An electric heater in an AHU overheats when airflow stops. At the factory test, witness each link of its safety chain: the airflow interlock, the automatic-reset and manual-reset high-limit cut-outs, the fan run-on, the clearance to filters and sensors, and a contactor that can cut power if the solid-state relay fails conducting.
ICH M4Q(R1) divides the body of data in Module 3 into drug substance, drug product, appendices and regional information. A finished-product plant appears under manufacture in the drug product branch, where equipment type and working capacity should be identified, and, for biotech products, in the facilities and equipment appendix.
The Site Master File is read when an inspection is planned, and section 4 of the content list in the EU Explanatory Notes is engineering content: layouts with room classification and pressure differentials, HVAC principles, water schematics, other utilities and the major equipment list. Each appendix can carry its own effective date, so check them after every project.
EU GMP Annex 1 (2022) asks for an integrity test of final filters at requalification and puts no age on the filter. A terminal HEPA is replaced when a leak cannot be repaired within site limits, when pressure drop reaches the manufacturer's final value, or after damage or wetting. On glass-fibre media, dust loading alone does not lower efficiency.
Conductivity sees ions, so sugar, alcohols and most biofilm products pass it unnoticed. Total organic carbon catches them: the limit for bulk Purified Water and Water for Injection is 0.50 mg/L, or 500 ppb. A well-run loop sits far below that, so a slow rise or a step after maintenance is the signal to act on.
Fan curves are measured with a uniform inlet and a straight outlet duct. An elbow on the outlet or a wall near the inlet makes the fan deliver less pressure at the same speed, a loss the ordinary fitting coefficient does not include. AMCA Publication 201 gives system effect factors to add at selection, or keep the straight lengths.
EU GMP Chapter 8 says recall operations should be capable of being initiated promptly and at any time, and that the effectiveness of the arrangements should be periodically evaluated, in and out of office hours. A mock recall is to be considered, not prescribed. A small out-of-hours test of the contact numbers is a fair start.
An hour after a deviation is found, the room may be cleaned and the controller's trend close to being overwritten. Record it as found, contain it and export the data first. Then classify by risk, investigate the root cause with evidence, set corrective and preventive actions with owners, and have QA close and trend it.
30 Pa across a 2 m² door is about 60 N on the leaf, and doors that are hard to open get propped. EU GMP Annex 1 gives a minimum of 10 Pa between grades as a guidance value; typical design is 10–15 Pa. Larger steps add little separation, add up across a suite and need doors designed for them.
A signed printout with an electronic file behind it is a hybrid system. The MHRA guidance (2018) says it should be clearly documented what constitutes the whole data set, with all of it reviewed and retained. Check the procedure, the link from printout to file, the review, and the completeness and retention of both.
A pressure transmitter above a sealed ceiling with no test connection, a duct probe fixed out of reach, a pipe sensor with no second pocket: each turns calibration into a shutdown or a scaffold job. Test tees with isolating valves, removable probes and spare thermowells cost little on the installation details and save years of postponed calibrations.
Analytical method transfer is the documented process that qualifies a receiving laboratory to use a procedure that originated elsewhere. Routes include comparative testing, co-validation, revalidation and a justified waiver. The WHO guideline on technology transfer says procedures should be implemented before process validation samples are tested, with acceptance criteria agreed before the protocol is executed.