How many measurement points are needed for a temperature mapping?

In short

• Small volume (up to ± 2 m³ / 2000 L): 9 measurement points as standard — the eight corners plus the geometric centre.
• Larger volume (up to ± 20 m³): 15 measurement points as standard, with intermediate positions and extra heights.
• Warehouses and large rooms: a grid of loggers, every 5 to 10 metres in length and width, at multiple heights (WHO).
• Always add points at doors, cooling and heating units, air inlets and outlets and known risk zones (ISPE).
• The number is not a formula, but a justified choice based on a risk assessment (ISPE: "there is no formula for the number of sensors").

Why the number of measurement points matters

A temperature mapping shows how temperature is distributed across a room or piece of equipment. The right number of points sits between two extremes.

• Too few points: you may miss the warmest or coldest location (hot spot or cold spot). The study is then not representative and not audit-proof.
• Too many points: unnecessary cost and logistics, without added certainty. Two loggers measuring the same effect add nothing (ISPE calls this redundant).

The goal is therefore not a maximum number of loggers, but representative coverage: demonstrating that every storage location in use stays within the limits. GDP therefore asks for monitoring at the locations with the greatest fluctuations.

The rule of thumb: 9 or 15 measurement points

The well-known rule of thumb comes from the geometry of a space. A cube-shaped room has eight corners; with one point in the geometric centre you reach 9 measurement points. The corners capture the extremes, the centre gives the average.

For larger volumes you add intermediate positions and extra heights, often reaching 15 measurement points. The boundary between "small" and "large" is usually placed around 2 m³ (2000 litres) in practice.

An honest nuance: the ISPE guide shows minimum diagrams for about 2 m³ and about 20 m³. The "2000 litres" boundary is a common practical convention, not a hard regulatory figure. 9 or 15 is a fine starting point, but never the final answer without justification.

What do the guidelines say?

This is where a well-founded mapping sets itself apart. Three sources complement each other.

GDP — EU Good Distribution Practice

GDP requires an initial temperature mapping under representative conditions before commissioning. Fixed monitoring equipment is then placed at the locations with the greatest fluctuations. Repetition is risk-based or follows significant changes to the room or the cooling and heating installation. For small rooms at room temperature, a risk assessment (for example of heat sources) can suffice (GDP 2013/C 343/01, §3.2.1; for active substances 2015/C 95/01).

WHO — TRS 961 Annex 9 Supplement 8

The WHO describes the practical grid approach for storage areas:

• Loggers in a grid, every 5 to 10 metres in length and width (up to 20–30 metres in very large facilities).
• Height distribution per grid point: for a ceiling height up to 3.6 metres about three levels (for example floor, 1.2 m and 3.0 m); above 3.6 metres several levels across the full working height.
• High racking warehouses are sensitive to stratification: measure across the full height.
• Loggers with traceable calibration, an error of no more than ±0.5 °C and a measurement interval of about 1 to 15 minutes.

ISPE — Controlled Temperature Chambers

ISPE develops the measurement strategy further as a risk-based approach (Good Practice Guide, Table 6.1 and §6.4.2):

• Measurement points next to the room's control sensor, in the corners and the geometric centre, at doors, air inlets and outlets, fans and air distributors, near shelving and obstructions and at the maximum storage height.
• There must be a scientifically based rationale for both the number and the location of the sensors.
• There is no fixed formula. A generic rule such as "one sensor per 4 metres" without justification is explicitly discouraged.
• Rooms with multiple HVAC units may be divided into zones and assessed per zone.

See the central guidelines page for the full context per source.

The real answer: a risk-based approach

The right number of measurement points follows from the number of thermal influences in your space. The idea is simple: one measurement point per distinct influence. Multiple sensors for the same effect add nothing. The factors below drive the number up.

In every case: record the choice with a documented rationale approved by QA. That is what makes the study audit-proof.

Height and grid: distributing measurement points in 3D

A mapping is three-dimensional. You first lay out a grid in length and width (WHO: every 5 to 10 metres) and then distribute vertically per grid point. Why height is crucial: warm air rises, causing stratification — especially in high racking warehouses. Follow the WHO height rules: up to 3.6 metres usually three levels, above that more.

Common mistakes

• Measuring only the corners and forgetting the geometric centre.
• Ignoring height and stratification in high warehouses.
• Not placing a measurement point next to the control sensor.
• Copying a generic rule ("1 per 4 metres") without justification (ISPE discourages this).
• Skipping the door or another worst-case position.
• Not recording a written rationale, leaving the study not audit-proof.

From mapping points to fixed monitoring sensors

Mapping and monitoring are not the same. Mapping is temporary and uses many measurement points to find the hot and cold spots. Monitoring is permanent and uses fewer sensors, placed at the worst-case locations found during the mapping. In practice a mapping is often repeated every 3 to 5 years, sometimes annually for critical product storage, or sooner after changes and deviations — always with a documented justification. Read the difference between monitoring and mapping or how often to requalify.

Preparing for an audit? Also read what auditors want to see in a temperature mapping.