Particle monitoring involves using particle counters to count airborne particles and classify them by size in order to ensure compliance with cleanroom air cleanliness standards. There are two types of particle monitoring: manual monitoring conducted periodically and continuous monitoring using online measurement systems. Standards like ISO 14644 provide guidelines for particle monitoring, including defining air cleanliness grades based on particle concentration limits and specifying sample volumes and locations for qualification measurements. Particle counters work by detecting and analyzing the light scattered when particles pass through a laser beam to measure particle size and concentration.

1. 3.K Particle Monitoring
Up22 Thomas von Kahlden
Here you will find answers to the following questions
● What is particle monitoring?
● Which standards and guidelines apply to particle monitoring?
● What is the difference between particle measurement carried out during qualification and measurement carried out
during cleanroom monitoring?
● How does a particle counter for airborne particles work?
● What is the difference between conventional particle counters and particle counters that are integrated in monitoring
systems?
● What must be taken into account when taking and transporting samples?
● How are particle counters calibrated?
● What has to be observed when operating manual particle counters?
● What parameters are included in a monitoring plan?
● What types of automatic monitoring systems are available?
● What are the main components of an automatic monitoring system?
● Which parameters influence the choice of sampling locations for an automatic monitoring system?
● What needs to be observed when the data is evaluated?
3.K.1 What is particle monitoring?
Cleanrooms require clean air. But how can air cleanliness be defined? We normally differentiate between particulate
and microbiological contamination of air. Contamination must remain below certain levels which are specified in the
GMP regulations (see Chapter 3.C Air Cleanliness Classes and Grades). This chapter deals with the monitoring of the
particulate cleanliness of air, also referred to as particle monitoring.
Compliance with the specified cleanliness grade must be tested at regular intervals or on a continual basis. Monitoring
has established itself as the preferred method for determining airborne particles. Particle counters are used that suck in
an air sample, count the particles in the sample and classify them by size. However, they do not differentiate between
viable microorganisms and other particles in the air. As opposed to air samplers, particle counters deliver the result
immediately after completion of the measuring interval.
There are two different particle monitoring processes:
● manual monitoring carried out at regular intervals, and
● continuous monitoring using online measurement systems.
GMP-compliant continuous monitoring systems have been used for more than 20 years. The systems were initially
developed for existing Windows platforms and the data were stored, e.g. as Excel files. The requirements have become
stricter on account of the increasing demand for data protection and manipulation protection of raw data, and because
monitoring data is considered to be production-related information. The software of computerised monitoring systems
must comply with the requirements for computer validation (e.g. GAMP® 5). A large number of automated monitoring
systems for recording and documenting production-related data are now available. Particle monitoring systems have a
large market share. To ensure that these systems can be used in pharmaceutical companies, the applicable
requirements had to be met. The standard functions include an audit trail, different password levels for access rights and
encryption of raw data (i.e. tamper protection).
For further information on computerised systems and detailed information on monitoring, please refer to the following
chapters:
● Chapter 3.J.6 Validation of a monitoring system in accordance with GAMP® 5
● Chapter 9 Computer System Validation
● Chapter 24.J Monitoring
3.K.2 Requirements for particle monitoring based on standards and
guidelines
There are a number of different standards and guidelines that contain requirements for particulate cleanliness of the air
and information on how particles levels can be determined. The pharmaceutical industry must comply with the
requirements in Annex 1 of the EU GMP Guidelines "Manufacture of Sterile Medicinal Products" (see Chapter
C.6.1 and Chapter 3.C Air Cleanliness Classes and Grades). ISO 14644, "Cleanrooms and associated controlled
environments" and the VDI Guideline 2083 are important from a technical point of view.
Annex 1 of the EU GMP Guidelines only defines air cleanliness grade limit values for particle sizes that are ≥0.5 µm and
≥5 µm. A distinction is made between systems in operation and at rest. The particle concentration limit values for grades

2. B and C for systems in operation are 100 times higher than for systems at rest. The particle concentration limit values
for both operational states are identical for grade A rooms only.
When the air cleanliness grades in EU GMP Guidelines Annex 1 and DIN ISO 14644 are compared, there are
noticeable differences.
For particles ≥0.5µm, the grade A classification of the EU GMP Guidelines corresponds to grade 5 in ISO 14644-1. For
particles ≥5.0µm, however, grade A corresponds to ISO grade 4.8. ISO 14644-1 outlines three different operational
states (as built, at rest and in operation), but does not specify different particle concentration limit values for the
individual grades.
In addition, the EU GMP Guidelines differentiate between initial qualification and requalification (also referred to as
classification), and routine monitoring (see Chapter 3.C Air Cleanliness Classes and Grades). EU GMP Guidelines
Annex 1 clearly states that particle measurement during the qualification of cleanroom areas should be carried out in
accordance with ISO-14644-1. It specifies that the tubing used for sampling should be as short as possible because the
risk of particle losses increases with the length of the tubing. This applies to larger particles in particular. Isokinetic
sampling probes are prescribed for unidirectional air flow systems (only affects grade A). A sample volume of 1 m3 per
sample location in Grade A zones as specified in Annex 1 is only relevant for measurements that are carried out during
qualification. It does not apply to continuous monitoring.
The EU GMP Guidelines also state that continuous particle measurement must be carried out in Grade A zones during
production. Continuous measurement is also recommended for the surrounding areas if they are classified as Grade B.
With regard to technical cleanroom requirements, the following parts of ISO 14644 "Cleanrooms and associated
controlled environments" are particularly important:
● Part 1: Classification of air cleanliness
● Part 2: Specifications for testing and monitoring to prove continued compliance with ISO 14644-1
● Part 3: Test methods
Part 1 defines the cleanliness grades in terms of concentration limit values. Part 2 contains information on the testing
and measurement intervals for continuous or recurring monitoring of compliance with cleanroom parameters. Part 3
describes how during an initial qualification or requalification, particle measurement has to be carried out to determine
the cleanliness class. The processes described in part 3 should be carried out during manual monitoring whereby the
measurement durations and sampling locations can deviate. The specifications in ISO 14644-3 do not apply to
automatic monitoring.
For further information on monitoring, and on particle monitoring in particular, please refer to the VDI Guideline 2083
Part 3.1 "Metrology in cleanroom air - Monitoring" (see also Chapter 3.C Air Cleanliness Classes and Grades, Figure
3.C-7).
3.K.3 Qualification and monitoring of cleanrooms
Measurements to determine the grade of cleanliness are carried out for different reasons:
● manual monitoring drive
● initial qualification of the cleanroom
● scheduled routine requalification
Different types of particle counters and measuring processes are used. Standalone manual particle counters are used
for manual measurement. Automatic monitoring systems use scaled-down particle counters that consists of a
measuring chamber, electronic evaluation system and, if applicable, a vacuum pump. These devices are operated by
the monitoring system controller; manual operation is not possible.
3.K.3.1 Measurement during qualification
If measurement is carried out to qualify/classify cleanrooms, the methods described in ISO 14644-1 must be used. This
document defines the number of sampling locations per room and the minimum sample volume for each individual
grade. If requalification measurement is carried out in Grade A areas, a sample volume of 1 m3 (1000 l) per sampling
location is required in accordance with EU GMP Guidelines Annex 1. ISO 14644-1 specifies a smaller sample volume of
684 litres (see Figure 3.K-1).
Figure 3.K-1 Graphic representation of sample volumes in accordance with ISO-14644-1. The sample volume flow
rates are only shown for particle sizes ≥0.5 and ≥5 µm.

3. Why does the sample volume depend on the cleanliness grade?
Small particles outnumber large particles in the airborne particle size distribution of naturally occurring particles in a
cleanroom. The naturally occurring particle size distribution in the atmosphere corresponds more or less to the limit
values for the cleanliness grades outlined in ISO 14644-1. Because a representative determination of particles ≥5 µm
must be carried out in accordance with the EU GMP Guidelines, a sufficiently large sample volume must be taken
depending on the cleanliness grade. The sample volume increases for zones with a higher cleanliness grade because
there is a lower particle concentration. The graphic in Figure 3.K-1 shows the correlation.
3.K.3.2 Measurement during routine monitoring
When measurement is carried out during routine monitoring, ISO 14644-1 specification deviations are permitted in
accordance with the monitoring plan. However, the specified sample volumes must be observed.
Are the minimum sample volumes specified in ISO 14644-1 justified?
ISO 14644-1 specifies a minimum sample volume of 2 litres. This value should be looked at in a critical light. If, for
example, a particle counter with a sample volume of 28.3 litres/min is used, the minimum sample volume would be
reached after a measuring period of only 4 seconds. In addition, a sample volume of 2 litres seems to be extremely
small compared to the total volume of the cleanroom. A 3-minute sampling period for each location might well be
required if a reasonably representative sample is to be taken.
3.K.4 Particle measurement terminology
To ensure that the sections that follow dealing with particle measurement technology are understood and interpreted
correctly, a number of terms must first be defined.
Particles
Particles are fragments of matter in a solid or liquid aggregate state that have defined physical properties (e.g. dust
particles, steam droplets, microorganisms, etc.).
Particle size
The particle size is the largest dimension of a particle. It can be determined using a microscopic method. If scattered
light is used, the equivalent diameter is also given as the particle size. The equivalent diameter is the diameter of a
comparison sphere with known properties that produces the same signal in the measuring device as the particle being
measured. The equivalent parameter can be used if calibration has been carried out, e.g. using latex spheres with a
known diameter, refractive index and density.
Particle concentration
If the airborne particles are counted using a particle counter, the measurement result is always expressed as a number
of particles in relation to the sample -volume, i.e the result represents a concentration value (particles per volume unit
[n/cft]). Even though we talk about a number of particles, it is in fact the particle concentration value.
Aerosol
An aerosol is made up of liquid or solid particles suspended in a carrier gas (e.g. air) for a certain amount of time.

4. Sample volume flow rate
The sample volume flow rate refers to the amount of air that flows through the measuring chamber in a defined period of
time. It contains the particles to be counted. The accuracy of the sample volume flow rate is decisive when determining
the particle concentration and should only deviate from the defined value by a small percentage. The sample volume
flow rate is a fixed value and depends on the device. Cleanroom particle measurement devices with a sample volume
flow rate of 2.83 l/min, 28.3 l/min, 50 l/min and 100 l/min are currently available.
Coincidence
If several particles overlap (coincidence) in the measurement system, two overlapping particles, for example, may be
measured as one particle. Depending on the design of the device, a specified air particle concentration should not be
exceeded to avoid incorrect measurements of this type. The coincidence limit is well above 500,000 particles per cubic
foot for modern particle counters, i.e. well above the concentrations that would be expected in a grade D cleanroom.
However, coincidence can occur when the particulate air filters are being tested because artificially high aerosol
concentrations are generated that must also be measured.
Zero count rate
The zero count rate of a particle counter expresses the background noise of the device. To determine the zero count
rate, an appropriate filter is attached to the sample inlet and the device is switched to measuring mode. When low
particle concentrations, in particular, have to be detected, e.g. for Grade A and B air quality standards, it is important to
know the zero count rate of the particle counter. State-of-the-art particle counters usually have a zero count rate well
below one particle per cubic foot in the smallest channel. If the determined zero count rate is too high, this might
indicate an "internal particle source". This is usually caused by contamination of the measuring cell after extended
operating periods with high particle concentrations.
Depending on the condition of the device and the environmental conditions, particle counters may detect particles even
though the particles sizes that can be detected by the device are not present in the sample volume. These incorrect
counting pulses can be caused by:
● electrical interference inside the device or from external sources
● devices that are sensitive to cosmic radiation and cause incorrect counts, e.g. photomultipliers
● detachment of particles from the aerosol sensor in the measuring chamber
● maladjustment between the aerosol sensor, measuring cell and eventual purge air flow
Counting efficiency
Counting efficiency refers to the ratio between the count impulses triggered by the particles suspended in the sample
volume of the measuring cell.
When individual particles in a stream of ultrapure air are measured, the lowest particle concentration measurement limit
is significant. The counting efficiency of smaller particles does not decrease suddenly, but gradually, because particles
that are close to the limit of detection only trigger counting pulses if they are optimally lit.
Modern particle counters usually detect particle sizes of 0.3 µm and above. It is normally expected that the smallest
measuring channels in particle counters have a counting efficiency of at least 50%.
Classification accuracy
The classification accuracy shows how accurately a defined particle size (e.g. latex particles) is assigned to the correct
particle size channel. Figure 3.K-2 shows the difference between "ideal" latex particles and real particles.
Figure 3.K-2 Left: ideal latex particles used for calibration. All the parameters of the latex particles, e.g. density, optical
properties and diameter, are known. Right: a real particle whose shape is very different from the shape of the
calibration particle (source: BS-Partikel GmbH – Wiesbaden and Infineon Regensburg)

5. Resolution
The resolution defines how accurately a particle counter can differentiate between different particle sizes.
3.K.5 How does a particle counter work?
Optical particle counters count and classify particles based on their scattered light intensity (standardised scattered light
diameter). The individual particles are guided through a laser beam. The scattered light impulse generated by the
particle is photoelectrically detected.
Figure 3.K-3 Schematic representation of an optical particle counter with 90° light-scattering detection (source: VDI
3489, Part 3)
How does a particle counter function?
The particles pass through the laser beam in the measuring cell. The beam is ideally several times thicker than the
particle. This ensures that the particle passes through a practically homogeneous light field. The theoretical analysis is
based on perfectly round particles, e.g. latex spheres that are generally used for the calibration of particle counters. The
particle counter always detects the intensity of the scattered light emanating from the particle being measured.
As soon as the particles start to pass through the laser beam in the measuring cell, scattered light is created that is
measured by the scattered light detection system, and the data is transferred to a downstream electronic system. An
analogue signal is created from the light data. If a particle passes through the measuring cell, short-term scattered light
intensities are created and, as a result, electrical pulses. These are counted by the optical particle counter and classified
using a pulse height detection system. This facilitates the detection of different scattered light intensities and thus
different particle sizes. After the pulse height analysis, the counts are digitally assigned to the individual size classes
and displayed at the end of the measuring interval.
Figure 3.K-4 contains a summary of this functional principle.
Figure 3.K-4 Functional principle of a particle counter

6. Functional principle of a particle counter
● The laser source sends a permanent laser beam through the measuring cell.
● To avoid scattered light from the laser beam hitting the wall of the measuring cell, a light trap is installed in this
position.
● As soon as a particle passes through the laser beam, scattered light is created.
● The collecting lens bundles the scattered light and directs it to the photodetector.
● The photodetector converts the light signal into an electric signal.
● The particle size is then determined based on the height of the signal.
The lower particle size limit of detection depends on the type of device. The upper limit is determined by sample-taking
and electronic overload rather than by the method used for measuring.
The count rate is used to determine the particle concentration, and the amplitude of the pulse is used to determine the
particle size. To determine the particle concentration, the volume flow rate through the measuring field must also be
known.
Apart from the particle diameter, there are other factors that have an impact on the intensity of the scattered light
created by the particle:
● intensity of the laser light
● wavelength of the laser beam
● diameter of the particle
● density of the particle
● surface properties
● various other photo-optical parameters (see Figure 3.K-5)
Figure 3.K-5 Schematic representation of a "real" particle when hit by a laser beam and the relevant photo-optical
parameters that have an impact on the creation of scattered light (source: MT-Messtechnik, Adelzhausen)
Scattered light measurement using real particles is a double indirect measurement. For this reason,
larger measurement errors must be expected during the actual measurement depending on the material and the
properties of the particles.
Double indirect measurement means:
● The calibration is carried out using latex particles that are absolutely round and whose physical properties are known.
● The particle itself is not measured, but the scattered light created by the particle (indirect measurement).
● Real particles do not have an ideal shape. The measured scattered light pulse does say anything about the shape
and size of the particle. It only indicates that the particle corresponds to the scattered light pulse created by a size X
latex particle (double indirect measurement).

7. To ensure compliance with the limit values for the air cleanliness grades, the particle size concentrations measured
during monitoring should be clearly below the limit value.
A major advantage of particle measurement is the immediate availability of the result after each measuring interval. On
the other hand, the fact that this method does not differentiate between viable and non-viable particles is a
disadvantage. For this reason, the operator has to carry out additional air sampling (see Chapter 12.J Microbiological
monitoring).
3.K.6 Conventional particle counters and counters used in monitoring
systems
Conventional particle counters that are used for manual measurement when determining the cleanliness grade are
stand-alone devices. They include
● an internal vacuum pump that creates the volume flow
● a comprehensive user interface for configuring all relevant parameters required to carry out measurements
● an evaluation unit that can be used to send the data to a printer
● a digital interface for transferring a protocol that contains all of the data via the network (alternatively by using an USB
stick)
Modern monitoring devices have a high performance battery so that they can be carried from one measuring location to
another.
The development and use of fully automatic monitoring systems has resulted in the modification of particle counters to
meet the requirements of these systems. This means that particle counters were more or less reduced to the measuring
cell and evaluation unit. The evaluation unit of devices for use in the pharmaceutical environment was reduced to 2
channels for particle sizes ≥0.5 and ≥5 µm . There is no control unit, only a status display. The sample volume flow can
be created using two different methods: an external system, usually a vacuum pump, or a vacuum pump integrated in
the device. A decision on which kind of system to used, depends on the spatial and structural situation. After each
measuring interval, the measurement data is transferred to the computer used for recording data via data interface.
These devices usually have an internal control sensor which checks the laser intensity and the actual sample volume
flow, for example. When a deviation occurs, a signal is sent to the computer via the data bus and the computer can then
trigger an alarm.
Counts for the individual particle size classes are available for each measuring cycle, defined by the measuring period
and sample volume. This data is referred to as raw data. It can usually be printed as cumulative and/or distributive data,
stored in an internal buffer or transferred directly to a PC with appropriate software via digital interface (usually RS-232,
RS-485 or by Ethernet). The data is displayed on an LCD display.
The particle counters that are often used in monitoring systems are completely controlled by PC. The particle counter
itself usually only has an LED status indicator.
In principle, manual particle counters can also be integrated into monitoring systems. However, this may not make
sense because of the high costs involved.
Figure 3.K-6 shows different types of devices.
● Top left: a hand-held particle counter with a sample volume of 2.83 l/min
● Centre: a portable battery-operated device with a sample volume of 28.3 l/min.
● Right: a particle counter with an built-in pump for integration into a monitoring system and without a control panel
● Bottom left: a particle counter without a vacuum pump with a sample volume of 2.83 or 28.3 l/min for integration into a
monitoring system
Figure 3.K-6 Particle counters from different manufacturers (source: Deha, Heimsheim)

8. 3.K.7 Taking and transporting air samples
To sample air for particle monitoring, a volume of air is taken from the cleanroom using the sampling probe and fed
through the sample tubing to the particle counter. Sampling, the transport of the air sample and, if applicable,
processing of the sample take place between the measuring location and the measuring device (see Figure 3.K-7).
Figure 3.K-7 Components of a sampling system
To ensure that the sample is representative and that no change has occurred during transport to the measuring device,
the actual values of the sample (temperature, concentration, air pressure) should correspond to the values measured at
the measuring location. The speed of the sample in the measuring cell of an optical particle counter is so high that it
can have an impact on small drops of liquid, for example, because high speed also causes pressure changes. This
could lead to incorrect measuring results.
When a hot gas sample is taken from a sterile tunnel, for example, the sample has to be processed.
In this example, the gas flow has to be cooled to below 50°C because conventional particle counters do not permit gas
flows above this temperature.
Assembly of equipment during sampling

9. To sample particles from the air of the cleanroom, sampling probes are used that suck in the sample volume flow and
transfer it to the particle counter. Figure 3.K-8 shows the standard assembly of equipment during sampling. The particle
counter is placed on the floor to ensure that it does not impact the sampling process. The sampling probe is positioned
on a stand at a height of 1 to 1.5 m (working height). The sample tubing is kept short to minimise particle loss in the
tubing.
Figure 3.K-8 Standard assembly of equipment for air particle measurement
Impact of the air flow
Samples can be taken from turbulent or laminar aerosol flows. Figure 3.K-9 shows the respective flow patterns. The
particles that can be seen here range in size from size 0.1 µm to 10 µm. The image on the right shows that the aerosol
flow is strongly deflected towards the sampling tubing.
Figure 3.K-9 Left: sampling in an area with a turbulent flow Right: super-isokinetic sampling in a laminar displacement
flow
Isokinetic sampling can only be carried out in the undisturbed flow areas of cleanrooms with laminar laminar
displacement flow. Isokinetic sampling is not necessarily required when measurement is carried out in cleanrooms
because most of the particles are smaller than 5 µm and larger particles are present in small numbers only. Particles <5
µm follow the air flow even at the very high flow rates that occur in cleanrooms
The sampling probe is usually positioned with the opening facing upwards. There is usually a turbulent flow in the
process area which means that the sampling probe can be pointed in a different direction.
Important criteria for sampling
The following criteria should, in theory, be met when samples are taken from an aerosol flow:
● Isokinetic sampling The main gas flow rate corresponds to the suction flow rate of the sampling gas, i.e. the sampling
probe is adjusted to the volume flow and air speed. For example, in the case of a turbulent displacement flow of 0.45
m/s and a sample volume flow rate of 28.3 l/min, the diameter of the sampling probe must be approximately 38 mm.
● Isoaxial sampling Ideally, the sampling probe should be facing the direction of the main gas flow.

10. ● Design of the sampling probe The sampling probe should have thin walls and sharp edges to ensure that the flow is
impacted as little as possible.
Figure 3.K-10 shows examples of sampling probes.
Figure 3.K-10 Left: stand-alone sampling probe with temperature and moisture sensor. Right: isokinetic sampling probe
in a safety cabinet (source: Deha, Heimsheim)
Sample transport requirements
The distance between the sampling location and the measurement cell should be kept as short as possible to limit the
transport duration and keep the particle loss caused by deposition on the walls of the tubing to a minimum. The
following factors have an impact on particle loss in the sampling system during the transport of the sample:
● the force of gravity that affects the particles
● particle inertia in bends
● electrostatic forces between the particle and the sample tubing
ISO 14644-3 "Metrology" contains information on particle loss during transport for a Reynolds number of 3000 (the
value should not drop below 3000). For example, if a sample volume flow rate of 28.3 l is generated in tubing with a 6-
mm diameter, the Reynolds number is approximately 6000. Assuming that the particle loss increases linearly with the
tubing length, the particle loss in sample tubing with a length of 1 to 2 metres is negligible for small particles.
The tubing should be made of materials with a smooth surface that also prevent static electricity. If electropolished
stainless steel pipes are used, static electricity is not an issue.
3.K.8 Calibration of particle counters
Calibration must be carried out to ensure the comparability of the measurement results. If the values measured during
calibration are outside the defined limits, the relevant parameters must be adjusted using the test piece. Adjustment
itself is not part of the calibration. After adjustment, recalibration has to be carried out. For detailed information on
calibration, please refer to Chapter 4.F Calibration.
When calibrating optical particle counters, the parameters of the test piece are compared with those of a reference
device. As already mentioned above, the following parameters are important for the measurement accuracy of a
particle counter:
● sample volume flow rate
● classification accuracy
● resolution
● counting efficiency
The sample volume flow rate must be calibrated because deviations from the sample volume flow rate have a direct
impact on the calculated particle concentration.
Particle counters can be calibrated on site or in the laboratories of the device suppliers or manufacturers. During
calibration, the resolution and classification accuracy are checked. Calibration is carried out in accordance with ISO
21501-4. Latex particles are usually used as calibration particles (see Figure 3.K-2 and Figure 3.K-11).

11. Figure 3.K-11 Analogue electronic signals from latex particles of identical size examined with an oscilloscope (source:
AC laboratory Spietz, Switzerland, 1988)
The latex suspension is diluted as required and sprayed using a generator. The water dries off during the process with
only the latex particles reaching the particle counter and reference device. This type of generator is described in detail in
the VDI Guideline 3491 Part 3. The classification accuracy and resolution for the measuring range of the particle
counter can be determined using latex particles of different sizes.
The calibration of the counting accuracy or counting efficiency is much more complex, especially when the most
sensitive channels are calibrated. This calibration can, therefore, only be carried out in a specially equipped laboratory.
Together with a particle generator that is operated continuously over an extended period of time, other equipment is
also required, e.g. a high-resolution reference device.
After the devices have been manufactured, an initial calibration is carried out by the manufacturer. An annual
recalibration is then required because of the wear and tear of some of the components, e.g. vacuum pumps, LEDs, etc.
It may be necessary to shorten the calibration interval if the devices are used in tough conditions that frequently change.
Comprehensive documentation is part of calibration and should include the following:
● identification of the device using the serial number
● documentation of the state of the device at the time of the recalibration (in particular, deviations from the target
values)
● list of test equipment and other types of equipment
● test processes and standards used (e.g. VDI guidelines, ISO standard or ASTM standards)
The calibration processes used in different institutions are not necessarily identical. For this reason, the process that
was used must be described to ensure that the documentation of the test device is complete.
3.K.9 Operating manual optical particle counters
The following recommendations are based on practical experience:
● The personnel carrying out particle counter measurement should be trained and familiar with the functions of the
devices. They should be trained in the consequences of measurement.
● Only calibrated devices may be used during measurement. This applies to the qualification of cleanroom facilities as
well as measurement during requalification.
● The calibration and servicing intervals specified by the device manufacturer must be observed. The date of the last
calibration should be indicated on the device.
● Particle counters must be commissioned in accordance with the user manual.
● If particle counters are used in fluctuating climatic conditions, an appropriate waiting time must be observed until
thermal balance is reached and the first measurement can be carried out.

12. ● When setting up a particle counter at the measurement location, it must be ensured that the exhaust air from the
cooling fans or the emission of sampling air do not impact the measurement location. The cooling air and discharged
sampling air may both be contaminated with particles. If necessary, the exhaust air must be discharged through a
hose and filtered.
● Before a measurement or a complete series of measurements is carried out, the sample volume flow rate must be
checked.
● If high concentrations are measured (e.g. in Grade C and D rooms), particles are usually deposited in the tubing
leading to the measurement device (sample tubing) which can be set free at a later date and cause incorrect results.
This can lead to incorrect results during subsequent use in Grade B and A rooms. For this reason, the zero count rate
should always be determined before measurement after a switch from one measuring location to another.
● If a particle counter is not used for an extended period of time between measurements, a protective cap or filter
should be attached to the sampling probe or sample input of the measuring cell.
Some of the functions of optical particle counters can be checked before measurement is carried out. The checks are
described below.
● Checking or determining the zero count rate Before measurement is carried out, the zero count rate should
always be determined. This should take place at the measuring location, i.e. in the cleanroom.
● Checking the sample volume flow rate Particle counters display the sample volume flow rate on the display or in
the printout of the measurement data. If there is a major deviation between the sample volume flow rate and the
target value, the cause must be determined and the device serviced, if necessary. A minimum or maximum deviation
for the sample volume flow rate is not specified. There is a linear relationship between the volume flow rate and the
concentration measurement. A deviation of 1 litre for a device with a sample volume of 28.3 l/min results in a
measurement error of about 3%.
● Checking the plausibility of the counts The person carrying out the measurement should have the ability to check
the latest measured particle concentrations for plausibility. This requires some experience working with particle
counters. Irrespective of the relevant limit values for 0.5 and 5 µm particle sizes, the complete particle size distribution
of a measurement should be checked for plausibility.
3.K.10 Manual particle monitoring in accordance with the monitoring plan
Regular particle monitoring measurements of the air must also be carried out in cleanrooms that are not equipped with
an automatic monitoring system (e.g. grade C and D cleanrooms). This is referred to as manual monitoring.
Depending on the layout of the room and the activities carried out in the room, a monitoring plan is created that
includes the following parameters:
● measurement or sampling locations
● measuring devices to be used
● measuring procedures to be followed
● frequency and duration of the measurements for each measurement point
● type and amount of manual documentation during the measurement
If particle measurement is carried out in operational mode, the activities (number of persons, operational processes,
etc.) in the rooms must be documented.
When monitoring new cleanrooms, the intervals between the individual measuring drives should be kept short so that a
relatively large amount of data can be collected in a few weeks.
A review of the results should then be carried out. This can show whether the results from individual sampling locations
are conspicuous or not. The insights gained from this process can be used to modify the monitoring plan, i.e. the
measuring locations and sampling frequency.
3.K.11 Automatic monitoring systems
The following section deals with automatic monitoring systems and focuses on particle measurement.
3.K.11.1 Types of monitoring systems
Since automatic monitoring systems were first used, companies have always developed their own system versions. The
performance and functionality of most of the systems make them suitable for general use. They can be used for all the
usual pharmaceutical applications (pharmacy, laboratory, production, etc.) and configured for each individual
application. There are 3 types of system:
● Systems with a recorder Data is recorded and saved using an electronic recorder.
● Computer-aided systems These systems are the most commonly used.
● Systems that are connected to the building control system (BCS) These systems use the available control and
regulating components of the ventilation system and provide the hardware required for monitoring. Special monitoring
software must be installed in the system because the control and regulating software cannot be used for monitoring.
For qualification, there must be a clear separation of control and regulating functions and monitoring functions.

13. The regulatory requirements are identical for all of these systems. All systems must also be operated in a GMP-
compliant way. However, the system requirements can vary depending on the type of cleanrooms and processes as
well as on the philosophy and requirements of the users. The basic requirement for GMP-compliance is that the
software has been specially developed for the purpose and that the configuration can be customised. For further
information on computerised monitoring systems, please refer to Chapter 3.J.6 Validation of a monitoring system in
accordance with GAMP® 5.
Figure 3.K-12 Schematic representation of a PC-based system (source: MT-Messtechnik, Adelzhausen)
Figure 3.K-12 shows the schematic representation of a PC-based monitoring system that is explained in detail below.
● The computer that records the measurement data is positioned close to the production area, i.e. close to the sensors.
This computer records the data at the configured intervals, saves it and transfers it to the network, e.g. to the server
(centre). The software of the evaluation computers evaluates the data online, i.e. it decides immediately whether a
warning threshold or limit value has been exceeded and generates an alarm via the user interface (LEDs, horn,
monitor) depending on the configuration.
● The server is used to back up the data and, if applicable, to configure the system.
● Several evaluation computers can be integrated in the network so that users who are not directly involved in
production can access the data. Users can carry out their assigned tasks based on the access rights granted to
different password levels.
● The particle measurement probes and necessary peripheral devices are at the heart of particle monitoring, in this
case an external vacuum system. With the help of network or bus-compatible particle counters, a number of different
measurement probes can be integrated in the system.
● The type and number of digital converters used with the analogue sensors that measure differential pressure, climatic
data from coolers, etc. always depends on the requirements of the user.
3.K.11.2 Main components of a particle monitoring system and their requirements
In principle, automatic systems contain the same components as mobile systems used to carry out manual
measurement (see Chapter 3.K.6 Conventional particle counters and counters used in monitoring systems). However,
some additional requirements have to be observed because the components are integrated in the system. These
requirements are described below.
● Sampling probe: The sampling probe measures the sample volume and is positioned directly in the cleanroom area.
Therefore, the material of the probe and the holder must be resistant to disinfecting agents. If spray disinfection is
carried out, it must be possible to protect the probe with a cap or filter in order to protect the sample tubing and
particle counter against aerosol droplets of disinfectant (see also Figure 3.K-14).
Figure 3.K-13 Schematic representation of a sampling probe installed through a wall. The sample tubing is protected
by the pipe and can be replaced at any time (source: MT-Messtechnik, Adelzhausen)

14. Figure 3.K-14 Sampling probe with a protective filter and cap, e.g. during disinfection of the sampling area (source: MT-
Messtechnik, Adelzhausen)
● Sample tubing: Tubing that is coated on the inside is used for most applications. Ideally, the tubing is installed in a
pipe that leads to the particle counter. This has the advantage of making it easy to replace the sample tubing. The
tubing has to be replaced after a certain amount of time because even if due care is taken, particle deposits can build
up inside the tubing, resulting in a zero count rate that is unacceptable.

15. ● Particle counters: The particle counters used in monitoring systems have already been described in the sections
above. Long-term stability in continuous monitoring devices is important. This applies, in particular, to the laser diode.
Depending on design, particle counters can have an integrated vacuum pump or the volume flow can be created
using an external vacuum system. If the vacuum pump is integrated in the particle counter, its availability and
operational stability must meet the same stringent requirements.
● Vacuum pump or vacuum system: If the volume flow is created using an external vacuum pump, the pump should
ideally be placed in the grey room, i.e. in the direct environment of the cleanroom area (see also Figure 3.K-15).
External vacuum pumps are a good choice if they have to supply several particle counters simultaneously. A
disadvantage of the system is that if the pump fails, the entire particle monitoring process comes to a standstill.
However, a second pump can be installed that is activated by a pressure-driven redundancy switching mechanism.
This is also useful if the vacuum pump has to be serviced while the cleanroom is being used. These types of vacuum
systems usually have a pressure monitoring system which records and monitors all data.
All monitoring systems have to include a data backup system that saves all of the recorded data. The data is stored by
the backup system in a tamper-proof way to ensure it is accessible at a later stage, e.g. for evaluation and trending.
This chapter does not cover user interfaces such as monitors or acoustic and visual alarms that are required for each
system, e.g. for user login and for alarm comments.
Figure 3.K-15 Left: example of a system with a vacuum pump and control cabinet for all of the electrical and sensor
controls. Right: vacuum distribution system with vacuum sensors and valves that are used to switch off the vacuum
supply to the individual particle counters. This system is useful if a particle counter has to be deinstalled or reinstalled
when a particle counter fails occurs or when calibrating the device.
3.K.11.3 Selection of sampling locations in Grade A and B areas
The selection of sampling locations for automatic particle measurement is time-consuming and requires a systematic
approach. The process is relatively simple in the case of Grade B cleanroom areas because of the turbulent mixed air
flow. The particles are spread around most of the room which means they can be easily detected by a sampling probe
installed at a fixed location.
The airflow in Grade A areas is usually a low-turbulence displacement flow that flows vertically from top to bottom.
Isokinetic sampling measures only a small part of the flow.
The following approach should be taken when searching for the correct sampling locations:
● Risk analysis It makes sense to carry out an initial risk analysis to define the areas for which a particle measurement
seems necessary.
● Process analysis When the areas have been defined, the exact sampling location has to be determined. On the one
hand, this should be as close as possible to the potential source of particles, e.g. at a filling area. On the other hand,
aerosols that are created, especially when low viscosity liquids are filled, can lead to high particle concentrations in

16. the filling area which result from the actual product filling process. For this reason, it is important to understand the
theoretical and practical implications of the process in order to select a suitable sampling location and interpret the
results correctly.
● Test measurements at the defined sampling locations A final check of the sampling locations is carried out by
taking particle counter test measurements during a simulation. A flow visualisation should ideally be carried out at the
same time that shows the actual flow pattern. The flow is usually no longer low-turbulence in the area where the
process is carried out.
3.K.11.4 Data evaluation and alarms
Two approaches can be taken when evaluating particle data:
● Online evaluation of the recorded data while measurement is in progress
● Evaluation after the measurement process has been completed
Specific requirements apply to online evaluation because the resulting data is used to determine if there are any
violations of the warning and action limits during ongoing production, and this information is then displayed for
personnel.
Alarm limit values must always be set for production-related parameters such as particle data and differential pressure.
In an ideal case, there should also be warning limits. The advantage of warning limits is that warnings (pre-alarms) are
displayed at an early stage if the measurement values increase. The user can then respond before a violation of a limit
value occurs which has to be commented. The warning limit values can be set individually, whereas the alarm limit
values are fixed in accordance with the cleanroom grades (see also Chapter 3.C.2 GMP Requirements for Cleanrooms:
Air Cleanliness Grades and Chapter 3.J.3.2 Alert and action limits).
The evaluation of particle data for trending or for analysing the course of concentration per measurement day or per
sampling location are of secondary importance, but should not be neglected. A trend analysis, for example, can show
the influence of the climatic conditions on cleanroom technology in winter and in summer. Similarly, this type of analysis
can be used to determine whether the cleanroom conditions deteriorate continuously during the filling period, even if the
limit values are not exceeded.
Major changes occurred with regard to the evaluation and interpretation of online data when the reference volume for
the classification of cleanrooms was changed from cubic feet (ft³) to cubic metres (m3) (cancellation of the Federal
Standard 209E in 2001). Particle counters used with automatic monitoring systems usually have a sample volume flow
of 1 cubic foot per minute. However, ISO 14644 specifies one cubic metre as a reference value for the classification (1
m3 = 35.3 ft³). This means that it takes 35 minutes for this type of device to reach a volume of 1 m3. This volume results
from 35 individual measurements at intervals of one minute. The limit value specified in Annex 1 refers to 1 m3. Whether
this limit value was exceeded or not can (and must!) only be evaluated based on the total number of particles after 35
individual measurements have been carried out. However, this (permitted!) process does not provide information about
when particles occurred and in what numbers. The two examples in Figure 3.K-16 show results for the measurement of
particles ≥0.5 µm in a Grade A area. The limit value is 3,520 particles per m3.
Figure 3.K-16 Examples of measurement results for particles ≥5 µm in Grade A rooms: both graphs show particle
measurement results at 1-minute intervals. In example 1 (top), the total number of particles is 2,600, in example 2
(bottom) 3,450 (source: MT-Messtechnik, Adelzhausen)

18. Discussion of the particle results shown in Figure 3.K-16
● During both measuring processes, a sample volume of 1 ft³/min was used for each measurement interval and each
measurement point. The required sample volume of 1 m3 was reached after 35 minutes (1 m3 = 35.3 ft³).
● Top graph: if a total limit value of 3,520 per 1 m3 is specified for the evaluation, the limit value is not exceeded. If,
however, the limit value refers to the sample volume of 1 ft³/min (3,520/35 = 100 particles), the limit value was
exceeded at 8 intervals during the 35 minutes. 750 particles/ft³ approx. was the highest value.
● Bottom graph: during measurement, the limit value for 1 ft³ was exceeded at 14 intervals. However, the highest value
was only 350 particles/ft³. Again, the total limit value of 3,520 particles for 1 m3 is not exceeded in this example.
In both cases, the value is below the limit if a reference volume of 1 m3 is used. Even if the limit value was exceeded,
the result would only be available after 35 minutes. The question of what happened in the previous 34 minutes then has
to be asked. If only the total value is measured, this cannot be traced.
If the individual measurement results per cubic foot are evaluated, an additional evaluation for each measurement can
be carried out at any time. If all of the results for the 35-minute period are added, the question of whether the limit value
was exceeded can be checked; however, it is not likely in this case.
The EU GMP Guidelines do not specify a process to be used for the evaluation and alarms. The user can decide
whether they evaluate the cumulative measurement result for 1 m3 or whether they take a closer look and use the
individual results for the evaluation. Both approaches are GMP-compliant.
A different approach is required when evaluating particle sizes ≥5µm. In this case, the limit value per cubic metre (m3) is
20. If one event occurs in one cubic foot (≥5µm ), the limit value is exceeded (the limit value for 1 ft³ would be 20/35, i.e.
<1). This also means that no additional 5-µm particle should be detected in the first or in the following 34 intervals.
There is a pragmatic way of correctly interpreting such individual measurements:
● If particle events with particles ≥ 5 µm are to be expected, they will not be restricted to the channel for this particle
size.
● A strong increase in the number of particles ≥0.5µm is also to be expected. If this is not the case, it is very likely that
this is a random event. In this sort of case, the next measurement interval has to be monitored to find out whether the
number of particles is still too high.
3.K Summary
Particle monitoring is an important part of cleanroom technology. Continuous and non-continuous (manual) monitoring
of air cleanliness contribute significantly to QA in a cleanroom. Different types of particle counters are available for
carrying out the measurements. The measuring principle based on the measurement of scattered light is the same in
all of the devices, regardless of whether they are used in manual or continuous monitoring systems. Calibration with
latex particles is carried out in the same way worldwide. This means that the measurement results for all cleanrooms
are comparable.
Automatic monitoring systems with different numbers of particle counters are available for different platforms. There
are solutions that use computers as control and data recording systems. There are also systems that are integrated in
the building control system and solutions that use built-in recorders. All of these systems have been developed and
designed in such a way that they facilitate GMP-compliant monitoring. Particle counters with reduced functionality are
available for use in monitoring systems. The selection of representative sampling locations can be a challenge. A risk
evaluation and flow visualisation including the respective test measurements must be carried out to find suitable
locations.