Step 1. Build the pressure cascade first

In a classified space, pressure is what holds the classification. EU GMP Annex 1 paragraph 4.14 states that adjacent rooms of different grades should have an air pressure difference of a minimum of 10 Pascals, described as a guidance value. Paragraph 4.16 turns that into an instrumentation requirement: indicators of air pressure differences should be fitted between cleanrooms and between isolators and their background, and differences identified as critical should be continuously monitored and recorded.

Match the transmitter range to the boundary

A Grade C to Grade D boundary moves by around 10 Pa, and a door swing pulls it further, so the accuracy has to be a small fraction of the limit you are watching. S+S publishes range and accuracy together, and the comparison is instructive. The PREMASGARD® 2121 spans ±1000 Pa at ±10 Pa accuracy, while the 2120-SD spans only +100 Pa at ±3 Pa. Both can watch a boundary that moves by tens of pascals, but on the 2121 the resolution is the same size as the limit you are policing; on the 2120-SD a 10 Pa deviation is unmistakable. Wide range protects the sensor from over-pressure, narrow range protects the reading. On a critical boundary, choose the narrow one and protect it with the mounting instead. The 2110 (±1100 Pa, ±3 Pa, 0-10V or 4..20mA selectable by DIP switch) and the 11D9-Modbus (±500 Pa, ±3 Pa, Modbus RTU, IP65) cover the same ground where a selectable output or a bus connection is needed.

Step 2. Lay out the temperature and humidity grid

The warehouse side has the most explicit published method: WHO TRS 961 Annex 9 Supplement 8 gives a spacing rule instead of a sensor count.

Horizontally, arrange loggers in a grid along the length and width of the area, with loggers every 5 to 10 metres; a footnote allows 20 or 30 metres in very large facilities. Vertically, repeat the grid at several levels. Where the ceiling is 3.6 metres or less, position loggers above one another at high, medium and low level, with the document giving floor level, 1.2 m and 3.0 m as the example; above 3.6 metres use vertical arrays, illustrated for a 6 metre store at 0.3 m, 1.8 m, 3.6 m and 5.4 m.

Risk overrides the grid. Positions should coincide with where products are actually stored, and high bay pallet racking is particularly susceptible to temperature stratification, so it must be mapped over its full working height. The grid is the starting geometry; the product map decides where it is populated, so the output is a shortlist of worst-case locations worth instrumenting permanently, not a sensor count.

Inside a cleanroom, Annex 1 handles placement deliberately. Paragraph 4.28 points to ISO 14644 Part 1 for the minimum number and positioning of classification sampling locations, and requires critical processing locations to be set by documented risk assessment; paragraph 9.7 makes the same point for grade A. So use the warehouse grid as a sanity check, not a substitute, and put permanent points where product is exposed, at operator intervention interfaces, and in return air paths that prove the cascade.

Step 3. Choose the parameters, not just the parameter count

Annex 1 paragraph 9.2 lists what a monitoring programme typically comprises: total particle, viable particle including personnel, temperature and relative humidity and other specific characteristics. Paragraph 9.6 adds that temperature and relative humidity should be controlled within ranges aligning with product, processing and personnel requirements. Those are qualities of the room, which is what the S+S families measure.

The HYGRASGARD® accuracy figure is usually misread. S+S publishes typical accuracy of ±2 % RH across 20 to 80 % RH at +25 °C, degrading to ±3.0 % outside that window. A cleanroom at 45 % RH sits inside the good window; a dry store at 15 % RH does not.

Step 4. Two alarm levels, and a written reason for any delay

This is where most installations are technically working and still not compliant. Annex 1 paragraph 9.9 requires appropriate alert levels and action limits, established from cleanroom qualification results and reviewed against trend data. Paragraph 9.10 adds the purpose: alert levels should be set so that adverse trends indicating deterioration of environmental control are detected and addressed.

Paragraph 5.2 adds the mechanism: alarm events should be acknowledged and evaluated for trends, with review frequency based on criticality and critical alarms reviewed immediately. A defensible scheme therefore has three properties. Two thresholds per critical parameter, because alert prompts review against trend while action prompts investigation. A review cadence that varies by criticality. And delays that are argued, since paragraph 4.16 requires alarm delays to be assessed and justified within the contamination control strategy and forbids overriding the warning signal without assessment.

Warehouse alarms use the same two-level structure for a different reason: the risk is product, not classification, and the question is how long the space can be out of band before the load is affected. WHO TRS 961 Annex 9 Supplement 8 sets a useful benchmark — where a mapping study includes door openings, temperature should stay within the defined limits except for a maximum of 30 minutes following the opening.

Step 5. Instrument the worst case permanently, map the rest once

The most common specification error here is treating a mapping exercise and a permanent monitoring system as one purchase. The WHO document describes mapping as a project with a protocol, an exercise and a report, whose outputs include identifying the best places to locate sensors for routine monitoring. So map the space first, using loggers that meet the published requirements: a sampling period allowing intervals from 1 to 15 minutes, a range wide enough for all anticipated extremes such as −30 °C to +60 °C, a three-point calibration certificate traceable to a national standard with error no more than ±0.5 °C, and software meeting the applicable data integrity requirements. Then leave permanent sensors only at the worst-case locations, and re-map when anything material changes. Paragraph 4.29 requires classification in both the at rest and in operation states, so a room that passes empty can still fail once it is working.

Which document owns which number

Audit findings here usually come from attributing a requirement to the wrong framework, so keep the mappings separate. ISO 14644-1 owns the particle classification and the minimum number and positioning of sampling locations; ISO 14644-2 owns the particle monitoring plan that provides evidence of performance. EU GMP Annex 1 owns the cleanroom pressure and alarm rules: paragraph 4.14 the 10 Pa guidance value between grades, 4.16 continuous monitoring and warning with justified delays, 9.9 and 9.10 the alert and action levels, and 5.2 the trend review cadence. WHO TRS 961 Annex 9 Supplement 8 owns the warehouse side: the 5 to 10 metre grid, the height levels, the 1 to 15 minute interval, calibration to ±0.5 °C, and the 30 minute door-open recovery allowance.

Where to buy S+S Regeltechnik monitoring sensors

Ruilianxin Technology supplies S+S Regeltechnik sensors sourced through the manufacturer and its authorised channels, with original packaging and traceable documentation. Send us the parameter, the range and the boundary you are monitoring, and we will match the family and order code. Single-piece orders for spares and pilot installations are welcome.

FAQ

Q1. Where should monitoring points be placed in a cleanroom?

Start from the pressure cascade, not the room plan. Fit a differential pressure transmitter across every boundary between two different grades, because EU GMP Annex 1 paragraph 4.14 sets a guidance value of at least 10 Pa between adjacent rooms of different grades and paragraph 4.16 requires critical differences to be continuously monitored and recorded. Then add temperature, humidity and particle points inside the room, chosen by documented risk assessment as paragraph 4.28 expects.

Q2. How many temperature sensors does a warehouse need?

WHO TRS 961 Annex 9 Supplement 8 gives the method: a horizontal grid at 5 to 10 metre intervals, stretchable to 20 or 30 metres only in very large facilities, repeated vertically. Where the ceiling is 3.6 metres or lower, place units at high, medium and low level, such as floor level, 1.2 m and 3.0 m; above that use vertical arrays. High bay racking must be mapped over its full working height.

Q3. How should alarm limits and delays be set?

Two levels, following EU GMP Annex 1 paragraph 9.9, which requires alert levels and action limits. The alert level is the earlier trend signal that triggers review; the action limit triggers investigation. Paragraph 5.2 expects alarm events to be evaluated for trends, with critical alarms reviewed immediately. On delays, paragraph 4.16 requires them to be assessed and justified within the contamination control strategy, so a justified 30 second door delay is acceptable and a silent one is not.

Q4. What documentation does the system have to produce?

Three things from three sources. WHO TRS 961 Annex 9 Supplement 8 covers warehouse mapping: the protocol, the exercise and a report, using loggers with no more than 0.5 deg C error and a 1 to 15 minute interval. EU GMP Annex 1 covers the cleanroom: continuous records for critical air pressure differences and trend evaluation of alerts. ISO 14644-2 covers the particle monitoring plan.

Related reading

• S+S Regeltechnik brand profile and range overview: S+S REGELTECHNIK
• Other German and European brands we supply: Brands
• Choosing between the sensor families model by model: S+S REGELTECHNIK Sensor Selection Guide
• S+S Regeltechnik sensors and other instrumentation in stock: Products
• More guides and technical resources: Knowledge Base