask specifically which of those a candidate has executed rather than advised on, because the gap between the two is large.

They can, but the distinction between validation and compliance is worth keeping clear. Validation produces the evidence; compliance is the state of having that evidence current, complete and consistent with what the plant actually does. Validation work directly supports inspection readiness by producing qualification packages, cleaning validation, process validation and computerised system records. Where it commonly falls short is the maintained state: a system qualified three years ago that has since had an undocumented control change, an overdue calibration and unexplained excursions is not in a validated state regardless of what the original report says, and that gap is exactly what inspectors probe. Ask any provider how they support periodic review and change control, not just initial qualification.