Main feature

Ensuring the air supply rate to a cleanroom

complies with the EU GGMP and ISO 14644-3
recovery rate requirements
W Whyte 1, N Lenegan 2 and T Eaton 3
1
School of Engineering, University of Glasgow, Glasgow G12 8QQ; 2 Energy and Carbon Reduction Solutions,
Ashton-Under-Lyne, Lancashire, OL5 0RF; 3 AstraZeneca, Macclesfield, Cheshire, SK10 2NA

Abstract According to the EU GGMP, a Grade B measurement, C0 is the initial

The European Union Guidelines to cleanroom should have a maximum concentration, and C is the
Good Manufacturing Practice (EU concentration of particles *0.5µm concentration after time t.
GGMP) includes a recommendation during operation of 352,000/m3 and in The recovery rate and recovery time
for a ‘clean-up’ of airborne particles the at rest condition of 3520/m3, and both measure the particle decay and one
in the cleanroom after completion of therefore the required maximum drop test result can be easily converted to the
operations, where the concentration in the concentration of particles during other. No required recovery rate or
should decay by up to 100, or 10 fold, ‘clean-up’ is 100 fold. In a Grade C recovery time is specified in ISO 14644-3,
in 15 to 20 minutes. When designing a room, the maximum ‘operational’ but the EU GGMP requirement of a 100
ventilation system for non-unidirectional condition is 3,520,000/m3 and the ‘at fold particle concentration reduction
airflow EU GGMP Grade B and C rest’ condition is 352,000/m3, which is a in less than 20 minutes is often applied
cleanrooms, it is necessary to determine 10 fold reduction. Particles *0.5µm are to cleanrooms not regulated by the
if the proposed air supply rate will be normally measured during the clean-up EU GGMP.
sufficient to provide the air change rate tests, and not particles *5µm, as the When a non-unidirectional airflow
for the clean-up specified in the EU larger particles give a greater decay rate cleanroom is being designed, the air
GGMP, and such a method is provided than actually occurs, because of particle supply rate has to be determined.
in this article. The air change rates for deposition onto cleanroom surfaces However, it is usually unclear whether
other decay times and reductions in caused by gravity (ISO 14644-3:2005 this air supply rate is sufficient to ensure
particle concentrations in cleanrooms suggests that test particles should be the ‘clean up’ requirements given in the
can also be calculated by this method. less than 1 µm). The decay time given EU GGMP. If the calculated air supply is
in the EU GGMP for both Grade B and insufficient the recovery time will be
Introduction C rooms is between 15 to 20 minutes longer than desired and the recovery
Annex 1 of the EU GGMP covers and, as a decay time of 15 minutes is rate slower. It is not until the cleanroom
various aspects of sterile products a more stringent requirement than 20 is built and manufacturing starts, that
manufacture, including contamination minutes, it is the time often applied. testing can be carried out, and it would
control. It suggests that manufacturing ISO 14644-3:2005 gives two recovery be useful if a method was available to
cleanrooms will quickly recover from tests to determine a cleanroom’s predict what air supply rate is necessary
the generation of high concentrations of ability to recover quickly from a high to ensure a specified reduction in
airborne particle contamination and concentration of particles, namely, particle concentration in a given time.
should ‘clean up’ in 15 to 20 minutes the recovery time and recovery rate. Carrying out a recovery test provides a
(guidance value) after completion of Test particles are introduced into the qualification test that gives confidence
operations, to the particle limit stated cleanroom and the recovery time is that the airflow in the non-UDAF
for the ‘at rest’ state. This applies to obtained by measuring the time for the cleanroom is well designed and will
cleanrooms designated Grade B and airborne particle concentration to decay provide effective particle removal.
Grade C, which have non-unidirectional by a factor of 100:1. If the recovery time The concentrations of airborne
airflow, but not to Grade A cleanrooms is set at 15 -20 minutes, this test is similar contamination in non-unidirectional
which have unidirectional airflow, to that specified by the EU GGMP for a airflow rooms can be calculated by a set
or to Grade D which has no specified Grade B cleanroom. of equations known as the ‘ventilation
‘operational’ particle concentration. The recovery rate is obtained from equations’, and their application to
Achieving this ‘clean up’ time is measurements of the decay of cleanrooms has been discussed by
considered to demonstrate that the concentration of test particles in the Whyte, Whyte and Eaton (2012).
cleanroom has a ventilation effectiveness cleanroom and calculated by Equation 1: These equations can determine the
that cannot be guaranteed by a simple concentrations of particles or microbe-
Equation 1
air supply rate specification which carrying particles (MCPs) in cleanroom
provides the required airborne air as the particle concentration (a)
contamination in the steady-state builds up when activity starts, (b)
Where, n = recovery rate, t is the time
condition. remains relatively steady during
elapsed between the first and second
manufacturing, or (c) decays when

22 Clean Air and Containment Review | Issue 26 | April 2016 www.cleanairandcontainment.com

 Main feature

activity stops. The decay equation can Or, when logarithms to the base 10 time of between 15 and 20 minutes is
be used to determine the air change rate are used, given in the EU GGMP as a ‘guidance’
required to satisfy the EU GGMP ‘clean time but 15 minutes is used as it is the
Equation 5
up’ requirements, or another specified most stringent requirement. Equation 5
recovery time or recovery rate of is used, as follows, to calculate the
particles measured by the ISO 14644-3 required air changes per hour.
recovery test methods.
It is interesting to note that the right
hand side of Equations 1 and 5 are
Decay and recovery rate equations
identical, and therefore the recovery
When people leave a cleanroom and
rate (n) is equal to the air change rate
machinery is turned off, the concentration If the same calculations are carried
(N) at the test location.
of particles in the cleanroom air will out for an EU GGMP Grade C room,
Equations 4 and 5 can be used to
decay. Similarly, when the introduction where the maximum ‘operational’
calculate the room air change rates (N)
of test particles is stopped during the concentration is 3,520,000/m3, the ‘at
required by a cleanroom to comply with
ISO 14644-3 recovery tests, there will be rest’ condition is 352000/m3, and a
the EU GGMP, or other recovery times
a decay of particles over time. These clean-up time assumed to be 15 minutes,
or rates. How this is done is illustrated
decays occur in an exponential way that the required air change rate per hour is 9.2.
by the following example, which calculates
is predicted by the Equation 2, and in The air change rates per hour obtained
the air change rate for EU GGMP Grade
the manner shown in Figure 1. for Grade B and C cleanrooms of 18.4
B and C cleanrooms, and can be used
and 9.2, respectively, are applicable
Equation 2 for all Grade B and Grade C cleanrooms
to all cleanrooms of the same grade,
as long as air mixing is effective.
although only if perfect air mixing
Where, C = airborne concentration is achieved throughout the cleanroom.
Calculating air change rates to
of particles after a given decay time, The next section of this article discusses
comply with the EU GGMP clean-
CO = initial airborne concentration how the result should be modified for
up requirements
of particles, N = room air change rate, situations where good air mixing is
An EU GGMP Grade B cleanroom is
t = decay time not achieved and there may be less air
being designed with an air supply rate
It is important to note that in Equation supply at the test location.
of 3.33 m3/s, and the question to be
2 it is the ‘air change rate’ that affects
answered is whether this air supply
the decay rate and not the ‘air supply rate’. Consideration of ventilation
is sufficient to achieve the clean-up
This is different from the steady-state effectiveness of cleanrooms
requirements given in EU GGMP for a
condition, where the particle concentration If efficient air diffusers are used to
Grade B cleanroom. The cleanroom has
in the operational condition during supply filtered air to a cleanroom, and
a floor area of 10m x10m and is 3m high
manufacture is determined by the air the room air is extracted at low-level
i.e. the air volume is 300m3, and the air
supply rate. The air change rate and exhausts around the cleanroom, then,
change rate is therefore 40 air changes
the air supply rate are related as shown as shown by Whyte et al (2014) and
per hour. According to the EU GGMP,
below. Lenegan (2014) there will be good air
the cleanroom should have a maximum
mixing between supply and cleanroom
Equation 3 concentration during operation of
air. However, if air mixing is not perfect,
particles *0.5µm of 352,000/m3, and this
then locations in the cleanroom will
concentration is taken as the ‘worst
receive less clean air than average and,
If Equation 2 is rearranged, the case’ initial concentration after activity
at that location, the decay of airborne
following equation is obtained stops. The particle concentration given
particles will be slower than required.
in the EU GGMP for particles *0.5µm in
Equation 4 The ventilation effectiveness at a
the ‘at rest’ condition is 3520/m3, which
location in a cleanroom can be determined
is the concentration that must be achieved
by measuring the air change rate at the
at the end of the clean-up test. A recovery
test location and comparing it to the
overall cleanroom average (Whyte et al,
ϰϬϬϬϬϬ
2014). The ratio of the air change rate
ϯϱϬϬϬϬ at the test location to the overall
ϯϬϬϬϬϬ cleanroom average is called the Air
Change Effectiveness (ACE) index.
ŽŶĐĞŶƚƌĂƚŝŽŶ;ͬŵϯͿ

ϮϱϬϬϬϬ

ϮϬϬϬϬϬ It is calculated as follows.

ϭϱϬϬϬϬ

ϭϬϬϬϬϬ
If the air mixing is perfect, the ACE
ϱϬϬϬϬ
index will be 1 but if the test location
Ϭ
Ϭ ϭϬϬ ϮϬϬ ϯϬϬ ϰϬϬ ϱϬϬ
receives more clean air, the ACE index
dŝŵĞ;ƐͿ will be higher than 1. Locations that
Figure 1 – Decay of small particles in a non-unidirectional cleanroom receive less clean air will have an ACE
supplied with 40 air changes per hour index lower than 1. When the ACE

www.cleanairandcontainment.com Clean Air and Containment Review | Issue 26 | April 2016 23

Main feature

index is less than the room average i.e. correct ‘clean-up’ performance. The requirements, or may specify a required
1, the air change rate will have to be correct air supply rate will be that particle decay in terms of recovery time
increased to achieve the required which is needed to provide for the most or recovery rate. Where the specification
clean-up, recovery time or recovery rate demanding of the four parameters. differs from that of the EU GGMP,
at the location under consideration. It has been previously difficult to Equations 4 or 5 can be used to calculate
It has been shown in a previous calculate the air supply rate needed for the required air change rate, and an
section of this article that the recovery the clean-up requirements of Annex 1 of example of how this is carried out is
rate (n) is the same as the air change the EU GGMP, or another recovery rate given in this article.
rate (N) at the test location, and this fact or time. This article describes a method
can be used to obtain the ACE index. to calculate the clean-up requirement. References
The decay of test particles introduced If the cleanroom has to comply with 1. Lenegan N (2014). Diffuser
into the cleanroom can be used to the EU GGMP and be capable of performance in cleanrooms. Clean
obtain the recovery rate, which is the reducing the airborne concentration in Air and Containment Review, Issue
same as the air change rate at the test a Grade B cleanroom by a 100 fold in 15 18, pp8-14.
location. The recovery rate (air change minutes, then, if there is perfect mixing
2. Whyte W, Whyte WM and Eaton T
rate) at the location can then be of supply and room air, an air change
(2012). The application of the
compared to the overall air change rate rate per hour of 18.4 is sufficient. If the
ventilation equations to cleanrooms:
of the cleanroom and the ACE index cleanroom is a Grade C, an air change
Part 1; The equations. Clean Air
obtained. Investigations carried out by rate per hour of 9.2 is required. These
and Containment Review, Issue 12,
Whyte et al (2014) showed that in two air change rate assume that the
pp4-8.
non-unidirectional airflow cleanrooms, supply and room air are perfectly mixed
with effective air diffusers fitted, the and that no location within the 3. Whyte W Ward S Whyte WM and
ACE index is unlikely to be less than 0.7. cleanroom receives less clean air than Eaton T (2014). Decay of airborne
An ACE index of 0.7 can therefore be other location. Good air mixing can contamination and ventilation
used to compensate for lower air change be achieved by means of efficient air effectiveness of cleanrooms.
rates that might occur at some locations diffusers and low level extracts around International Journal of Ventilation,
in this type of cleanroom. the cleanroom. Poor air mixing will 13(3), pp211-219.
The air change rate required to require more air to be supplied to
4. ISO 14644-3: 2005. Cleanrooms and
produce the correct ‘clean up’ in an EU the cleanroom but when efficient air
associated controlled environments.
GGMP Grade B cleanroom with perfect diffusers filters and low level extracts
Part 3: International Organization
air mixing was previously calculated are used, an ACE index of 0.7 will
for Standardization, Geneva,
to be 18.4/h. Applying an ACE index compensate for non-uniform air mixing.
Switzerland.
correction factor of 0.7, the minimum This will result in a required increased
air change rate per hour to ensure the air change rate for an EU GGMP Grade 5. EU GGMP (2008). The rules
correct clean-up should be increased B cleanroom of 27 per hour and for a governing medicinal products in the
from 18.4 to 26.3. For a Grade C Grade C cleanroom 13 per hour. European Union –Volume 4 -EU
cleanroom, where a 10 fold reduction The calculations in the previous guidelines to good manufacturing
within 15 minutes is required, the air paragraph assume the airborne practice – medicinal products for
change rate per hour should be concentration during operational human and veterinary use – Annex 1
increased from 9.2 to 13. conditions is the maximum acceptable –Manufacture of sterile medicinal
The air supply rate should now be by the EU GGMP, and the shortest products. European Commission,
checked in the example to make sure decay time of 15 minutes. However, Brussels.
that it is sufficient. In the example, an it would be unusual to find that the
air supply rate of 3.33m3/s, which is airborne concentration in the operational
Full biographical notes for
equivalent to an air change rate of 40 condition was as high as the maximum
W (Bill) Whyte are on page 10.
air changes per hour, was thought to allowed by the EU GGMP and, therefore,
be appropriate. This is greater than the the required drop in particle concentration
Tim Eaton, B.Sc. (chemistry), is
required air changes for the clean-up, would be less than 100-fold. This would
Sterile Manufacturing Specialist at
and the proposed air supply rate is require a lower air change e.g. a 10 fold
AstraZeneca. His full biographical
therefore sufficient. drop instead of a 100 fold would half the
notes may be found in CACR20
air change rate. Similarly, if a ‘clean up’
(October 2014) on page 9.
Discussion and Conclusions time of 20 minutes instead of 15 minutes
It is necessary when designing non- is acceptable, then the air change rate
Nigel Lenegan, B.Eng., C.Eng.,
unidirectional airflow cleanrooms to can also be reduced. The air change rate
MCIBSE, is founder and managing
ensure that the air supply rate will be for these less stringent requirements can
director of Energy & Carbon
sufficient to (a) achieve the correct be calculated by the method described
Reduction Solutions. His full
particle concentration in the cleanroom in this article.
biographical notes may be found
in the steady-state condition, (b) control Cleanrooms that are not regulated
in CACR18 on page 13.
the heat gains in the cleanroom (c) by the EU GGMP may have similar
compensate for room air leakage and clean-up requirements to those that are
process air exhaust, and (d) provide the regulated, or may have different clean-up

24 Clean Air and Containment Review | Issue 26 | April 2016 www.cleanairandcontainment.com