VWG DD003 Safe Change Paper Issue 01

Download as pdf or txt
Download as pdf or txt
You are on page 1of 30

NUCLEAR INDUSTRY GUIDANCE

NNVF Discussion
Document - Guidance on
“safe change” filter housing
design.
Issue 01

VMG DD003 June 2015

This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating good practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and
comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or
used for contractual agreements. Anyone who uses information from this document will therefore do so at their own risk.
Page | ii

Foreword

This document has been issued as a National Nuclear Ventilation Forum discussion document aimed
at formulating best practice within the UK Nuclear Industry for possible adoption by the Nuclear Site
License holding companies. As such it may contain ideas and comments which may not reflect the
consensus opinions of the NNVF attendees and should not be relied on as a source of information or
used for contractual agreements. Any one who uses information from this document will therefore do
so at their own risk.

Comments should, in the first instance, be sent to :-


Grant Hall
Bld C21.1
AWE(A), Aldermaston
Nr Reading
RG7 4PR
[email protected]

Acknowledgments: The following companies are acknowledged for their contributions in writing this
document and supplying photographs. AWE Plc, Sellafield Ltd, EDF, MC Air Filtration, Emcel and
Cam Fil Farr.

Current at Date of issue. 4 June 2015

RECORD OF REVISIONS

Document Issue Revision Date Changes Made

Issue 01 June 2015 First Issue

Page | ii This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at
formulating best practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it
may contain ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a
source of information or used for contractual agreements. Anyone who uses information from this document will therefore do so at their
own risk.
Page | iii

CONTENTS

RECORD OF REVISIONS ............................................................................................................................ II

1. INTRODUCTION ................................................................................................................................... 1

2. EARLY FILTER CONFIGURATIONS .................................................................................................... 1

3. THE ROLE OF CHEMICAL FILTERS IN FILTER HOUSING DESIGN. ............................................... 3

4. BAG CHANGE HOUSINGS ................................................................................................................... 3

5. SQUARE TYPE SAFE CHANGE FILTER HOUSINGS ......................................................................... 6

6. CIRCULAR PLUG IN SAFE CHANGE FILTER HOUSINGS ................................................................ 7

7. BAG CHANGE TECHNIQUES. ........................................................................................................... 10

8. CANISTER FILTERS ........................................................................................................................... 12

9. FILTERS WHICH ARE SAFE TO ENTER INTO THE WASTE STREAM. .......................................... 13

10. PUSH THROUGH FILTER HOUSINGS USING RADIAL FLOW FILTERS. ....................................... 14

11. “PUSH PUSH” CIRCULAR FILTERS. ................................................................................................. 22

12. COMPARISON OF FILTER CHANGING IN HEPA FILTERED TEMPORARY AIR MOVERS. .......... 24

13. ACTUAL CONTAMINATION RELEASES DURING FILTER CHANGING .......................................... 25

14. CONCLUSION ..................................................................................................................................... 26

Page | iii This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 1

1. INTRODUCTION

This paper has been written by the UK National Nuclear Ventilation Forum filter
subcommittee. This sub group has expert members from the UK Nuclear licence sites and
UK filtration supply chain. The document has been written to provide some education on the
meaning of “safe change” when applied to filter housings. As such it may contain ideas and
comments which may not reflect the consensus opinions of the NNVF attendees and should
not be relied on as a source of information or used for contractual agreements. Anyone who
uses information from this document will therefore do so at their own risk.

There are many misconceptions regarding the meaning of the word “safe change” when
applied to filter housing design. This document has been written to detail what “safe change”
means in practice for filter housing design and looks at safe for whom. This document also
examines the practical application in mobile fan/filter units for decommissioning.

This document should give the reader enough knowledge to recognise possible
contamination migration during filter changing and to dispel the myth surrounding the phrase
“safe change”.

2. EARLY FILTER CONFIGURATIONS

To understand how the term “safe change” was coined we first need to examine some of the
early less “safe” filter configurations.

Two popular filter configurations of the 1960’s were the Ladder Rack and the walk in Active
Filter Room. To change these filters man entry to the ladder rack plenum or Active Filter
Room is made wearing pressurised suits or suitable RPE. When changing the filters the
filter is simple removed from the wall of filters or housing and a clean unit inserted. The filter
change operative and the “clean” side of the housing/filter could easily become
contaminated. Therefore from a safety view these were not safe change for the filter
changing team or the down stream ductwork.

Note: Ladder Rack systems were still being installed in the 1980’s. However these were for
low activity applications, hence with reduced risk when changing

Fig 1. Ladder Rack Style Filter Housing (Filters Shown To The Left).

Page | 1 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 2

Fig 2. Modern Ladder Rack Design.

Fig 3. 1960’s Non Safe Change Housing In Exclusion Area.

Fig 3. Shows the “walk in” style of filter plant room, where the filter is simply withdrawn into the room
with no protection as the whole room was classified as a dirty area.

The non-safe change housing in the 1950’s normally were of the “re-packable” design. Basically the
filter wadding was made up larger than the filter housing and compressed into the housing, so it made
its own seals. The obvious problem here is as the wadding is removed the particulate would mostly sit
on the top of the wadding and may become airborne as it is pulled out the housing.

Fig 4. Re-packable non safe change housing and new wadding.

Page | 2 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 3

3. THE ROLE OF CHEMICAL FILTERS IN FILTER HOUSING


DESIGN.

Chemical filters were developed during both world wars to counter the effects of gas attacks. It is
documented by the DOE that the US Army Chemical Corps supplied filters to the Manhattan Project
nuclear weapon programme. There are many simularities with filter housings for chemical and
Nuclear protection. Its difficult to say where improvements in housing design came from. However the
important factor is that the effort applied improved the technology available for both industries.

Safe Change Systems within the Nuclear Industry tend to be used to keep contaminants from getting
out of an area or building however, systems can also be used where contaminants are trying to be
kept out or from gaining entry into an area/building. Examples could be in an extract system from a
process plant or alternatively on a ventilation inlet or recirculation system to a critical control room.

Military Chemical, Biological, Radiological, Nuclear (CBRN) applications tend to operate safe change
systems in a way that protects occupants of certain environments such as those within armoured
vehicles, ships, bunkers and buildings. Systems exists within vehicles that use the Push through style
of filter to eject contaminated filters from within the clean area into the contaminated zone. Of course
with vehicles that are challenged by CBRN agents the immediate reaction is to withdraw to a clean
zone before changing filters. Once at this clean zone the whole vehicle can be decontaminated and
then filters changed using protective suits etc.

However, clearly this process is not possible for building protection and in this instance buildings are
more likely to have two systems set up in parallel to allow one to be changed while the second is
offering protection to the environment. Isolation of each system is achieved in the same way as for
Nuclear type applications with the use of dampers. Changing of such filters can be complex and
extremely cumbersome since building CBRN filters tend to be significantly larger than those for
vehicles.

In Marine applications it is rare that the traditional bagging system is used when changing of filters.
Systems are available that simply cap off the contaminated filter before withdrawal and change. This
type of process leaves opportunities for cross contamination between clean and dirty sides of a
ventilation system.

It should be noted that for CBRN applications the large majority of filters tend to be built as a
combination of both HEPA and Carbon since protection is usually required for both chemical &
biological agents.

4. BAG CHANGE HOUSINGS

Ozonair produced a round “safe change” housing in the late 1950’s. These were still being installed in
the 1980’s relatively unchanged. Fig 4 shows a filter in position with the bag removed for clarity. The
handwheels on the lower part of the housing lift the filter up to seal on its top surface. The change bag
is on the clean side of the housing. So technically as the dirty filter is withdrawn into the bag the clean
side of the housing could becomes contaminated. However this was a good improvement from the
previous housing designs.

Page | 3 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 4

Fig 5. Ozonair Filter Housing.

Fig 6. Original Filter Change Sequence With Added Red Markings For Contamination Movement.

From these early configurations the square bodied “safe change” filter housing was developed which
is generally used today. The old filter could be removed from the housing into a PVC bag and then a
new filter put into the housing using a second new PVC bag, and inserted into the housing. As with
the round design in most circumstances this is “safe change” for the filter change team.

Page | 4 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 5

Fig 7. Modern “Safe Change” Housing with filter support table. Note the air flow is horizontal.

Fig 8a and b. Modern Multi Bank “Safe Change” Filter Housings With Pre-Filters (Top) and HEPA
Filter arrangement at the bottom.

There are several different styles of “safe change” housings ready to accept standard square filters.
The most common application is to have the flow going from top to bottom to help “retain” any
particulate in the filters as they are changed. In addition, using filters in this configuration will mean
that the seal face of the filters is uppermost such that they are not damaged or dragged upon
installation. However, as the filters are changed particulate can spread around the change bag and
onto the “clean” faces of the filter housing. If we have the down stream “clean” ductwork below the
filter, loose contamination will settle on the “clean” side. Therefore this type of housing, although may
protect the operative from the filter during change out, may not be considered as “safe change” for the
down stream ductwork.

Page | 5 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 6

The housing needs to be isolated from the headers either side during filter change as you would not
want the facility flow running through the housing when the filter is removed. This may be designed as
a complete system isolation where all filter may be changed without airflow passing through or
systems may be designed such that individual housings may be isolated so that other filters within the
bank arrangement may continue to operate albeit with a slightly increased flow rate until all are
brought back on line. On plant dampers are normally found above and beneath each housing. Also
very often the depression within the ventilation duct will be too great to be able to use the change bag
effectively if only one isolation valve was fitted. Because the isolation dampers either side of the
housing are normally a little leaky a small filtered air in-bleed is provided. This inflow can be adjusted
so it balances with the leakage of the isolation dampers and will provide a small depression in the
housing. It can therefore easily be seen that any contamination which makes its way to the clean side
of the housing during a filter change will either be carried into the clean side of the duct during a
change or when the ventilation is restored.

These types of housing are used for atmospheric abatement filtration and normally would be very
lightly loaded with contamination, with the majority of any particulate arising from the processes being
stopped by local process filtration. Therefore this very minor carry over of particulate is not normally a
problem, although an allowance should be made in the plant design for particulate carry over during
changing. What is important is the overall Decontamination Factor the housing provides over its life
time. The subject of the required facility Decontamination Factors is fairly complex and therefore can’t
be discussed in any great detail in this general document.

In practice the post filter atmospheric discharges are carefully measured. These measurements have
shown over many years that the discharges can be kept within the discharge limits set by the
Enviromental agencies when the plant is operated correctly.

These types of change methods are often more accurately described as “Bag In Bag Out”. This
description does not give the illusion of providing any safety for the down stream ductwork.

5. SQUARE TYPE SAFE CHANGE FILTER HOUSINGS

The square type arrangement preceded the now more common Circular type filters but can still be
found in many Nuclear sites across the UK and Europe. The movement towards the Circular type
housings and filters was driven by the more consistent sealing of the circular filter and by the
suitability to dispose of the circular filter without the need for dismantling or cutting up.

The picture below, see fig 9, shows a typical Prefilter and HEPA filter housing with safe change bag in
place for the prefilter but removed on the HEPA filter for clarity. The HEPA filter has seals fitted to the
upstream face that seat onto an internal flange within the main housing. The middle picture shows the
eccentric “Cam Bar” arrangement that is used to lift the filter into place and compress the seals within
the housing. This arrangement allows the filters to be inserted into the housing without dragging or
damaging the seals and the eccentric bars control the amount of compression on the seals to be set.
In the example shown the flow of air is from top to bottom of the unit although it is also possible to
mount similar types of units such that horizontal flows can be accommodated.

Collection of the particulate is on the top of the filter and within the pleats of the material and so there
is potential for this to be disturbed during filter change and thereby contaminant the clean side of the
housing however, unlike the circular filter there is no need to jerk the filter off its spigot and so
opportunity for disturbing particulate is limited.

Page | 6 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 7

Past experience has led to seals becoming over compressed due to oversized filters being fitted or
filters becoming stuck to the seal face of the housing due to seals being used that have not properly
been cured. In such instances it has been particularly difficult to withdraw the filters through the safe
change bags.

Due to the weight of HEPA filters (particularly Type II) some systems employ a support table or
platform to withdraw the HEPA filter onto however, this can make changing of the filter more complex.

Fig 9. Square Safe Change.

6. CIRCULAR PLUG IN SAFE CHANGE FILTER HOUSINGS


Fig 10 Modern “Safe Change” Plug In Filter Housing.

Page | 7 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 8

Fig 10 shows a plug in filter being removed from its filter housing, with the change bag not in place for
clarity. The filter seals onto a round spigot, which can be seen at the back of the housing on the left.
Here the flow of air is normally from inside of the filter to the outside. As the filter is changed this helps
to keep any loose particulate off the clean side of the housing.

The circular plug in filter was developed by AEA Technology during the 1970’s and aims to provide a
more reliable sealing arrangement to the proceeding square filter, while also being directly disposable
into a waste drum without size reduction.

This type of housing suffers from the same problem of potential contamination carry over during a
change as the square safe change filter housing. However where the flow is from the inside to the
outside for the filter element the geometry lends itself better to retain any loose particulate.

There is a style of multiple filter “Plug In” housing, the EP2, which has multiple plug in filters. This style
was originally introduced with all the filter flow coming into the centre of the filters and radially flowing
outwards, identical to the single housings. The inlet and outlet were both from the top. However in
some installations this original design has been modified, to save space the air first enters the inner
diameter of the first filter and then flows radially inwards through the second filter, i.e. the flow
direction through the second filter is the wrong way to normal to save space. However by the time the
air stream has reached the second filter it is practically particulate free so this compromise is not
normally a problem, although the variety of derivatives may start to become confusing to some
building operatives. There may be good reason why the standard filter housing design has been
modified, i.e. design flows may have changed and the now built plant room is not big enough. Any
modifications normally introduce other issues so great care and thought need to be applied to
changes. With the modified EP2 combining two filter stages, the mixing distances for DOP, etc. cant
be fully achieved. The use of ‘quick mix’ injection for DOP does not reduce the mixing distance to
zero, so many of the ‘alternative arrangements’ are not fully compliant and leave the operators with a
problem of proving that there efficiency testing results are sufficiently accurate.

With this housing the filter slide back on two rails before it is lifted onto the back spigot. There have
been reports of the rails having sharp edges so damaging the filter as it is inserted. Because of this
problem these rails need to be nicely rounded. See fig 11 for poorly rounder guide rails.

Fig 11. Guide Rails With Insufficient Corner Rounding.

Page | 8 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 9

When this “Plug In” filter is changed from a horizontal to a vertical orientation the amount of particulate
retention will change.

With sintered metallic liquid process filters, which are basically the same shape as their larger air filter
cousins, the elements can be mounted with the removable bowl on top of the mounting spigot or
below, see fig 12 and 13. Note; Pall Europe now advise that all filter elements should be exclusively
mounted with the bowl downwards. With the upper filter housing in the sketch at fig 12 the element
can be pulled off the housing spigot with the particulate collected in the removable element retained in
the element. However with the lower filter housing in fig 12, when the filter element is removed the
particulate collected has a high chance of dripping onto the clean side of the filter housing, which is
unexpectedly what is actually found in practice. This would effectively render this micron particulate
removal filter useless. The same should be true for its larger air filter cousin to some extent. The liquid
plug in filter has been further developed to enable a bowl upwards configuration by adding a weir to
the filter element for remote change applications. However this still wont be as effective as the bowl
downwards configuration.

Fig 12. Standard Liquid Plug In Filter

Fig 13. Sketch Of Poor Liquid Plug In Filter Installation

Bowl downwards (good change)

Bowl upwards (very poor change)

Page | 9 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 10

7. BAG CHANGE TECHNIQUES.

When bagging in/out filters through the change bag there are several popular techniques to separate
the filter from the housing. The technique for bagging out final stage filters is normally different than
for bagging out components from active glovebox lines. The reason for this is that the filters bag is on
the clean side of the filter housing and so should be contaminated to many orders less than the
average glovebox, except for the push through wall mounted types. Also the change out bags for
filters are normally much larger than used for gloveboxes, although this is not always the case.

The bags are either cut with scissors/knife or using a hot wire. It is believed that the hot wire will seal
in any loose contamination.

A popular method in the UK is to swan neck the bag and then cut and tape the loose ends. The filters
needs to be rotated to form the swan neck and is sometimes lowered onto a lower filter change table
and simply rotated, see fig 14 or just simply rotated. Some of the air around the filter may be pushed
back into the clean side of the housing.

Fig 14. Swan Necking At Bag Out.

Portion twisted and wrapped with tape

Housing

Filter

Fig 15. Swan Necking.

Cable ties or Steel banding is also commonly used. These are positioned to neck the bag down and
the bag cut between the ties/bands.

Page | 10 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 11

Bag welding techniques are also available for filter changes which are identical to that used for more
hazardous glovebox work.

Fig 16. Bands applied to a change bag before separation.

Fig 17. Cable Ties And Scissors.

Fig 18. Camfil Bag Welding Method (Weld, Cut and Seperation).

Page | 11 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 12

8. CANISTER FILTERS

Although there are normally no claims that canister filters are safe change, they do offer some safety
advantages. Techniques have been developed for changing canister filters using ventilated
enclosures or change bags. The main advantage for this type of filter is that normally during a filter
change there is little likelihood of particulate migration to the down stream ductwork, so they achieve
a good safety rating from a down stream ductwork view, although they have a poor safe change rating
for the person performing the change and contamination spread to the laboratory. For some of the
smaller canister filters there is a large selection of end fittings available. The standard 25 l/sec has 4
end flange designs, so it would be good practice to double check that you have the correct
replacement filter derivative before removing the old one.

Fig 19 Picture shows a 5Ltr/s Process HEPA Filter

Page | 12 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 13

9. FILTERS WHICH ARE SAFE TO ENTER INTO THE WASTE


STREAM.

Under certain circumstances the holdup with a HEPA filter can require it to enter the intermediate
level waste stream. In short, for the current low level RA waste stream the filter can enter the waste
route as a double wrapped package. For intermediate level waste the disposal route requires the
filters to fit into the standard oil drum. Where standard square 609 x 609 filters are disposed of this will
require their size reduction to allow for this disposal route. Unfortunately size reducing a HEPA filter
which is contaminated to an extent where it is classified as intermediately level waste is not a risk free
operation. To lessen this size reduction hazard a split 609 x 609 unit has been manufactured, which in
operation acts similar to the standard filter, but each half of the filter can be directly disposed of within
the standard oil drum package. In this way we now have a filter which is “safe” to enter the
intermediate level waste stream.

Fig 20. Split 609 x 609 unit for “Safe” Intermediate Level Waste disposal (fitted with split pre-filters).

Page | 13 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 14

10.PUSH THROUGH FILTER HOUSINGS USING RADIAL


FLOW FILTERS.

There are several different types of push through filter housings and push through filters. The main
distinction in these filters is between the designs which have basically radial or axial flow. This section
will look at the advantages and disadvantages of radial flow push through filters, with the next section
looking at axial flow push through filters, which are often called “Push Push” filters.

Radial flow push through filters are normally used to accommodate gas process flows. As with the
other filters we will first look at the hazard to the operator and then the hazard to the down stream
ductwork. The radial push through filters come in standard sizes ranging from 12.5 l/sec to 160 l/sec.

Fig 21. Photo Of Radial Push Through Filters.

Page | 14 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 15

Fig 22. Radial push through filter housing types.

Original Harwell
Design Wall Mounted Unit

Stand By Filter
Stand By Filter

Single Filter Housing


Primary(Duty) Filter

With these types of filters there are three basic types which vary the hazard to the downstream
ductwork and filter change team. We have the original “Harwell” design, the modified one filter
element housing and the wall mounted design. With the original design the maintenance man is
protected from particulate release by the “Stand By” Filter. However with this design it was recognised
that the “Stand By” filter may be 20 years old before it is put into service. The truncated single filter
housing version was then developed. To protect the maintenance crew either the new filter is placed
in a special filter changing tube or a bag placed over the housing. During filter changing there is the
risk that the ventilation duct will be exposed to the laboratory as the secondary filter is pushed through
if precautions are not taken. This could risk tripping the ventilation system. The third type of housing is
the wall mounted unit where the filter is displaced into a catch bag. However with all these systems
the risk to the maintenance crew from contamination spread are minor with no problems reported on
the sites where these types of filter housings are installed.

The risk to the down stream ductwork is a little more complex. The risk can be explained from the
sketches of the changing process for each type of element, fig 16.

Page | 15 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 16

During the Filter Element change it is apparent that if the glovebox inlet flow and process gasses are
not stopped, potentially dirty gas will be passed directly to the extract duct when the filter insert is
bypassed when the seal of the used and secondary filters are passed by the extract duct. This may
lead to some contamination entering the extract duct, although it should be noted that HEPA filters
are not absolute and will let through 0.02% of the challenge at the most penetrating particle size for a
99.98% filter during their normal use, i.e. the downstream duct will not be guaranteed as
contamination free. It should also be noted that the extract is Blocked for part of the change. These
effects can be seen in Figure 23. Also during the last part of the filter change the primary containment
will be slightly pressurised as the used filter is pushed in. This blockage situation is normally
overcome by providing Run and Standby filter housings or by using a small filtered bypass line to
provide some ventilation for filter changing.

Fig 23. Original Standard Harwell Push Through Filter Housing Design.(Type A)
Stand By Filter

Standard Air Flow


Housing
Direction Position 1b (Mid
Position 1a. Blocked
Change)
Position 1c (Mid
Air by Pass
Change)
Extract Duct

Page | 16 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 17

With a derivative of the standard original Harwell push-through filter housing the design has been
slightly modified giving an “Engineered Gap” between the primary filter and the containment, see fig
24. This prevents the “passing” of the filter element during a change, see fig 17 position 17b, however
the “Engineered Gap” means the filter can not be tipped/rotated to release the seal to allow it to be
removed from the housing, see fig 24 position 24c. This is only a problem where the extract needs to
be isolated from the containment, which is usual where Run and Standby housings are provided, in
such cases a filtered relief in-bleed needs to be fitted into the rear of the housing to stop a vacuum
being formed behind the filter.

Where there is a single extract on the containment the “Engineered Gap” will lead to a greater
pressurisation of the containment when the sealed filter end is passing through the lower portion of
the filter housing during a filter change. Based on a 35 l/sec filter of 200mm dia and 300mm length the
filter displacement will be approximately 9.5 litres with the pressurisation relative to the glovebox
volume. In some designs a low capacity filtered extract line is provided to prevent pressurisation of the
glovebox during filter changes.

Fig 24. Radial Circular HEPA Filter With Engineered Gap.


Stand By Filter

Extract Duct
Primary (Duty)
Filter

Engineered
Gap

Standard Air Position 24b


Flow Direction
Position 24a.
No By Pass As Filter Front
Low capacity Filtered Extract To Prevent
Seal Is Still In Housing Position 243c
Pressurisation During Filter Change

As detailed earlier the HEPA filters are not absolute protectors and a very small amount of activity
may be expected to pass the element. If these particles settle near the upper seal and the standby
filter is pushed too far into the housing or a slightly shorter filter fitted, this may expose a ring of
contamination. It can thus be seen that filter positioning is more important with single filter designs
then with dual filter units, see fig 25. However monitoring of actual in-service extract systems show
that the extract ductwork remains very clean and not the problem it was once expected to be.

Page | 17 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 18

Fig 25. Single Filter Radial Flow Circular Filter Housings

Type C – Single Filter Type D – Single Filter


Housing With No Housing With Possible Contamination
Engineered Gap Area
Engineered Gap

See Enlarged View “A”

(Duty) Filter
Primary
(Duty) Filter
Primary

Enlarged View “A”

Engineered Gap

Standard Air Flow Direction Extract Duct

A further development of the push through housing design is the remote wall mounted unit where the
old primary filter is posted into a filter change bag, see fig 26. With the standard wall mounted design
the filter seals are by-passed during a filter change as the extract is placed less than a filter length
from the inlet. Increasing the distance between the extract and inlet will then cause the primary filter to
be delivered further away from the bagging ring, see fig 27. It has been found that the compromise to
make this non by-pass design operational is to introduce spacers between the secondary and primary
filters, with the added waste disposal during a change. In some cases the spacer has been bonded
directly to the rear of the Filter Element so it is not forgotten during a change. Serious thought should
be given to the number of filter changes needed to be made over the facilities life and multiple
canister filters considered, although most filter changes are initiated through penetration test failure
(Note: Normally Push Through Filter Housings are not in situ tested). Both of the standard wall
mounted units suffer from problems associated with having a standby filter which may be very old
when put into use. It would be possible to use a single filter wall mounted unit. However in practice
these are not normally used.

It should be noted that the wall mounted push through filter housing is the only “safe change” design
which has the change bag on the “dirty” side of the housing.

Page | 18 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 19

Fig 26 Standard Radial Flow Circular Wall Mounted Push Through Filter Housing.

Stand By Filter
Primary (Duty)

Extract

Duct
Filter

Filter By
Passed During
Change
Inlet

During
Duct

Change

Filter Movement
Prevention Device
(Re-Used After
Change)

Page | 19 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 20

Fig 27 Radial Flow Circular Push Through Filter Housing With Extended Distance Between Inlet And
Extract Ducts.

Spacers

Position 5a Position 5c
Standby Filter
Stand By Filter

Exhaust
Primary (Duty)

Duct
Filter

Extract
Blocked

Inlet

Duct

Filter Movement
Prevention Device
(Re-Used After
Change)

Page | 20 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 21

There is not always a choice due to space constraints if the filter housings can be vertical or
horizontal. However there are some technical issues which must be understood with both designs.
With vertical filter housings, if the filter is not properly vertically constrained it has been known for the
filter insert to slowly creep down and start to extend into the glovebox, thus no longer providing any
filtration, see the mid position filter element in Fig 23. This problem would not normally occur in
horizontal units.

Both types of housing have been known to have Filter Elements incorrectly positioned within the
housing. This could lead to the housing being blocked by the end of the Filter Element or for the
element to be bypassed. Therefore all designs need a means to ensure the Filter Elements are
placed in the correct position in the housing and that they stay there.

With horizontal housings as particulate is disturbed and settles within a Glovebox the highest dust
levels will be on horizontal surfaces, when compared with a vertical surface. Thus if there is any
horizontal shelf at the front of the filter housing particulate will be collected here which could be wiped
under the front filter seal when changing, see fig 28. Obviously here the original Harwell design has
the shortest horizontal shelf but may also suffer from by-pass problems.

Theoretically if the housing and Filter Element have been incorrectly toleranced in a horizontal design,
the Filter Element could settle leading to a passing above the Filter Element. However the Housing
and Filter Elements tolerances as normally defined have not led to any problems of this type being
reported.

Fig 28. Horizontal Filter Housings

Horizontal Type C
Inlet Housing

Glovebox
Possible Contamination Shelf
Wall

Horizontal Type D Housing


Inlet

Page | 21 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 22

11.“PUSH PUSH” CIRCULAR FILTERS.

Circular filters which have an axial flow direction can be mounted so the flow passes through two sets
of filter packs. These have been termed “push push” filters, and are more common in the
Pharmaceutical then the UK Nuclear Industry, see fig 29.

Fig 29. Photo of a Push Push filter element.

Fig 30 Example of a Push - Push Housing.

Page | 22 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 23

Fig 31. A Push Push Filter Assembly In Operation.

Fig 32. A Push Push Filter Element Change.

There is little experience of “Push Push” filters in the UK Nuclear industry, so this section has not
been written from lessons learner, but only from a desk based analysis.

The obvious advantage of these filters is that the housing could contain two stages of series filtration.
These filters are normally manufactured to a H14 standard, therefore giving a very high
decontamination factor. The down side would be the higher pressure loss when compared with a
single filter.

When changing these filters the containment the housing is attached to would not suffer the
pressurisation which the standard radial circular filter gives with its sealed rear end. The maintenance
person would be exposed to the containment via one or two stages of filtration, depending on the
housing configuration during the change.

As the rear of the housing is opened, to insert another filter, the extract line is exposed to the
laboratory. If the extract has been designed to shut down on loss of depression, this may cause the
extract fan to trip off or cause a change over of the fans set. What may be required is the filter
housing to be isolated during a filter change.

Filter positioning should be as important as with the radial circular filters. Therefore provision to hold
the filters in the correct position will need to be seriously considered.

Page | 23 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 24

There are obvious advantages and disadvantages over the radial flow filters. The Push Push filters
look to be better suited to reverse purge primary extract type applications on air boxes over Radial
Filters as two serial filters can be in operation together.

12.COMPARISON OF FILTER CHANGING IN HEPA FILTERED


TEMPORARY AIR MOVERS.

With modern mobile HEPA filter air movers the HEPA filter housings can either be “safe change” or
bolt together non-safe change, see fig 33 for a photo of a modern safe change unit.

At AWE these units are normally used downstream of a tested HEPA filter, and so should remain
relatively contamination free.

Fig 33. Nuclear Standard “Safe Change” 950 l/sec Plug In Mini Mobile Fan and Filtration Unit.

Air In

Air Out

The “safe change” unit normally does no afford the filter change team much of an improvement in
safety as the clean side of the housing is exposed to the dirty side of the filter during a change. The
monitoring access is normally difficult due to the size of the units. Any contamination which is moved
to the clean side of the housing during the change will be blown out of the unit when it is next started
up. Due to the size of these units it is not normally practical to move them inside an enclosure for filter
changing and monitoring. People also have the mis-conception that these units are “safe change” for
the environment and the filter change team, leading to less precautions being taken than may be
required.

For the non-safe change industrial units, fig 34, the air mover needs to be disassembled within a
suitable enclosure, should some contamination become airborne during the filter change process.
Once disassembled the inner faces of the filter housing can be fully monitored and cleaned. If the
contamination levels after this process are low enough the filtration unit can be removed from the
enclosure for use on the next task.

Page | 24 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.
Page | 25

Fig 34. Standard industrial air mover with square filter.

These two cases of a circular plug in filter or a square filter are not ideal. Operators of this type of
equipment need to control the amount of activity reaching these air movers if they are to be re-used.
Given they both have high particulate arrestance efficiencies, then in theory all the RA reaching them
should get caught in their filters. The use of a terminal filter before the air mover will keep
contamination levels at the second stage filter to a level where they can be safely changed using
RPE. There is therefore in practice very little contamination spread problems if terminal HEPA filters
are used.

13.ACTUAL CONTAMINATION RELEASES DURING FILTER


CHANGING

There has been very little research into how much particulate is actually release during a filter change.
It is expected that there are many contributing factors which will determine the potential for a release,
some of the major factors will be:- housing/filter design, volume of free particulate, mobility of
particulate, care taken during the change, etc…..

There have been reports of spikes in facility discharges during filter changing, but very little evidence
on the magnitude of these “spikes” has been collected to give any credence or validity to their actual
significance.

From a discussion with building operators who managed the non-safe change filter design in Fig 3 it
was hope that a rough estimate of particulate release could be made. However these filters were in a
pressurised suit area where loose contamination was expected. The management regimes were
concentrated on contamination levels at the pressurised suit entry area, looking to limit the spread of
contamination to the other building rooms. The highest contamination levels from the surveys found
were around the Electrostatic Precipitation filters which required to be opened up and washed down
from time to time.

Several “clean room” managers were interviewed to see if they had the answer. However as they
understood the situation particulate would be released when changing filters. Therefore precautions
were taken to limit the spread of particulates into the clean areas using flexible tents and PVC
sheeting where possible. No measurements of the particulate levels released were taken. The filters
are simply very carefully removed, replaced and clean up measures applied. Only after the full clean
up measures have been applied is the clean room counted for particulate.

Page | 25 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best
practice within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain
ideas and comments which may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of
information or used for contractual agreements. Any one who uses information from this document will therefore do so at their own risk.
Page | 26

14.CONCLUSION

There are many different styles of filter housings and the amount of protection a “safe change” system
will offer the down stream ductwork and filter change crew will vary enormously.

This paper only looks at the filter housing styles from a contamination perspective. In areas of high
radiation the doses received by the filter change teams may be excessively high if a full “safe change”
procedure is followed. Therefore in high radiation areas the full safe change method may not be
followed. Alternatively high radiation filters may need to be changed on radiation levels and not on
dust blockage or efficiency.
When we consider the safety of a filter housing design we need to consider:-
A, The possibility of particulate escaping from the housing during a change.
B, The possibility of the clean side of the housing becoming contaminated during a change and
migrating to the downstream ductwork.
C, Any size reduction of the filter required to allow it to enter the waste stream.
D, The radiation levels the filter change crew will be exposed to when changing the filter.

Abatement of the process arisings is a full topic on its own. Generally facilities are designed to contain
the majority of the loose contamination within the process containments with the later stage
atmospheric discharge filters being very lightly loaded. Therefore the choice of filter housings can only
be determined though an analysis of the process flow sheets, hazards, discharge limits and site
standards.

The designer must understand how the plant will be operated and maintained. From a close analysis
of what will happen to the contamination on the filters as they are changed we can see if the design
will meet the intent of the abatement system. In hindsight it is clear that the intent of using a vertically
mounted liquid “plug in” filter (bowl upwards) will cause down stream problems after filter changing.
Unfortunately this type of issue would likely not be picked up during commissioning. The advice on
how these filters are used is now changed so they are mounted with the filter below the housing. It is
only though questioning how operation and maintenance will affect the plant can the design be
improved and the correct equipment developed and installed. The significant design information must
be passed to the operators and maintainers of the plant to ensure it can be used safely. This
information exchange is especially true where standard filter designs have been modified so slightly
different change procedures or ancillary support equipment is required.

It is therefore up to all of us to assess our hazards and question if we have the correct equipment
available for the task. As can be seen from this simple analysis of what “safe change” means, with
some thought and reflection on operating experience we can better understand the tools at our
disposal and the tasks which they are suitable to perform.

Page | 26 This document has been issued as a National Nuclear Ventilation Forum discussion document aimed at formulating best practice
within the UK Nuclear Industry for possible adoption by the UK Nuclear Site License holding companies. As such it may contain ideas and comments which
may not reflect the consensus opinions of the NNVF attendees and should not be relied on as a source of information or used for contractual agreements.
Any one who uses information from this document will therefore do so at their own risk.

You might also like