Why TOC and conductivity trends flag a failing water system first
Many teams treat purified water compliance as settled by the heterotrophic plate count and, for Water for Injection (WFI), the bacterial endotoxin result. Those tests are periodic and slow. Total organic carbon (TOC) and conductivity can be read continuously, so a drifting trend shows in your data well before a plate count or endotoxin result crosses a limit. This is an operational rationale, not a guarantee.
A continuous TOC trend can cross an alert level while periodic plate counts still look unchanged.
USP <1231> describes alert and action levels as process-control tools, not substitutes for pharmacopoeial specifications. PIC/S PE 009-17, Annex 1, section 6.13 requires regular chemical and microbial monitoring, with alert levels based on qualification data and reassessed periodically. Section 6.15 adds that WFI systems should monitor TOC and conductivity continuously because these parameters may indicate overall system performance better than discrete sampling. Inspections often find that data were never trended, so an isolated in-limit result hid a slow deterioration.
This guide gives the USP and European Pharmacopeia (EP) limits, standard methods for TOC, conductivity, plate count, and endotoxin, and practices that keep a system inspection-ready.
Chemical limits for USP and EP water, and what TOC and conductivity reveal
USP and EP acceptance limits at a glance
Parameter | USP value | EP value | Source chapter |
|---|---|---|---|
TOC (bulk purified water and WFI) | Not more than 0.50 mg/L carbon | Not more than 0.5 mg/L carbon | USP <643>; EP 2.2.44 |
Conductivity, WFI and Highly Purified Water | Stage 1: 1.3 µS/cm at 25 °C | 1.1 µS/cm at 20 °C | USP <645>; EP 2.2.38 |
Heterotrophic plate count, purified water | <100 CFU/mL, action level | Not more than 100 CFU/mL | USP <1231>; EP 5.1.4 |
Heterotrophic plate count, WFI | 10 CFU/100 mL, action level | 10 CFU/100 mL | USP <1231>; EP 5.1.4 |
Bacterial endotoxin, WFI only | 0.25 EU/mL | 0.25 IU/mL | USP WFI monograph; EP 0169 |
Purified water has no endotoxin limit. The HPC figures are action levels, not release limits. EP conductivity for bulk Purified Water (monograph 0008) follows its own temperature table, so check the current edition.
TOC testing methods and what they measure
Total organic carbon (TOC) is an indirect measure of organic molecules in water, expressed as the carbon they contain. The TOC Testing Method described in USP <643> oxidizes organic compounds to carbon dioxide and detects it. The limit for USP Purified Water and WFI is 0.50 mg/L carbon.
Before routine use, the analyzer must pass a system suitability test using 500 ppb sucrose and 1,4-benzoquinone carbon solutions. Response efficiency must fall between 85% and 115%.
TOC does not identify the contaminant, so the trend carries the information. A steady rise can point to biofilm shedding, resin or membrane degradation, or a change in source water. See our pharmaceutical testing services for broader quality-control context.
Online and offline conductivity testing under USP <645>
Online sensors give continuous readings, but compliance requires the uncompensated value, because temperature-compensated readings are not permitted. Stage 1 compares that reading with a temperature-dependent limit; 1.3 µS/cm applies at 25 °C. Sensor and meter calibration keeps readings credible.
Water that fails Stage 1 moves to laboratory testing. Stage 2 reads a sample at 25 °C ± 1 °C against a stage limit. Stage 3 is a pH-dependent test after stirring and equilibration.
EP 2.2.38 has no three-stage structure. Its table gives 1.1 µS/cm at 20 °C and 1.3 µS/cm at 25 °C. Because USP and EP set Water Conductivity Limits at different reference temperatures, the two numbers should not be swapped.
TOC and conductivity as early indicators
Conductivity reacts to ions, so ionic breakthrough, exhausted deionisation and membrane leaks raise it first. TOC rises with organic ingress, biofilm, and degrading media. Read together, they separate an ionic failure from an organic one.
Trend both against alert levels from qualification data, and review the slope as well as the latest value. A reading that stays inside its limit while climbing for weeks is the earliest warning a system gives.
Microbial action levels and routine HPC sampling
USP <1231> and EP 5.1.4 give the heterotrophic plate count (HPC) for purified water as fewer than 100 CFU/mL. That figure is a recommended action level, not a hard monograph release limit. It triggers an investigation, and a site sets its own alert levels below it from qualification data and historical trends.
WFI is commonly held to 10 CFU/100 mL. The two figures are not interchangeable because the sample volume differs: 1 mL for purified water against 100 mL for WFI. Ten per 100 mL is a much tighter count per milliliter.
Standard HPC method
The method described for pharmaceutical water is membrane filtration of a defined volume, often 100 mL for WFI, with the membrane placed on R2A agar. R2A is a low-nutrient medium that helps recover stressed or slow-growing organisms. USP <1231> indicates incubation of at least 48 hours at 30 to 35 °C, or 96 hours at 20 to 25 °C, subject to a validated procedure. Report incubation conditions with the count so the result can be reproduced. The lab reports article covers how results are documented and traced.
Membrane filtration places the captured organisms onto low-nutrient R2A agar, where stressed cells can recover over extended incubation.
Sampling practice
A count is only as good as the sample behind it. FDA's high purity water guide and USP <1231> point to these practices:
- Flush the sampling valve for a consistent, validated time or volume. USP <1231> gives at least 30 seconds at high flow as an example.
- Collect with the same hoses and delivery devices used in manufacturing.
- Take at least 100 mL from WFI systems. FDA treats volumes under 100 mL as unacceptable, with 100 to 300 mL preferred.
- Collect aseptically, and test promptly or hold under controlled refrigeration.
- Where point-of-use filters are fitted, sample before and after the filter.
FDA describes daily sampling after each purification step and at every point of use during the first two to four weeks of validation. For WFI in routine use, it requires at least one point-of-use sample daily, with all points tested weekly.
Chemical trends come first
Plate counts take days to read and capture one moment. Rising TOC or conductivity may warrant investigation while the HPC still sits below its action level. A TOC slope that climbs for weeks, followed by a count creeping toward 100 CFU/mL, is the sequence trending is meant to catch.
LAL testing and the 0.25 EU/mL limit for WFI
Pyrogens are substances that cause fever when injected. Bacterial endotoxin, the lipopolysaccharide from the outer membrane of Gram-negative bacteria, is the pyrogen of main concern in water systems. It persists after the organisms that shed it have been killed or filtered out, so a clean plate count does not rule it out. In an injectable, contaminated water can cause fever and other reactions in the patient.
The USP WFI monograph sets the limit at less than 0.25 USP Endotoxin Unit per mL. European Pharmacopeia monograph 0169 gives the same figure, less than 0.25 IU/mL, so only the unit name differs. The purified water monographs carry no endotoxin limit.
Standard method
USP <85> and Ph. Eur. 2.6.14 describe the Limulus Amebocyte Lysate (LAL) test, which uses a lysate from horseshoe crab blood that reacts with endotoxin. USP <85> covers three method families: gel-clot, turbidimetric, and chromogenic. Ph. Eur. 2.6.14 divides these into methods A to F.
Gel-clot gives a pass/fail reading against the lysate's labeled sensitivity, λ. The geometric mean endpoint used to confirm sensitivity should fall within 0.5λ to 2λ. The photometric methods give a quantitative result, but sample turbidity, color, or precipitates can distort the reading. Inhibition and enhancement controls show whether the sample interferes, and any dilution used to overcome interference must not exceed the maximum valid dilution.
In a gel-clot test, an intact clot and free-flowing liquid are the two outcomes, which is why interference controls matter for a valid reading.
Where earlier trends fit
The endotoxin result confirms quality at release but gives no early warning. USP <1231> treats TOC, conductivity, and microbial counts as complementary to the endotoxin test, so investigation should start with chemical trends.
What inspectors look for in a purified water system
FDA's guide to inspecting high purity water systems points reviewers to preventive maintenance records, critical-instrument calibration, in-line TOC and conductivity data, trend reports, and excursion investigations.
Routine upkeep usually covers:
- Sanitization to a validated procedure and schedule, which USP <1231> also allows as a remedial step after an out-of-limit result
- Calibration of conductivity sensors and TOC analyzers
- Inspection of filters, resin beds and membranes
USP <645> requires the cell constant to be known within ±2%, meter calibration against NIST-traceable resistors accurate to ±0.1%, and temperature accuracy of ±2°C. USP <643> calls for a calibrated TOC instrument and periodic apparatus suitability checks. Calibration services should be traceable to recognized standards. Under USP <1225>, validation shows a procedure fits its purpose, while <1226> verification shows a compendial procedure works under your actual conditions.
An alert-level excursion calls for documented review and trend assessment, not proof the water is unsuitable. An action-level result under EU GMP Annex 1 needs prompt notification, a formal investigation, a product impact assessment, and corrective action. The flow runs from alert to investigation, root cause, CAPA, and an effectiveness check on later trend data. A defensible OOS investigation follows the same logic.
Articles on ISO 17025 and pharma quality control, and legally defensible evidence explain why traceable results hold up under review.
Review TOC and conductivity trends weekly, link every excursion to a CAPA, and keep calibration certificates beside the data they support.
Sources
- USP <1231> Water for Pharmaceutical Purposes — USP
- GUIDE TO GOOD MANUFACTURING PRACTICE FOR MEDICINAL PRODUCTS ANNEXES — PIC/S
- 〈643〉 Total Organic Carbon - USP-NF ABSTRACT — United States Pharmacopeial Convention
- QbD1200 USP System Suitability - Beckman Coulter — Beckman Coulter Life Sciences
- Water Purification Conductivity: Compliance and Calibration — METTLER TOLEDO
- 2.2.38. CONDUCTIVITY — European Pharmacopoeia
- High Purity Water System (7/93) | FDA — FDA
- EUROPEAN PHARMACOPOEIA 7.0 Water for injections — European Pharmacopoeia
- FDA document on bacterial endotoxin testing — U.S. Food and Drug Administration
- APR Endotoxin Supplement, Part X: Pyrogen Testing — Associates of Cape Cod
- Gel-Clot Technique — Microbe Investigations
- Bacterial Endotoxins/Pyrogens — U.S. Food and Drug Administration
- USP <85> Bacterial Endotoxins Test — Veolia Water Technologies
- USP Monographs: Water for Injection — United States Pharmacopeia
- Guide to Inspections of High Purity Water Systems — U.S. Food and Drug Administration