Water for Injection Testing Under Canada GMP: What Health Canada's Guidelines Actually Require

Most pharmaceutical manufacturers entering Canada are prepared to discuss their active ingredients, their finished product release testing, and their batch record systems. What consistently catches them off guard is the depth of scrutiny Health Canada inspectors bring to their water systems — specifically Water for Injection (WFI).

WFI isn’t just a utility. It’s a raw material. It makes up the majority by weight of almost every parenteral and sterile drug product, and its quality directly determines patient safety. The margin for error is narrow, and the testing requirements — across Health Canada’s own GMP guidelines, USP Chapter <1231>, and the Ph.Eur. monograph 0169 — are more demanding than many manufacturers realize until they’re sitting across from an inspector.

Here’s what you actually need to know before that conversation happens.

What Canada GMP Requires for WFI Production and Storage

Health Canada’s GMP framework for human drug products is set out in Division 2, Part C of the Food and Drug Regulations, with detailed expectations articulated in the GMP Guidelines (GUI-0001) and the sterile products annex (GUI-0104). For parenteral preparations and sterile drug products, these documents align closely with ICH Q7 and EU GMP Annex 1, but incorporate specific Health Canada expectations around water system qualification and ongoing control.

The starting point is production. Historically, WFI was required to be produced exclusively by distillation — a position Health Canada shared with the USP until the USP revision of 2017, which explicitly permitted purification technologies achieving equivalent quality, including reverse osmosis paired with additional downstream steps. The Ph.Eur. followed with its 2017 revision, permitting membrane-based methods meeting the same monograph limits. Health Canada’s current guidance is consistent with this evolution. That said, when a non-distillation process is in use, inspectors expect robust documented justification demonstrating that the chosen technology consistently meets all WFI specifications under worst-case operating conditions.

Storage is where many programs have their first vulnerability. WFI maintained at ≥ 80°C in a continuously recirculating hot loop resists microbial growth — biofilm simply doesn’t establish itself at those temperatures. Facilities that run cold storage systems (≤ 4°C) face more demanding sanitization frequency requirements and a tighter window between sanitization cycles before microbial risk accumulates. What creates inspection problems isn’t choosing one approach over the other; it’s running a hot loop at 74°C or 75°C as an energy-saving measure without a validated rationale showing that temperature is still adequate. If that decision is documented anywhere — in a change control, an SOP, an energy audit — an inspector will find it and ask about it.

The Four Testing Parameters That Define a Compliant WFI Program

A fully compliant WFI testing program under Canada GMP covers at minimum four analytical dimensions. Each has specific compendial limits, and each carries its own inspection focus.

Conductivity, specified in USP <645>, is typically measured in-line and at point-of-use. The Stage 1 limit for WFI is 1.3 µS/cm at 25°C — a threshold sensitive enough to flag ionic contamination before it becomes a safety issue. Most modern WFI systems use continuous conductivity monitoring, which is efficient and well-suited to trending. But the calibration records for those conductivity sensors are a consistent inspection target. A sensor that hasn’t been calibrated on schedule makes your continuous data practically undefendable. We’ve seen facilities with clean conductivity readings receive quality system observations purely because the instrument logbook was incomplete.

Total Organic Carbon (TOC), measured per USP <643>, monitors dissolved organic contamination. The limit is 500 ppb (0.5 mg/L) for both Purified Water and WFI. Like conductivity, TOC can be monitored continuously using online analyzers, or periodically via grab sampling. Either approach is acceptable under Canada GMP, provided the sampling frequency is risk-justified, documented in your water quality plan, and actually followed. The system validation should confirm the TOC analyzer’s response factor using sucrose and 1,4-benzoquinone system suitability standards — something inspectors may ask to see in the raw instrument records.

Bacterial Endotoxins represent the highest patient safety stakes of any WFI test. Pyrogens cause fever, systemic inflammatory responses, and in severe cases, septic shock in parenteral recipients. The limit for WFI under USP <85> is 0.25 EU/mL, determined by Limulus Amebocyte Lysate (LAL) or recombinant Factor C (rFC) assay. Health Canada inspectors reviewing endotoxin programs look at sampling frequency, positive product control results, analyst training and authorization records, reagent lot qualification, and out-of-specification investigation history — not just the passing results. A facility that passes every single endotoxin test but can’t produce documented analyst training for the person who ran them is in a weaker position than it appears.

Microbial Enumeration uses the action limit of 10 CFU/100 mL for WFI, per USP <1231>. For context, Purified Water’s action limit is 100 CFU/mL — WFI is ten times more stringent per unit volume. That 10 CFU/100 mL figure is an action limit, not a finished product specification: exceeding it mandates investigation and CAPA, not automatic batch rejection. But many facilities either over-react to any result above zero or, more dangerously, treat consistent counts of 8–9 CFU/100 mL as passing without noticing the trend. Both misapplications will generate an inspection observation.

One figure that surprises manufacturers new to the Canadian market: Health Canada expects not just individual test results but trend data analyzed at defined intervals — monthly at minimum for most WFI systems. A facility with 12 months of passing tests and no trend analysis documentation is still non-compliant.

Loop Qualification, Alert Limits, and What Inspectors Actually Check

Before a WFI system enters routine production use under Canada GMP, it must be formally qualified. The standard approach runs through Installation Qualification (IQ), Operational Qualification (OQ), and Performance Qualification (PQ), with PQ conducted over a minimum of three consecutive weeks of daily sampling. That three-week baseline must cover all point-of-use locations, all sanitization cycle phases, and ideally seasonal or operational extremes if the system is being commissioned in a controlled environment that might vary.

After PQ, facilities must establish site-specific alert and action limits from their own baseline data — not simply import the USP defaults as the entire program. Alert limits, set below the action limit, trigger increased sampling frequency and investigation before any limit is actually breached. Action limits trigger formal CAPA. A program that sets only an action limit, with no alert limit, has a structural gap. It’s the kind of gap a thorough inspector will spot in your water quality plan during the document review phase — before they’ve even walked onto the production floor.

Point-of-use sampling is the most common substantive gap we see in WFI programs at inspection-ready reviews. Sampling the storage tank and one or two convenient use points is not sufficient. Every point of use that contacts product, product-contact equipment, or is adjacent to open sterile product must be sampled on a schedule justified by your risk assessment. Dead legs, infrequently used sampling ports, long spur lines at the ends of distribution runs — these are the biofilm formation hotspots, and inspectors know the floor plan of a pharmaceutical water loop as well as your facilities team does. They will ask about those specific locations by name.

For facilities using third-party ISO 17025-accredited labs for WFI testing — a practical choice for smaller manufacturers, new market entrants, or CMOs scaling up in Canada — the quality agreement is the critical document. It must be specific about sampling procedures, sample hold times (microbial samples in particular degrade quickly and must reach the lab within defined windows), chain of custody documentation, and reporting timelines. A generic quality agreement that covers “testing services” without addressing these parameters isn’t adequate for a Health Canada inspection.

Reading a WFI Certificate of Analysis — and Knowing When It Falls Short

A WFI COA from a contracted laboratory should include: the specific USP or Ph.Eur. method used, the numerical result with units, the specification or limit, a pass/fail conclusion, instrument calibration references, and the authorizing analyst’s signature. For endotoxin results specifically, the COA should also reference the LAL or rFC reagent lot number, the positive product control result, and the standard curve or calibration verification data. That level of detail isn’t bureaucratic overkill — it’s the documentation demonstrating that the lab ran a valid, traceable test rather than simply issuing a number.

COAs that list “Endotoxins: Pass” without a numerical result and method reference are a red flag in the eyes of both your QA team and a Health Canada reviewer. If your contracted lab can’t provide this level of documentation, that’s worth addressing in your supplier qualification process.

One Practical Review Before Your Next Health Canada Inspection

If you have a WFI system and haven’t formally reviewed your monitoring program in the past 12 months, start there — not with the results, but with the program structure. Pull six months of trend data for all four parameters. Check whether alert limits exist and are being evaluated in your monthly trend reviews. Verify that point-of-use sampling covers every location in your current system map (systems change; maps sometimes don’t keep up). Confirm that endotoxin analyst training records are current and that out-of-specification investigation reports close within your defined timeframes.

Water systems don’t fail dramatically. They drift — and the drift shows up in trending data before it shows up in any single test result. Finding it there, documenting the investigation, and demonstrating corrective action is what turns a potential inspection observation into evidence of a functional quality culture.

---

Written by Nour Abochama, Quality & Regulatory Advisor, Androxa. Learn more about our team

Talk to our team about Health Canada compliance Contact us

Related from our network

• ISO 17025 Accredited Contract Laboratory Services in the US — Pharmaceutical, supplement, and raw material testing with documented chain-of-custody COAs suitable for regulatory submissions.
• EU GMP and Pharmaceutical Regulatory Compliance for European Markets — Regulatory consulting for manufacturers seeking EU market entry, including GMP alignment and EU Annex 1 sterile product compliance.