Technical Note No 41

Site testing for watertightness

Site testing is frequently used to test the watertightness of windows and cladding as installed,
and is specified in Test Methods for Curtain Walling (CWCT, 1996). This technical note
adds further clarification on site testing procedures and equipment and the selection of test
method. It should be read in conjunction with Test Methods for Curtain Walling.

This Technical Note is based on and replaces TN10 first published in January 2000. It
incorporates changes following the publication of standard BS EN 13051.

Introduction The frame of reference

Although many cladding components and An important requirement before carrying

systems can be tested for watertightness out site testing is to have a frame of
in the laboratory or on a large scale mock- reference - the assessor must know
up, these tests neglect a critical issue with whether certain parts of a component or
watertightness - the impact of site system are capable of passing the
workmanship. specified test when properly fabricated
and installed. This is simple to define
The fabricator and installer of a cladding when a component or large-scale
system are relied upon to ensure that the specimen has been successfully tested in
joined surfaces of components are cut the laboratory. The site testing procedure
straight, gaskets properly fitted, and can then be applied at the laboratory to
sealants properly installed. However, the determine if the test is suitable.
installer is often left to resolve This approach will also generate a second
intersections between joints, overcome piece of important information - which
inaccuracies in the as-built structure and components or parts of the system will not
ensure proper sealing to adjacent cladding pass the test. It is known, for example,
systems. that the hose test generates a strong jet of
water with a penetrating power far in
For this reason it is often appropriate to excess of normal driven rain; this test will
test a small part of the installed cladding usually fail joints which are intended to be
system, to ensure that fabrication and opened (for example around doors and
installation have not in any way reduced opening lights of windows), unless a
the performance of the system, and to modification to the test procedure is made.
check the performance of interfaces with It is often possible, on a test mock-up, to
adjacent systems that did not form part of modify the parameters for a site hose test
the laboratory test. However, site testing to determine the condition under which an
itself can also be poorly applied, and this opening joint will pass the test with the
technical note aims to identify some of the agreement of all parties.
key issues of which the site test specialist
and specifier should be aware.

© CWCT 2004 January 2004

This document has been printed from the CWCT ‘Cladding Forum’, access to which is restricted to subscribing Members of the
Centre for Window & Cladding Technology. Information about the availability of CWCT publications and membership is
available at our website – www.cwct.co.uk - or from the address at the end of this note.
Site testing for watertightness TN No: 41

Specifying and witnessing site tests comfortably - these softer seals are
intended to prevent penetration of run-off
Non-specialist specifiers should seek water but are easily pushed aside by the
advice from a UKAS accredited test jet from a hose.
laboratory or cladding consultant on how If the hose test is used on open or
to specify and witness site tests. opening joints then the normal procedure
is to reduce the pressure of water entering
the nozzle appropriate for the joint under
Site tests test, and to maintain the distance from the
nozzle to the joint. This will ensure a
Watertightness can be assessed on site meaningful test and is preferable to
using three distinct approaches - the hose, holding the nozzle further away from the
the spray bar and the cabinet. joint as it does not require any change of
Hose testing action on the part of the operator.
However, the spray bar test is usually
This test is defined in Test Methods for more applicable to this type of joint.
Curtain Walling (CWCT, 1996) and the
AAMA standard 501-94. The CWCT hose Where it is not possible to obtain this
test varies from the AAMA test only in that frame of reference at the time of
joints within 120mm of each other can be laboratory testing, the following procedure
tested in one pass, providing there are no can be followed.
projections/obstructions that shield the
joint. All parties should agree on a good quality
window in terms of both fabrication and
Hose testing uses a compressor to drive a installation and this is used to determine
flow of water through a nozzle, forming a the frame of reference.
strong jet of water droplets. The nozzle is
defined, as is the pressure of water Where this benchmarking process results
entering the nozzle and the flow rate of in a very low nozzle pressure, say below
water through the nozzle. The water jet is 150kPa, or one party considers the results
always aimed perpendicular to the plane to be questionable, or the component
of the cladding system, and at a fixed selected for benchmarking to be
distance from it. unacceptable, it is recommended that a
cabinet test be carried out to determine if
Hose testing is intended for the testing of the selected component/area can achieve
permanently sealed joints. The high the specified performance. If the results
pressure water spray should not dislodge are successful, the hose test can be
gaskets or wet-applied sealants unless carried out on that area to determine the
they have been poorly installed or not frame of reference for further testing.’
been allowed to cure. Water will be forced
through small gaps in these types of seal, When testing joints at an internal corner
and will find its way through unsealed the hose should be positioned 0.3m from
joints between framing components. both walls (rather than 0.3m from the joint)
to take account of the water re-directed off
Hose testing is not suitable, in its the walls and towards the joint.
unmodified form, for use on open joints
(even if baffled, the flow of water from the The hose test is also suitable for use on
nozzle will overwhelm many open joints) sloped claddings, providing the jet is
or joints which are intended to open. aimed perpendicular to the joint. If a
Weatherseals around doors and opening volume of water is required to simply
lights of windows are made of softer observe its flow and drainage from the
rubber compounds, in part, to ensure that cladding/roofing then the spray bar test is
the door or window can be operated more suitable.

2/6
Site testing for watertightness TN No: 41

Apparatus for hose testing and method 30 second period for the specified 1.5
of test metre test length is comfortably achieved,
such that there are ten passes during the
A typical set of apparatus for hose testing five minute test.
is shown in Figure 1. Note that no
connections (e.g. quick-clip connections) A specimen that has been successfully
should be placed between the pressure tested in the laboratory can be used to
gauge and the nozzle as these can calibrate the site hose test: a window of
dramatically reduce the water pressure. acceptable workmanship is tested with the
The recommended nozzle is the Monarch hose pipe and the flow rate is gradually
Type B-25, #6.030 nozzle. increased until water leakage occurs. A
slightly reduced flow rate is then used on
site to distinguish between windows of
acceptable workmanship and those of
unacceptable workmanship. However, the
results are only comparative and do not
absolutely guarantee weathertightness.

Spray bar testing

Spray bar testing is covered by BS EN

13051. A spray bar is a long pipe fitted
with holes or nozzles at regular intervals,
to provide a spray of water over the face
Figure 1 Typical apparatus for hose
of a cladding system. A single line of
testing
nozzles should be used, and water
allowed to run down the face of the
cladding system.
Note that the nominal water pressure is
220±20kPa, which gives a water flow of This test is suitable for open-jointed
22±2 litres/minute through the standard systems (e.g. rainscreen cladding and
nozzle, with a cone angle of 30°. The unsealed patent glazing) and opening
standard distance from nozzle to joint is joints, as water is not forced into the joint.
0.3m, and the joint is tested in 1.5m steps. Moreover, the test is useful for assessing
water flow around penetrations through
Wetting of the test area begins at the systems - a penetration may redirect the
lowest horizontal joint, rising progressively run-off flow onto a joint, or perhaps onto a
upwards via the intersecting vertical joints drainage opening.
to the next horizontal joint. If water
leakage occurs but is not easily located,
an additional procedure is recommended: Apparatus for spray bar testing
all joints to be tested are sealed with
masking tape and progressively exposed The basic apparatus for spray bar testing
and tested by working the hose back and in accordance with BS EN 13051 is shown
forth across the stretch of joint for a period in Figure 2. The spray bar consists of a
of five minutes. Once the source(s) of bar with nozzles spaced at not more than
leakage have been found and rectified the 400mm centres and mounted 250mm
area is re-tested. from the face of the cladding system. The
system should be capable of spraying
One parameter which is not defined in the water at a working pressure range of 2 to
test procedure is the rate at which the 3 bar, so that a constant and continuous
hose should be played along the joint - a

3/6
Site testing for watertightness TN No: 41

film of water is applied to the face of the pressurising the cabinet and a spray
cladding system from the spray bar. system (usually a spray bar or grid of
nozzles).
The area of cladding to be tested should
be agreed. Essentially the spray bar test The basis of cabinet testing is to create a
is a test for runoff. Therefore, the position positive pressure difference on the
of the spray bar in relation to open and cladding system, whilst spraying water
protected joints (e.g. under projections) onto the external face. This technique is
should be properly considered. The suited to the testing of doors or windows
nozzles can be directed at a joint, but it is after installation. However, there are
preferable that they are directed at a point several problems associated with this test:
above the joint, so that the run-off flow
runs down over the joint or area under • The first problem is where to fit the
test. cabinet; if the cabinet is mounted on
the external face then it is necessary to
pressurise the cabinet, which will tend
to push the cabinet off the cladding
system. External fitting is also costly
when many storeys above ground level
because of the reliance on access
equipment. If mounted on the internal
face the cabinet must be de-
pressurised, which will help to hold it in
place, but this will limit access to the
surface of the cladding system to look
for signs of leakage.
• The next problem is how to prevent
lateral air movement through the wall -
pressurising or de-pressurising a
Figure 2 Apparatus for spray bar testing section of wall may draw air in from the
sides, rather than directly through the
The flow rate from the spay bar should be part of the wall under investigation.
adjusted to 5 litres/minute/metre length of • Finally, the finished surface of a wall
spray bar. As stated above, the spray bar may not permit easy attachment or
should generally be located above the sealing of the test cabinet - many
area of cladding to be tested, and at the cabinets are built in-situ and are
ridge, with water running down one slope, specific to one part of the cladding
when testing roofs. system.
The spray bar should be left running in These limitations make cabinet testing
each test location for 30 minutes and the costly, and it may be easier to remove a
internal surface observed for signs of window or door for testing in a
leakage. conventional test rig. The interface
between a window and a wall is more
suited to hose or spray bar testing.
Cabinet testing

Cabinet testing is based on the procedure Apparatus for cabinet testing

outlined in standards such as BS 5368:
Part 2. The basic apparatus comprises a The operation of the cabinet is generally
cabinet that can be sealed to the cladding based on a Standard such as BS 5368:
system, a means for pressurising or de- Part 2, which defines parameters such as

4/6
Site testing for watertightness TN No: 41

water flow rate and nozzle position and this approach is used the frame should be
spacing. braced against the building structure and
transmit no loads to the wall which should
A cabinet may be constructed of plywood carry only the pressure difference
inside or outside the wall as for a generated by the cabinet.
laboratory test.

A simplified method of cabinet testing is to General issues for site watertightness

seal the test area internally with polythene testing
and to de-pressurise the enclosure with an
industrial vacuum cleaner, measured with The following are general issues, which
a gauge. apply to all of the methods described
above:
If this simplified method is used then the
polythene sheet must not come into
contact with any of the sample area on the Surface cleaning
inside of the wall so that air is drawn
through the joints. However, the results Before testing, the area to be tested
are not comparable with the cabinet test should always be washed with a mild
unless the full net pressure across the wall detergent and then rinsed with clean
is developed and the true deflections of water. This prevents dirt from being
the wall or component occur. forced into the system and clogging
normal drainage paths, and the detergent
also helps to break surface tension (water
can otherwise be prevented from entering
a small opening by surface tension
effects).

Internal finishes

Site testing should always be carried out

before internal finishes are applied -
internal finishes prevent observation of the
internal surface of the cladding, and would
have to be removed anyway if remedial
work is necessary.

Test location

It is very easy with hose and spray bar

testing for the observer on the inside of
the building to be looking at a section of
cladding remote from that where the hose
Figure 3 Simplified test cabinet or spray bar is actually applied. Good co-
ordination is required, and radio
communication between assessor and
Whilst cabinet tests are perceived to observer is essential.
require rigid test cabinets similar to those
used in laboratories it is possible to
construct a cabinet from polythene sheet
and a frame of scaffold tubes or similar. If

5/6
Site testing for watertightness TN No: 41

Summary of site testing procedure References

The following steps are required if the site AAMA Standard 501-94 Methods of test
test is to provide genuine results: for exterior walls, Architectural Aluminium
• Determine which site testing Manufacturers Association, USA.
techniques are to be used and, if they
allow fine tuning, what BS 5368 Methods of testing windows, Part
form/modification to the standard 2, 1980, Watertightness test under static
sequence is to be used; pressure, British Standards Institution.
• ‘Site’ test a laboratory mock-up of the
sample which has already passed all CWCT, 1996, Test methods for curtain
other specified tests; walling, Centre for Window and Cladding
• Armed with a knowledge of the Technology, University of Bath.
expected site performance of the
component or system agree which BS EN 13051 Curtain walling -
parts of the installed system are to be Watertightness - Field test Without Air
tested; Pressure Using A Water Spray Bar, British
Standards Institution.
• Agree a suitable procedure for remedial
action should problems be found; Note
• Consider how water runoff will be A video demonstrating laboratory and site
managed to prevent flooding or hose pipe testing is available from CWCT.
damage to elements not designed to be
wetted;
• Proceed with site testing.

Summary of tests suitable for various

components/systems

• Permanent (fixed) joints - hose test,

spray bar test, cabinet test
• Opening joints - spray bar test, cabinet
test
• Open joints - spray bar test, cabinet
test

© CWCT 2004

University of Bath, Claverton Down, Bath,

BA2 7AY
Tel: 01225 386541; Fax: 01225
386556
Email: cwct@bath.ac.uk;
Website: www.cwct.co.uk

6/6