IGN 4-01-03 31 October 2015 PDF

IGN 4-01-03
WATER INDUSTRY October 2015: Issue 2

(Page 1 of 35)
INFORMATION &
GUIDANCE NOTE
UK Water Industry
GUIDE TO PRESSURE TESTING OF PRESSURE PIPES AND

FITTINGS FOR USE BY PUBLIC WATER SUPPLIERS
pressure pipelines. Note: This IGN is referenced by BS EN
Contents 1610 for testing of sewerage and drainage pressure
pipelines.
1. Introduction 1
2. Scope 1 This Information and Guidance Note (IGN) was originally
3. General considerations (all tests) 2 published in March 2011 and has been extensively adopted
4. Test report (all tests) 6 by the UK Water Industry and its contractors. This revision
5. Test for PE and PVC pipelines (Type II 6 has been prepared in the light of the experience gained from
test) the use of the document. The principal technical change in
6. Test for ductile iron, steel and GRP 10 this second edition of the document is in the manner in
pipelines which the amount of air present in a main under test is
7. Test for renewed mains (10 minute test) 11 calculated and the evaluation of its effect on test results. In
8. Test for service connections 14 addition, the document has been restructured so that the test
9. References 14 methods to be used by contractors undertaking the testing
are detailed in the main body of the IGN and additional
Appendix A Estimated water input volumes 15 background to the application and development of the
Appendix B Analysis to determine initial 20 method is included for reference in Appendices B, C and D.
volumes of air in a pipeline
Appendix C Overview of BS EN 805 tests 25 This document has been prepared on behalf of the Water
Appendix D Supplementary information on 28 UK Standards Board. Information contained in this
pressure testing of UK pipelines specification is given in good faith. Water UK cannot accept
any responsibility for actions taken by others as a result.
1. Introduction 2. Scope
One of the primary justifications for the refurbishment of This IGN details the procedure for the testing of PE and
water distribution, transmission and sewer pressure PVC water and sewerage pressure pipelines laid in the U.K.
pipelines is to replace and renovate pipes that have been The procedure involves raising pressure in a controlled
shown to have unsatisfactory leakage levels. It is paramount manner to the System Test Pressure (STP) and allowing the
that new pipeline installations are fit for purpose and pressure to decay after isolating the main. The assessment
therefore follows that the new systems should be follows the principle of specifying an acceptable limit of the
demonstrated to be as secure and leak-free as possible. pressure loss over test time and test length. The values of
allowable pressure loss rate for different materials are
In 2000, BS EN 805 was published and, for the first time, given.
there are now specified European criteria for assessing
acceptability for different pipeline systems. For PE and PVC pipelines, viscoelastic stress relaxation
effects affect the simple linear decay in pressure since they
BS EN 805 gives advice on different test methods that may will lose pressure without any leakage being present as the
be used to assess pipelines for leakage. These methods are molecular structure relaxes, a phenomenon usually referred
not mandatory; it is left to the Engineer / Client to choose to as ‘creep’. These effects dominate the pressure decay
the most appropriate procedure. until some 24 hr. - 36 hr. after reaching STP. Therefore, to
provide an earlier warning that pipelines are leaking at an
It has been decided to adopt one procedure to test all new unacceptable level, the analysis of results is modified for PE
PE and PVC water and sewerage pressure pipelines laid in and PVC and is aligned with that successfully used for many
the U.K. and another method for ductile iron, steel and GRP years in the 1980s and 90s by the U.K. Water Industry.
 Water UK 2015
Technical queries should be addressed to the Water UK Standards Board c/o The Technical Secretary on e-mail:mikeshepherd@thamesinternet.com.
For further copies please visit http://www.water.org.uk/publications/WIS-IGN.
IGN 4-01-03
October 2015: Issue 2
(Page 2 of 35)
This IGN also details the procedure for the testing of ductile Care needs to be taken:
iron, steel and GRP water and sewerage pressure pipelines
laid in the U.K. The procedure is essentially the same as that  Only competent and trained staff who are aware of the
detailed in BS EN 805 and the methods traditionally used risks should be allowed near to any exposed part of the
on these materials. pipeline when it is under pressure.
 The test area should be cordoned off and a warning
Note: Some specifications advise that PVC pipelines are notice erected when the test is in progress.
tested in the same manner as ductile iron but this is not
recommended.  When a long length of main is under test, staff involved
in the test should be in radio / mobile phone contact at
The test methods described above are only appropriate all times.
where a test section of the pipeline and associated fittings  Air in the pipeline should be minimised through
can be completely isolated. Where new pipelines are swabbing and the correct operation of valves.
already installed as part of a working water system – as, for
example, is the case for most rehabilitation projects – a 10 3.2 Choice of test section / length
minute mains renewal test for use on short test lengths is
provided in Section 7. There is no theoretical limit to the maximum length of main
that can be tested, but there are a number of practical issues
It is good practice to also test joints and connections to the which limit the length. These are:
pipeline to ensure a leak free system. A short, visual test is
provided in Section 8 for this purpose.  The number of joints and fittings on the main.
 The availability of potable water to pre-charge the
This IGN applies only to pipes and fittings which form part
main.
of the system that is owned by the utility. Pipes and fittings
which are part of privately owned pipelines or plumbing  The point of discharge of water after the test.
systems in premises, and which are supplied with water
from the public supply system, come under the scope of The  The differences in elevation on the main to meet the
Water Supply (Water Fittings) Regulations 1999, as System Test Pressure.
amended, in England and Wales, The Water Supply (Water  The ability to identify the source of any leak detected.
Fittings) (Scotland) Byelaws (2014) in Scotland or The
Water Supply (Water Fittings) Regulations (Northern  The time available in which to obtain a valid test result.
Ireland) 2009 in Northern Ireland.
The exception to this is the 10 minute test for renewed
3. General considerations (all tests) mains described in Section 7. This test is only suitable for
short test lengths with a small number of joints. The
3.1 Safety maximum recommended length is 200 m comprising two
jointed 100 m coils.
In all hydraulic testing, there are dangers involved when
high pressures are being employed. All applicable national 3.3 Choice of System Test Pressure (STP)
health and safety regulations should be taken into account.
The method for choosing the System Test Pressure in BS
Specific hazards: EN 805 is that STP should be the lowest of:
 High pressures could be dangerous if there is an  Surge calculated
unexpected pipeline failure.
 Forces on end fittings or thrust blocks during testing are o MDP + 1 bar
high and insecurely anchored ends could lead to the end
caps blowing off.  Surge not calculated
 Air in the pipeline is compressed during testing and can o MDP * 1.5
lead to a massive and sudden release of stored energy. o MDP + 5 bar
MDP is defined by BS EN 805: 2000 (clause 3.1.5) as the

maximum continuous operating pressure of the pipeline
2  Water UK 2015
IGN 4-01-03
(Page 3 of 35)
plus an allowance for surge pressures. The allowance for  Any service connections should not be tapped prior to
surge pressure (where this is not calculated) shall not be less pressure testing, even if tapping tees have been bolted /
than 2 bar. welded to the main.
The value of STP shall apply at the lowest elevation of the  Wherever possible, all joints made to the pipeline are
pipeline and should therefore include the static head due to in open trenches, visible for direct visual inspection.
elevation changes in the pipeline (P0) i.e. Raised Pressure +  The section of pipeline to be tested has been backfilled
P0 together equate to STP. and compacted prior to the test. This prevents any axial
movement or thermal effects due to weather changes.
The test pressure at the highest elevation should be at least
the allowable maximum operating pressure, PMA, (defined  Air valves are located at all high points to facilitate the
in 3.1.1 of BS EN 805: 2000). If this is not possible due to removal of air during filling of the main. Air valves
the elevations involved then the line should be split prior to should not be closed during the test but non-self-sealing
testing. air vents should be closed.
It is recommended that an automatic air valve is included in
Note 1: Some water suppliers prefer to use an STP value of all test sections of PE pipes with wall thicknesses thinner
1.5 * the pressure that designer expects the pipeline to see than SDR21, and for GRP pipes with a ring stiffness of
in service. lower than 5000 N/m2, in order to avoid the risk of collapse
under vacuum.
It is recommended that the minimum STP for PE pipelines
should not be less than 0.7 * PN where PN is a numerical 3.5 Test apparatus
designation related to the mechanical characteristics of the
component of a piping system. 3.5.1 Test fixtures
Note 2: When testing PE pipelines, some water suppliers  A typical setup requires tapped blank end plates, and
may prefer to use an STP value of 1.5 * PN of the lowest hydrants or ferrules at the lowest point of the pipeline
rated component in the system. Raising the pressure to facilitate the filling and pumping of the water and its
significantly above this STP will not affect the test analysis subsequent removal.
but may damage a plastic pipe or fitting if the pressure is
maintained for more than 2 hours (however, the creep  Duckfoot hydrant bends may be used as a temporary
should reduce the pressure in the pipeline significantly measure to allow easier removal of swabs used to purge
within 2 hours). It is strongly recommended that STP is not air from the system.
raised above 1.5 * allowable maximum operating pressure
Note: This setup may be varied for the 10 minute test for
(defined in 3.1.1 of BS EN 805: 2000) of the pipeline or
renewed mains – see Section 7.
1.5 * PN of the lowest rated component in the system.
3.5.2 Pressurisation
3.4 Pipeline testing set-up
The capacity of the pump varies according to the type of test
To carry out a quantifiable assessment of leakage by a
carried out:
pressure decay test, it is essential that:
 Type II test: Where available and practical, a pump
 The section of pipeline to be tested is isolated from the
with the capacity to raise pressure smoothly to System
rest of the pipeline with end load bearing end fittings
Test Pressure (STP) in a time period between 1 min. to
with sealed plates. End fittings should have pressure
30 min. should be used.
ratings of at least 1.5 * STP. For higher test pressures
such fittings will, of necessity, be specialist re-useable  Water Loss / Water Added test: A pump with the
items. capacity to raise and control the System Test Pressure
(STP). If the Water Added method is being used, a
 Any thrust blocks or other anchorages are sufficient to
separate, specialised pump is required. This pump
withstand the forces generated by the pressure test and
should be capable of maintaining the test pressure,
any concrete used has been adequately cured.
measuring the volume of water pumped into the
 Closed valves or ‘squeeze-off’ seals (for PE) are NOT pipeline to achieve this, and should have data logging
used to isolate the test length. capabilities. This pump could also be used with the
Water Loss method to accurately measure the water
3  Water UK 2015
IGN 4-01-03
(Page 4 of 35)
required to bring the pipeline back to STP. Any air vents should be located as close to the top of the
Alternatively, a hand pump could be used for the Water pipeline as possible.
Loss method.
3.5.5 Differences in pipeline elevation
 10 minute test for renewed mains: A pump with the
capacity to raise and control the test pressure for a If the test gauge and data logger are higher than the lowest
period of at least 10 min. point on the pipeline, the STP should be adjusted by
Note: An estimate of the volume of water to pressurise subtracting the additional pressure caused by static head
pipelines of different materials can be calculated using the between the lowest point and the point at which the pressure
guidance in Appendix A. reading is being taken. Without this adjustment, there will
be a significant effect on STP and air content calculation.
3.5.3 Measurement and recording equipment
The gauge pressure after filling the main should only reflect
 Pressure gauge: A digital pressure gauge with a that part of the natural static head lying above the gauge
0.01 bar resolution or better. For rigid or semi-rigid level. The height of the gauge above the lowest point of the
pipes, a calibrated conventional circular pressure main, when added to the gauge pressure, should equate to
gauge, minimum 200 mm diameter. the total static head.
 Flow Meter: A calibrated flow meter, ideally with a Figure 2 shows the correction for static head where the test
resolution of 1 litre or better, to measure the volume of point is above lowest point of main.
water added during the pressure rise phase.
 Data Logger System: a pressure transducer with an Example calculation – correction for static head where
accuracy of 0.25% of full scale connected to a logger test point is above the lowest point of main
that can record data at fixed time intervals or at fixed
pressure decay intervals. Time intervals of 20 sec. are Pressure gauge reading = 1.2 bar. Vertical distance
normally suitable unless the test is unusually long or between pressure gauge and lowest point = 7 m. STP =
short. Pressure decay intervals of a minimum of 0.1 bar 10 bar.
are recommended. It is recommended that the data
logger has GPS capabilities to allow the test location Additional static head pressure = 7/10 = 0.7 bar (where
and pipeline to be easily identified. 10 m static head = 1 bar of pressure)
 Volume container (Water Loss test): A calibrated Corrected STP = 10 bar – 0.7 bar = 9.3bar
container, suitable for the expected volume loss (e.g. a
200 ml measuring cylinder would be suitable for small Corrected gauge pressure (for the purpose of air percentage
volumes). This will allow the draw off to be accurately calculation) = 1.2 bar + 0.7 bar = 1.9 bar
measured.
It is not necessary to make any adjustment to the pressure
3.5.4 Test apparatus layout
gauge readings to establish the raised pressure for the
purpose of calculating the pressurised air content. It is
Figure 1 shows a general layout for testing equipment.
assumed that the test gauge and data logger will be at the
same height and therefore detecting the same pressure.
The data logger and pressure gauge should be mounted at
the point of lowest elevation of the pipeline; any deviation
from this should be noted in the test report (see also 3.5.5).
4  Water UK 2015
IGN 4-01-03
(Page 5 of 35)
Tap for air Calibrated Tap for air

bleeding & Pressure Gauge bleeding &
pressure release pressure release
Pump Datalogger
15 Pressure & Flow
10 20
0
235
18
Flowmeter
Thrust
Blocks
Low High
elevation elevation
end end
Figure 1: General layout for testing equipment
Figure 2: Diagram showing correction for static head where test point is above the lowest point of main
5  Water UK 2015
IGN 4-01-03
(Page 6 of 35)
4 Test Report (all tests) After filling the main and bleeding air from the system, the
valve at the point of highest elevation should remain open
For every test, a formal report containing complete details to ensure there is no residual head at that point.
should be prepared.
Any attempt to shorten the pumping phase by increasing the
The report should contain: initial pressure to local mains pressure (pre-pressurisation)
using an adjacent main is not allowed and will result in the
 The name, company and contact details of the person test being classed as invalid.
carrying out the test.
The following points should be noted:
 All details of the materials, dimensions, length and PN
ratings of the pipeline.  It will be obvious from input volumes and elevations
 The System Test Pressure (STP). that pre-pressurisation has occurred.
 Details of the date and start / end time of the test and its  The time for which the pipeline has been pre-
location, including GPS co-ordinates. pressurisation will distort the effective loading time.
The time correction factor will be in error. The error in
 The differences in elevation. correcting time will result in a larger slope change of
 A description of the type of pressure and flow meter, the pressure decay characteristics.
together with details of their maximum range, precision  There is a greater likelihood of the test failing the
and calibration history. specified criteria.
 The pump volume flow rate at rated speed, or the 5.2 Removal of air
logged data from the flow meter.
 All the recorded measurements of the pressure and As much air as possible should be removed from the test
water flow during the pressure rise phase. section during / after filling with water and before the start
of the test procedure to keep the test time to a minimum. To
 The pressure decay data. assist:
 Details of any analysis carried out in accordance with
 Ensure that all air valves are functioning properly.
the methods described in this IGN.
 It is recommended that a foam swab ahead of the water
column is used to assist the removal of air.
5 Test for new PE and PVC pipelines (Type II test)
 It is recommended that swabbing is carried out as one
5.1 Filling of the main continuous operation and at sufficient velocity to
prevent the swab stalling and consequently getting
Water for the testing of potable water mains should be taken trapped within the pipeline.
from the existing supply.
 All non-self-sealing air vents should be closed prior to
An adjacent main may be used to fill the main, but for PE testing.
and PVC pipelines every effort should be made to ensure There is an upper limit of 4% air in any main for a valid test.
the pressure in the pipeline at the lowest point does not
exceed the static head generated by elevation changes. If the The presence of air in a main will have a number of effects:
pressure is raised above this value then water should be bled
from the main to reduce the pressure prior to the test  Air will markedly increase the pressure rise time and
commencing. period over which pressure decay readings are taken.
The main should be filled from the lower end, with all air  Air will distort the interpretation of pressure decay
valves open and an open valve at the point of highest results.
elevation. Air content greater than 4% complicates the interpretation
of the test data. Whilst every effort should be made to
remove air, it is acknowledged that this will not always be
6  Water UK 2015
IGN 4-01-03
(Page 7 of 35)
possible. In such cases, guidance is given in Appendix B of Table 1 - Minimum test decay times by air content and
this IGN. pipe material
An estimate of the required water volume should be Pipe Material Air Content Suggested t3
calculated prior to pressurisation, with an assumed air (%) time
content of no more than 4%: see A.2 in Appendix A of this PE ≤ 4% 20*tP
IGN. PVC ≤ 4% 5 hr.
(1) tP is time taken to reach STP
Once the air has been removed from the test section, the data (2) When tP is less than 3 minutes, the minimum test
logger should be connected and all information logged. decay time t3 is 1 hr.
5.3 Test procedure The minimum test decay time, t3, is dependent on the
percentage air content and for PE, the time (tP) taken to
5.3.1 Preliminary conditioning of main reach the System Test Pressure at the start of the test. Where
air content exceeds 4%, the pressure decay time would need
For PVC pipelines with socket and spigot joints, the to be extended to counteract the effect of air in slowing the
pressure should be raised to the STP and allowed to settle pressure decay and to accurately determine whether the pipe
for 15 min. is leaking. The guidance in Appendix B should be followed.
There should be no preliminary conditioning of PE As the pressure decays, an analysis may be carried out to
pipelines; any such conditioning will result in the test being check whether there is reason to believe that the main is
classed as invalid and a retest will be required. leaking. This may be done whilst the test is in progress.
5.3.2 Raising pressure It is recommended that the test apparatus attached to the
main is not de-commissioned or the main put into service
The pressure should be raised to the test level (STP) by until there is strong reason to believe that the pipeline meets
pumping in a controlled manner. The pressure changes and the acceptance criteria (see 5.4.4) and is deemed to be free
the added volume of water should be continuously logged. from leakage.
A more accurate estimate of air volume can be calculated 5.4 Data analysis
by comparing the actual water input volume required to
bring the pipeline up to STP (obtained from the flow meter) 5.4.1 General
with the estimated water volume calculated prior to
pressurisation (see 5.2). PE and PVC materials creep under stress and therefore the
analysis of the test data is a little more complicated than for
If the actual input volume is significantly more than the other materials. The calculation steps below can be used
predicted, this suggests that the air volume is also greater to obtain two ‘n’ values which are used to indicate a pass or
than predicted, and the test should either be extended or fail.
abandoned and restarted once the air content has been
reduced. The guidance in Appendix B should be followed. 5.4.2 Correcting for creep during pressurisation time
5.3.3 Pressure decay phase For PE pipes only, a correction to the decay time is needed
to account for the amount of creep occurring during the time
After the test pressure (STP) has been reached, the system spent in raising the pressure. This is achieved by adding
should be isolated, the pump shut off and the pressure decay 0.4 * pressurisation time (tP) to the recorded decay time.
logged for the minimum test decay time, t3, given in The calculated time is the ‘corrected decay time’ (tc).
Table 1.
Corrected Decay Time (tc) = (0.4 * tP) + (Time since pump
shut off t1, t2, t3 etc.) (1)
where:
tP is the time to reach System Test Pressure
7  Water UK 2015
IGN 4-01-03
(Page 8 of 35)
5.4.3 Calculating rates of decay 5.4.4 Pass / Fail Criteria
Two rates of decay should be calculated using the pressure 𝑛2

If ≤ 1.25, then the test is a pass.
𝑛1
change or Raised Pressure (PA) data between the times t1
and t2 (slope n1) and between t2 and t3 (slope n2) as follows:
Note: If the ratio is slightly greater than 1.25, the data may
be analysed graphically and the “n” values obtained by a
[log(𝑃𝐴 𝑎𝑡 𝑡1 )−log(𝑃𝐴 𝑎𝑡 𝑡2 )]
trend line analysis to reduce single point errors. The
𝑛1 = (2) graphical “n” values should then be used to establish if the
[log(𝑡𝑐2 )−log(𝑡𝑐1 )]
test passes. Further guidance on the use of graphical values
can be found in Appendix D of this IGN or a specialist may
be contacted to offer advice.
[log(𝑃𝐴 𝑎𝑡 𝑡2 )−log(𝑃𝐴 𝑎𝑡 𝑡3 )]
𝑛2 = [log(𝑡𝑐3 )−log(𝑡𝑐2 )]
(3)
If pipelines fail to meet the acceptance criteria, the test
should be stopped and the excess water bled carefully from
Where:
the system. A search for potential leaks should be initiated.
PA is Raised Pressure = (Actual Pressure at time t1,
After leaks are found and repaired, the test should be
t2 etc.) – P0
repeated, but only after a time greater than four times the
P0 is static head at lowest point due to elevation
total original test time has elapsed to allow for complete
changes in the pipeline.
creep deformation recovery.
For PE
5.5 Post-test procedure
Times: tc1 = tP+0.4tP, tc2 = 8tP+0.4tP, tc3 = 20tP+0.4tP
Pressure readings: PA at t1 = PA at tP, PA at t2 = PA at 8tP, PA at
When a main has been positively accepted as being free
t3 = PA at 20tP
from leaks, the water should be released slowly from the
pipeline with all valves opened.
For PVC
Times: t1 = 1 hr, t2 = 3 hr, t3 = 5 hr.
The water should be discharged safely to a pre-planned
site.
Figure 3 shows a typical pressure v time graph for PE with
times at which pressure values are analysed to determine the
‘n’ values.
8  Water UK 2015
IGN 4-01-03
(Page 9 of 35)
Figure 3: Response of pressure v time for a PE pipe showing the times at which pressure values are analysed to
determine the “n” values
9  Water UK 2015
IGN 4-01-03
(Page 10 of 35)
6 Test for new ductile iron, steel and GRP pipelines  It is recommended that a foam swab ahead of the water
column is used to assist the removal of air.
6.1 Filling of the main
 It is recommended that swabbing is carried out as one
Water for the testing of potable water mains should be taken continuous operation and at sufficient velocity to
from the existing supply. prevent the swab stalling and consequently getting
trapped within the pipeline.
The main should be charged from the lower end, with all air  All non-self-sealing air vents should be closed prior to
valves open and an open valve at the point of highest testing.
elevation. After filling the main and bleeding air from the
system the valve at the point of highest elevation should 6.3 Test procedure
remain open to ensure there is no residual head at that point.
6.3.1 Raising pressure
Once the main has been charged (prior to pressurisation),
the data logger should be connected and all information The pressure should be raised to the System Test Pressure
logged. (STP) by pumping in a controlled manner.
6.2 Removal of air A more accurate estimate of the air content percentage can
be calculated from the actual water input volume (obtained
Where possible, efforts should be made to remove as much from the flow meter) and other known data, see A.4 in
air from the pipeline as possible to keep the test time to a Appendix A.
minimum. There is an upper limit of 4% air in any main for
a valid test. If the inputted volume indicated an air volume greater than
4%, the test should be abandoned and actions taken (as
Note: For the Volume of Water Lost/Added methods (see detailed in 6.2) to remove the air from the system prior to
6.6.3a and 6.3.3b), it is not necessary to remove the air from retest.
the system as this will not affect the results. It is advisable,
however, to remove the majority of the air to minimise 6.3.2 Preliminary conditioning of main
health and safety risks.
For ductile iron or steel pipes with epoxy linings or GRP
For the Pressure Decay method, air content greater than 4% pipes with socket and spigot joints, the pressure should be
complicates the interpretation of the test data. Whilst every raised to the STP and allowed to settle for 15 min. For
effort should be made to remove air, it is acknowledged that ductile iron or steel pipes with cement linings, the main
this will not always be possible. In such cases, guidance is should be allowed to settle overnight.
given in Appendix B of this IGN.
6.3.3 Test phase
An estimate of the required water volume should be
calculated prior to pressurisation, with an assumed air After the preliminary conditioning, the pressure should be
content of no more than 4%: see A.2 in Appendix A. raised to the STP. Once the STP has been achieved, use
either of the following procedures for measuring the amount
The air volume can then be estimated by comparing the of water required to maintain pressure.
actual water input volume required to bring the pipeline up
to STP (obtained from the flow meter) with this estimated a) Volume of Water Added
water volume. If the actual input volume is significantly
more than predicted, this suggests that the air volume is also Maintain STP for a period of one hour by additional
greater than predicted, and the test should be abandoned and pumping as necessary, accurately measure the volume of
restarted once the air content has been reduced. water added and record with a resolution of 5 ml or better.
Attempts should be made to purge air from the main b) Volume of Water Lost
during/after filling with water and before the start of the
pressure test: Maintain STP for a period of one hour by additional
pumping as necessary. Isolate the main by disconnecting the
 Ensure that all air valves are functioning properly. pump and closing all valves. Allow pressure to decay for a
period of one hour. The pressure should then be raised and
10  Water UK 2015
IGN 4-01-03
(Page 11 of 35)
returned in a controlled manner to STP. Water is then 6.5 Post-test procedure

drained back to the decay level pressure and captured in to
a calibrated volume container. When a main has been positively accepted as being free
from leaks, the water should be released slowly from the
6.4 Data analysis pipeline with all valves opened.
If the drawn off or added volume is less than the allowable The water should be discharged safely to a pre-planned site.
volume given in Table 2, then the test is a pass, if it is above
the test is a fail. If the pressure decay rate is close to the 7 Test for renewed mains (10 minute test)
allowed rate, the test may be repeated immediately.
7.1 Introduction
Table 2 – Standard allowable leak rates
(litres/km/hour) as a function of diameter The rehabilitation of mains is often conducted under severe
time constraints to ensure disruption to customers’ supplies
Nominal Leakage is kept to a minimum.
Pipe Rate
Diameter The test for renewed mains does not provide the same level
(mm) (litres/km/h) of robustness (especially with regards to data interpretation
100 0.18 and therefore identification of a small leak) as the longer
150 0.41 Type II or Water Loss test but can provide an element of
confidence in the system being installed, especially if there
200 0.72
are few joints (e.g. the testing of two 100 m coils with an
250 1.13 electrofusion coupler joining them). It can identify leaks at
300 1.62 joints or, if a straight coil is being tested, it can identify
350 2.21 damage such as that caused during slip lining.
400 2.88
450 3.65 This test is a constant pressure test rather than a constant
500 4.50 volume test. Modifications can be made to the testing setup
as follows:
600 6.48
700 8.82
 The test may be conducted against valves and ‘squeeze
800 11.52 off’ seals.
900 14.58
1000 18.00  There is no necessity to remove the air from the system
as this will not affect the result. It is advisable,
however, to remove the majority of the air to minimise
Note 1: Further information on the derivation of the values
health and safety risks.
in Table 2 is provided in Appendix C of this IGN.
Note: ‘Squeeze off’ units are not specifically designed for
If pipelines fail to meet the acceptance criteria, the test use in pressure testing. Care should be taken when
should be stopped and the excess water bled carefully from undertaking the test.
the system until only static head remains. A search for
potential leaks should be initiated. After leaks are found and 7.2 Filling of the test section
repaired, the test should be repeated.
Water for testing of potable water mains should be taken
Note 2: These are small volumes and as such the pressure from the existing supply.
gauge used to monitor the STP and the decayed pressure
needs to have a resolution of 0.01 bar or less. Small errors An adjacent main may be used to charge and initially
in the pressure may lead to relatively large differences in the pressure the test length.
drawn off volume and therefore lead to secure pipelines
apparently failing the test. Once the main has been charged (prior to pressurisation),
the data logger should be connected and all information
Where the calculated volume loss is very small, i.e. less than logged.
0.025 litres (25 ml), a pressure loss test of 0.2 bar/hr. may
be acceptable over the one hour test period.
11  Water UK 2015
IGN 4-01-03
(Page 12 of 35)
7.3 Test procedure been identified by the contractor when they conducted a
visual inspection of the test length. Although not ideal, the
The pressure should be raised to the STP by pumping in a use of ‘squeeze offs’ and testing against valves is needed to
controlled manner. For this test, the STP should be 1.5 * PN ensure this test is quick to conduct.
of the lowest rated component in the length under test, up to
a maximum of 15 bar. Once the STP has been reached, the A successful test is one where the pressure does not
pump should be left running to maintain the STP for a fluctuate significantly and there are low, reducing flow
minimum of 10 min. A visual inspection of all possible rates. If the volume inputted over three equal periods in the
sources of leakage (e.g. a ‘squeeze off’, valve or joint) duration of the test is analysed, there should be steady
should be made by walking along the length of the pipeline volume or reduction in the volume inputted to maintain the
under test. If these possible sources are present, a note pressure. Allowable volumes cannot be given as there is no
should be made on the results sheet. control of air volume and pre-pressurisation is allowed.
7.4 Data analysis An unsuccessful test is one where the pressure fluctuates
significantly and a significant amount of water, which does
The data is analysed by inspecting the graphs of pressure not reduce, is required to maintain pressure. An example
against time and flow against time. A typical graph is shown graph is shown in Figure 5.
in Figure 4.
Once a contractor has become familiar with the test it should
A successful test will have a high flow as the main is being be possible to identify if the test will be a pass or fail by
pressurised; this will reduce significantly when the STP is monitoring the flow meter – both the magnitude (depending
reached. To keep the pressure constant, more water will on the diameter and length) and how the flow rate varies
need to be added as PE creeps. The amount of water needed with time.
should reduce with time and, depending on the diameter and
length of main being tested, should be small in magnitude. If the test indicates there is a leak, this should be identified
A slightly increased level of flow can be attributed to a and the test repeated until a satisfactory result is obtained.
slight leak on a squeeze off or valve but this should have
Mains Renewal (10 Minute Test)
12.00
14.00
10.00 12.00
10.00
8.00
Flow (litres/min)
Pressure (bar)
8.00
6.00
6.00
4.00
4.00
2.00
2.00
0.00 0.00
00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00
Time (mins)
Figure 4: Typical graph for acceptable renewed mains test showing pressure increasing with high flows, reducing as the
test pressure is reached and decreasing as the rate of creep slows.
12  Water UK 2015
IGN 4-01-03
(Page 13 of 35)
Figure 5: Example graph for unacceptable renewed mains test indicating a leak - pressure fluctuates from the test
pressure, flow rate is high and does not reduce over the duration of the test.
Figure 6: Joints which could be assessed using the test for service connections
13  Water UK 2015
IGN 4-01-03
(Page 14 of 35)
8 Test for service connections 9 References
8.1 Introduction BS EN 805: Water supply. Requirements for systems and

components outside buildings.
Service connections are often overlooked and are a potential
source of leakage. To ensure a leak free system, all joints BS EN 1610: Construction and testing of drains and sewers.
and connections should be tested. Figure 6 shows three
joints which could be assessed using this service connection BS EN 545: Ductile iron pipes, fittings, accessories and
test. their joints for water pipelines. Requirements and test
methods.
As there can be a large number of services on a main, this
test has been designed to be quick and easy to conduct. It is IGN 4-32-18: The choice of pressure ratings for
essentially a visual test and there is no data logging to polyethylene pipe systems for water supply and sewerage
support the reported result. Neither the equipment detailed duties.
in 3.5 nor the reporting in Section 4 is required for this test.
8.2 Equipment
A hand pump capable of raising and maintaining a pressure

in the test length at 18 bar.
8.3 Test procedure and reporting
8.3.1 Choice of System Test Pressure (STP)
The test pressure is 18 bar.
Note: This value is specific to SDR 11 PE80 service pipes

with a maximum operating pressure of 12.5 bar. This may
seem high but the service will not suffer any damage from
this pressure if it is only applied for a short period of time.
8.3.2 Preparation and procedure
The service should not be tapped prior to this test being

conducted.
Fill the section to be tested. Raise the pressure to 18 bar and

maintain the pressure for 2 min.
Inspect the service visually during the test. Release the

pressure.
If a leak has been identified then the service should be

replaced. If no leak has been identified then the test is a pass.
8.3.3 Reporting
The observations should be recorded together with:
 The start time and date of the test.

 All details of the location of the service pipe.
 Details of the contractor who conducted the test.
14  Water UK 2015
IGN 4-01-03
(Page 15 of 35)
APPENDIX A: ESTIMATED WATER INPUT An increase in the temperature of the pipe material by
VOLUMES FOR DIFFERENT PIPE MATERIALS 1°C would cause the material modulus to reduce
AND DIAMETERS slightly and, if the pipeline is free to change in length,
would cause the pipe to expand. This leads to a slight
A.1 Introduction increase in the pipeline dimensions, so to compensate
for this change there would need to be an adjustment in
To assist contractors in the choice of pumps to raise the pressurisation water volumes. Similarly a decrease
pressure in mains, the estimated water input volumes have in temperature of the pipe material by 1°C would cause
been calculated for different pipe materials. the material modulus to increase slightly.
Tables A1 to A5 (column 3) show the volume of water Volumes should be decreased by approximately 1.35%
required to raise the pressure to STP with no air present in per degree for temperatures below 10°C and increased
the main. by approximately 1.35% per degree for higher
temperatures.
The calculations assume 100 m length of main and give the
estimated volume inputs for a 10 bar gauge pressure rise.  Values for all materials will be affected by the volume
Values for other lengths/pressures should be scaled pro rata of air in the pipeline.
in accordance with equation (A1) below.
A.2 Estimated water volumes for different pipe
The volume of water required to raise pressure to STP with materials and diameters
no air:
NOTES TO TABLE A1 TO TABLE A5:
Ex = estimated volume to add for pipe pressurisation (from  10 bar has been selected as the reference pressure for
column 3 of Tables A1 – A5) * PA/10 * L/100 (litres) (A1) tables A1 to A3 as the error margin is small, around
±5%, for a range of test pressures from 6 bar to 24 bar.
Where: The use of 10 bar as the reference pressure also renders
the pro-rata calculations for other air content and/or
PA is raised pressure (bar) = (Actual pressure at time t0 pipe SDR values simpler.
– Static Head P0)  10°C has been selected as the reference temperature as
L is tested pipeline section length (m) this is a typical value for buried water supply pipelines.
Variation from this figure is not significant except in
If air is present, a larger volume of water will be required to the cases of above ground pipelines or wastewater
compress the entrained air and scaled pro rata in pipelines carrying high temperature effluent. In these
accordance with equation (A2). instances, the pressure rating (PN) of the pipeline
should be de-rated in accordance with the guidance set
Water required to raise pressure to STP (litres) with v% air: out in IGN 4-32-18 which may reduce the allowable
system test pressure.
Vw = E x + A v (A2)  Provided the pipe wall temperature remains reasonably
constant, temperature would have no effect on the test
Where: results. For above ground pipelines where the pipe wall
Ex is given by equation (A1) temperature may vary during the course of the test
Av is estimated volume to compress initial air content period, there would be an effect on the test results and
of v% (from columns 4, 5, 6, 7 of Tables A1 – A5 this may result in an apparent test failure. It is
as appropriate) * L/100 recommended that a visual inspection of the full length
L is tested pipeline section length (m) of pipeline under test is carried out in such cases.
Values of unpressurised pipe volumes are also given in the NOTE TO TABLE A4:
tables (column 2). The tables given here are for guidance  The data assumes Class 40 cement lined pipe to BS EN
only: 545: 2010 is used. The actual pipe volumes vary with
manufacturing tolerances as these can increase or
 Precise values will alter for different types / grades of decrease the actual internal diameter. Other pressure
PE and PVC. classes are available and would also have different
water volumes.
 Values for all plastics will be affected by temperature
as plastics become stiffer as temperature is reduced.
15  Water UK 2015
IGN 4-01-03
(Page 16 of 35)
Table A1: PE100 pipes (SDR 11), 10 bar pressure increment for 100m of pipe at 10°C
Diameter DN / Unpressurised Volume to add Volume to add for air compression, no pipe pressurisation (litres)
OD (mm) pipe volume for pipe 1% air 2% air 3% air 4% air
(litres) pressurisation
no air (litres)
63 207 2 2 4 6 8
75 296 3 3 5 8 11
90 425 5 4 8 12 15
110 636 7 6 12 17 23
125 820 9 7 15 22 30
140 1031 11 9 19 28 37
160 1344 15 12 24 37 49
180 1702 19 15 31 46 62
200 2102 23 19 38 57 76
225 2659 29 24 48 72 97
250 3288 36 30 60 90 119
280 4126 45 37 75 112 150
315 5220 57 47 95 142 190
355 6633 73 60 120 181 241
400 8419 92 76 153 229 306
450 10648 116 97 193 290 387
500 13151 144 119 239 358 478
560 16504 181 150 300 450 599
630 20879 229 290 379 569 758
710 26512 290 241 481 722 963
800 33675 369 306 612 917 1223
900 42614 467 387 774 1161 1548
1000 52604 576 478 955 1433 1911
16  Water UK 2015
IGN 4-01-03
(Page 17 of 35)
Table A2: PE 100 pipes (SDR 17), 10 bar pressure increment for 100m of pipe at 10°C
no air (litres)
63 241 4 2 4 7 9
75 342 6 3 6 9 12
90 493 9 4 9 13 18
110 736 13 7 13 20 27
125 954 17 9 17 26 35
140 1196 22 11 22 33 43
160 1561 28 14 28 43 57
180 1976 36 18 36 54 72
200 2438 44 22 44 66 89
225 3085 55 28 56 84 112
250 3815 69 35 69 104 139
280 4784 86 43 87 130 174
315 6052 109 55 110 165 220
355 7685 138 70 140 209 279
400 9765 176 89 177 266 355
450 12354 223 112 224 337 449
500 15247 275 138 277 415 554
560 19136 345 174 347 521 695
630 24210 436 220 440 659 879
710 30758 555 279 559 838 1117
800 39058 706 355 709 1064 1419
900 49440 894 449 898 1347 1796
1000 61015 1101 554 1108 1662 2216
1200 87881 1588 798 1596 2394 3192
17  Water UK 2015
IGN 4-01-03
(Page 18 of 35)
Table A3: PE 100 (SDR 21), 10 bar pressure increment for 100m of pipe at 10°C
no air (litres)
63 255 6 2 5 7 9
75 361 8 3 7 10 13
90 520 12 5 9 14 19
110 776 18 7 14 21 28
125 1003 23 9 18 27 36
140 1259 29 11 23 34 46
160 1642 37 15 30 45 60
180 2082 48 19 38 57 76
200 2567 59 23 47 70 93
225 3249 74 30 59 89 118
250 4019 93 36 73 109 146
280 5035 116 46 91 137 183
315 6379 147 58 116 174 232
355 8103 187 74 147 221 294
400 10281 236 93 187 280 373
450 13010 299 118 236 354 473
500 16060 369 146 292 437 583
560 20157 464 183 366 549 732
630 25518 589 232 463 695 927
710 32391 745 294 588 882 1176
800 41146 949 374 747 1121 1494
900 52066 1200 473 946 1418 1891
1000 64269 1480 584 1167 1751 2334
1200 92561 2133 840 1681 2521 3362
Table A4: Ductile iron, 10 bar pressure increment for 100m of cement mortar lined Class 40 D.I. pipe to BS EN 545: 2010
OD (mm) pipe volume for pipe
(litres) pressurisation 1% air
no air (litres)
100 850 0 8
125 1327 1 12
150 1911 1 17
200 3391 2 31
250 5236 3 48
300 7489 5 68
350 10032 7 91
400 13010 10 118
450 16360 13 149
500 20189 16 184
600 28957 24 263
700 39058 34 355
800 51048 45 464
900 64525 59 587
1000 79548 73 723
18  Water UK 2015
IGN 4-01-03
(Page 19 of 35)
Table A5: PVC (SDR 26, PVC-A PN12.5 / PVC-U PN8), 10 bar pressure increment for 100m of pipe at 10°C
no air (litres)
90 542 7 5 10 15 20
110 810 11 7 14 22 29
160 1713 22 16 32 47 63
200 2677 35 24 49 73 97
250 4183 55 38 76 114 152
315 6640 87 60 121 181 242
400 10707 140 98 195 292 390
450 13552 178 123 246 369 493
500 16730 219 153 305 457 609
630 26561 348 242 483 725 966
19  Water UK 2015
IGN 4-01-03
(Page 20 of 35)
A.3 Example - calculation showing volume of water to APPENDIX B: ANALYSIS TO DETERMINE

raise pressure to STP with air content above zero INITIAL VOLUMES OF AIR IN A PIPELINE
Pipe specification: Diameter 180mm; Material PE100 B.1 Introduction

SDR17; length 250m; STP 15bar; static head 1bar.
Estimate based on maximum 4% air content. The purpose of this appendix is to provide additional
background information on the effect of entrained air on the
PA = STP – Static Head = 15 bar – 1 bar = 14 bar pressure test to support the guidance given in Sections 5
and 6.
Using equation (A2), water required Vw = Ex + Av
B.2 Modelling the effects of different air volumes
Ex = Estimated volume to add for pipe pressurisation (from
column 3 of Table A2) * PA/10 * L/100 (litres) Large volumes of air contained in a pipeline can be
= 36 *(14/10) * (250/100) = 126 litres dangerous if a pipe failure occurs. There will be a massive
sudden release of stored energy.
Av = Estimated volume to compress initial air content of 4%
(from column 7 of Table A2) * L/100 (litres) = 72 * All testing standards acknowledge that contained air will
(250/100) = 180 litres totally confuse ‘added volume’ measurements in constant
pressure tests and will have some effect on pressure decay
Vw = 126 litres + 180 litres = 306 litres characteristics.
Note: The values derived presume that pressure is being Having large initial air volumes will always increase
raised from zero to STP, above any static head due to pressure-rise times:
elevation changes.
 High air content will slow down the rate of
A.4 Pressurised air percentage calculator pressure rise. This is shown graphically in Figure
B1.
It is always advantageous to have less air in the system as
leak detection sensitivity is increased at shorter times. Air  Air may mask signs of leaks since air expansion
has an effect on the pressure decay, so it is essential to have will delay the time before leakage affects the
a quantitative measure of the air contained in a pipeline degree to which the pressure decays.
under test. Thus, having high air content always leads to increased test
times and the possibility of reduced test sensitivity.
It is possible to calculate the volume percentage of air
present from the actual volume of water pumped into a B.3 Effects of different air volumes on pressure decay
pipeline when the pressure is raised to its test pressure for PE pipes
(STP) using equation (A3).
An analysis has been made of the effects of air on pressure
Initial air content (%) = (Vw – Ex) / Av1 (A3) decay data in PE pipe systems. The results (Figure B2) show
how air produces a delay before leakage becomes apparent
Where: via a slope change in the logarithmic decay data.
Vw is actual input volume (litres) The results of modelling different pressure drops below the
Ex is estimated volume to add for pipe pressurisation expected creep power law stress relaxation profile are
(from column 3 of Tables A1 – A5) * PA/10 * shown in Figure B3. The presence of air will tend to mask
L/100 (litres) the pressure decay due to a leak. In pipelines with higher air
PA is raised pressure (bar) contents, it will take longer to detect any given percentage
L is tested pipeline section length (m) drop in pressure caused by a leak. This is shown in Figure
Av1 is required volume to compress 1% initial air B3 where the time to detect two different pressure drops
volume (from column 4 of Tables A1 – A5) below expected rates is shown vs air content. The time is
shown as a multiple of the loading time.
If initial air content is greater than 4%, the guidance in 5.2
should be followed. If the air content cannot be reduced to For a pipeline with no air, the pressure will have dropped
4% or less, the test would need to be either extended or by 2% more than expected at 6 * the pressurisation time.
abandoned. The guidance in Appendix B should be With 4% air, the pressure will have dropped by 2% more
followed.
20  Water UK 2015
IGN 4-01-03
(Page 21 of 35)
than expected at 9 * the pressurisation time. When the 𝑛3

If ≤ 1.25, then the test is a pass.
𝑛1
pressure has decayed by 5% more than expected, the power
law slope ‘n’ will have doubled. It is always advantageous
If the air volume estimate is greater than 8% or if the test
to have less air in the system as leak detection sensitivity is
period cannot be lengthened as set out in Table B1, the test
increased at shorter times.
should be abandoned as such large air volumes would
confuse data analysis. The pressure should be reduced to
Air has an effect on the pressure decay, so it is essential to
zero and efforts made to bleed air from the system. The test
have a quantitative measure of the air contained in a
should be restarted after a minimum period of four times the
pipeline under test. Methods of calculating volume of air are
period that the pipe was under pressure, including the initial
given in B.4 below.
rise time.
B.4 Estimating the volume of air in a pipeline B.4.2 Calculating air volume from pressurisation
characteristics
B.4.1 Calculating air volume percentage from data
The shape of the Raised Pressure (PA) vs Volume of water
The air volume percentage can be calculated from the actual added (Vw) characteristic is a good measure of whether
water input volume (obtained from the flow meter) and significant air is present. When there is negligible air, the
other known data: see A4 of Appendix A. pressure will rise quickly as water is pumped into the main.
Eventually, the pressure rise rate will slow down slightly, as
If the air content so calculated is up to the acceptable limit the plastic pipe becomes less stiff due to time dependent
of 4% of the total pipe volume, the test time should be 20 creep effects. This is shown as curve (1) in Figure B4.
times the pressurisation time tP plus 0.4 * tP.
If the pressure rises slowly in the initial phase and then more
Where the calculated air content is greater than 4% but less rapidly at higher pressures (curve (2) in Figure B4), this is
than 8%, the test may proceed with caution. The pressure a sign that there is significant air present. However, where
decay time needs to be extended to counteract the effect of the pipeline has already been pre-pressurised, the air content
air in slowing the pressure decay and accurately determine will already have been compressed which allows a normal
whether the pipe is leaking. The test time should be pressure rise time, hence masking the effect of the excess
extended as indicated in Table B1 plus 0.4 * tP. air.
Table B1 Extended test period for air content greater It is possible to calculate the volume percentage of air
than 4% present from the actual volume of water pumped into the
pipeline to raise the pressure to STP using equation (A3),
Pipe Material Air Content Suggested t4 see A.4.
(%) time (min.)
PE 4% < 6% 30* tP However, if the pipeline is pre-pressurised, equation (A3)
6% < 8% 45* tP needs to be modified to take into account an air compression
factor (Fc) derived from the raised test pressure and the
PVC 4% < 6% 8 hr. initial pressure.
6% < 8% 10 hr.
(3) tP is time taken to reach STP Initial air content (%) = ((Vw-Ex)*Fc) / (V/100) (B2)
(4) Where the suggested t4 time is less than 60 min., the
t4 time is taken as 60 min. The minimum t3 time of 1 Where:
hour no longer applies when a fourth reading is
being taken at t4. Vw is actual input volume (litres)
Ex is estimated volume to add for pipe pressurisation
By taking this additional pressure reading (PA at t4), a (from column 3 of Tables A1 – A5) * PA/10 *
further n value (n3) can be calculated using the following L/100 (litres)
formula: PA is raised pressure (bar)
L is tested pipeline section length (m)
[log(𝑃𝐴 𝑎𝑡 𝑡3 )−log(𝑃𝐴 𝑎𝑡 𝑡4 )] Fc is air compression factor derived from raised test
𝑛3 = [log(𝑡𝑐4 )−log(𝑡𝑐3 )]
(B1) pressure (PA) and initial pressure (i.e. pre-
pressurisation)
21  Water UK 2015
IGN 4-01-03
(Page 22 of 35)
V is the unpressurised pipe volume = pipe volume Table B2 Air compression factor where pre-
(from column 2 of Tables A1 to A5) * L/100 pressurisation has been applied
(litres)
Raised Air compression factor Fc (for given
B.4.3 The impact of pre-pressurisation on air content Test initial pressure at lowest point above
Pressure static head)
Any pressure in a pipeline over and above the static head PA (bar)
due to elevation changes just prior to testing has a dramatic 0 bar 1 bar 2 bar
effect on the initial air content.
10 1.10 2.40 3.90
Note: These values have been calculated using Boyle’s
Table B2 shows the air compression factors which would
Law.
be required for calculating initial air content (see A.4,
equation A3) if the pipeline had not been pre-pressurised
and (see B4.2 equation B2) when pre-pressurised to 1 bar
and 2 bar.
Using the example in A3, where there is an initial air

content of 4% and no pre-pressurisation, the initial air
content calculated using equation B2 and table B2 would
be 9% and 15.5% at 1 bar and 2 bar pre-pressurisation
respectively. These higher values are the result of the same
actual input volume (Vw) in each case.
22  Water UK 2015
IGN 4-01-03
(Page 23 of 35)
Figure B1: Pressure rise characteristics for 1 km of 800mm SDR 11 PE 100 Pipe
Figure B2: Effect of air on pressure decay in a PE Main with (a) No Leak and with (b) Leakage at 4 * Allowable Rate
23  Water UK 2015
IGN 4-01-03
(Page 24 of 35)
Figure B3: Effect of air on time (multiple of loading time) to detect a leak at BS EN 805 level
Figure B4: Pressure v Vol for (a) No Air and (b) 5% Air - 1km: 800mm SDR 11 Pipe
24  Water UK 2015
IGN 4-01-03
(Page 25 of 35)
APPENDIX C: BS EN 805 1.2 is an allowance factor (e.g. for air content) during
the main pressure test.
C.1 Introduction
For allowable pressure loss ∆p in bar (0.2 bar) and with
The purpose of this appendix is to describe the range of tests terms ‘D’ and ‘e’ in mm, equation (C1) is modified to:
available in BS EN 805 and the background to the selection
of certain tests by the UK water industry. 1 𝐷
∆𝑉max = 1.2 ∙ 𝑉 ∙ ∆𝑝 ∙ 100 ∙ ( + ) (C2)
𝐸w 𝑒∙𝐸R
C.2 Rate of leakage – general issues

It is noted that this formula has nothing whatsoever to do
with leakage rates per se. The formula is a simple stress –
Historically, the pressure testing of pipe systems in the U.K.
strain relationship, where ∆V /V is the strain and the ∆p *
has been governed by codes of practice and specifications
SDR component is the stress in the pipe wall. It should also
that have followed procedures developed by clients and
be noted that the pass/fail criteria is independent of test
suppliers over a long time period.
pressure, unlike criteria historically used.
For ductile iron (DI) and PVC pipes, there was specified a
The formula allows for calculation of the pipe contraction
maximum rate of volume loss of 0.02 litres per day per mm
(or expansion) if water is lost (or added) because of a
diameter per km per bar pressure applied.
pressure change, ∆p. ‘Allowable water loss’ is that water
loss which causes a drop in pressure of ∆p. The formula
This criterion was a sensible model of a leakage rate in that
works in terms of pipe volumes, so any allowable water loss
it is implicit that the rate of leakage will be directly
rate is automatically proportional to ∆D2. This may/may not
proportional to the pipe diameter and the Raised Pressure
be the real case.
(PA).
In practice, many leaks around flanges or socket and spigot
For PE systems where creep relaxation effects can make the
joints are likely to vary as a linear function of diameter.
discernment of leakage problematical, the procedure was to
They are controlled by changes in the pipe circumference
compare and contrast the actual decay in pressure with that
rather than the pipe volume. In other cases (such as with
which was to be expected because of stress relaxation due
leaks at tapping tee connections), there may be no pipe
to creep.
diameter sensitivity since the hole for the tapping does not
vary significantly.
BS EN 805 (issued in 2000) specifies that a pipe under
pressure should not suffer a pressure loss of greater than 0.2
For materials which do not have stiffness properties that
bar/hr. due to water loss. BS EN 805 requires that this
vary with time and temperature (e.g. ductile iron, steel), it
pressure decay be translated into an acceptable water
is simple to calculate the allowable water loss using known
volume loss by calculating the change in volume caused by
material properties. For plastics materials, which are
the pressure drop, using the pipe stiffness to calculate the
inherently more flexible and have modulus (stiffness)
diameter change.
properties that change very significantly with temperature
and time under load, there are major problems.
The standard gives a criterion for acceptance which is
described in terms of an allowable pressure loss per hour.
The lack of commonly agreed values of material modulus
values for plastics makes it impossible to define universally
The formula in BS EN 805 is:
agreeable water loss rates. There is also a strong
1 𝐷 implication, whether intended or not, that the lower
∆𝑉max = 1.2 ∙ 𝑉 ∙ ∆𝑝 ∙ ( + ) (C1) modulus materials (e.g. plastics) have a higher “allowable
𝐸w 𝑒∙𝐸R
water loss”. Obviously, this cannot be tolerated.
∆𝑉max is the allowable water loss in litres;
The U.K. water industry wishes to standardize on a common
𝑉 is the volume of the tested pipeline section in litres;
rate of water volume loss as the assessment criterion.
∆𝑝 is the allowable pressure loss as stated in Therefore, the water volume changes caused by the
kilopascals (20kPa); 0.2 bar/hr. decay rate stated in BS EN 805 have been
𝐸w is the bulk modulus of water in kilopascals; adopted as standard, using ductile iron pipe as the yardstick.
𝐷 is the internal pipe diameter in metres;
𝑒 is the wall thickness of the pipe in metres;
𝐸R is the modulus of elasticity of the pipe wall in the
circumferential direction in kilopascals;
25  Water UK 2015
IGN 4-01-03
(Page 26 of 35)
C.3 BS EN 805 – water loss calculation Having accepted that the BS EN 805 allowable pressure loss
rate of 0.2 bar/hr. shall apply for ductile iron pipes, the
In BS EN 805, the allowable pressure loss is 0.2 bar in any resulting acceptable pressure loss rates for pipes made in
test where the main is raised to the test pressure and then different stiffness categories for all other materials have
sealed. This drop in pressure is specified to be common to then been calculated.
ductile iron, steel, concrete and ‘plastic’ pipes.
The criterion is that the acceptable water losses are those
There is a major problem in adopting this universal criterion given in Figure C1. Equation (C1) has then been used to
for both metal and plastic pipes. Because plastics are much calculate ∆p, given ∆V as an input for pipes of different
less stiff the volume change caused by a decrease in SDRs and with different modulus values Ep.
pressure of 0.2 bar will be some 10 - 20 times larger than
for ductile iron where the volume change is small. The values of acceptable pressure decay rates (APDR) are
given in Table D2 in Appendix D of this IGN.
For example, consider the volume of water loss that would Allowable Water Loss for 0.24 bar/hr/km Pressure Drop
cause a pressure drop of 0.2 bar in 1 km of a 250mm 20
Water Loss litres/hr/km

pipeline:
15
 For a rigid / semi-rigid pipe such as ductile iron,

10
1.13 litres would need to be lost.
 For a flexible pipe such as PE100, some 16 litres 5
would be lost.
0
The specification of one simple pressure loss rate in BS EN 0 100 200 300 400 500 600 700 800 900 1000
805 leads to the acceptance of totally different water volume
Outside Diameter (mm)
losses for pipes of different wall thicknesses and made from
different materials. Figure C1 Allowable Leak Rates for ductile iron with a
Pressure Loss of 0.2 bar/hr.
Since water engineers are driven to maintain leakage at very
low uniform rates – independent of the pipe material, it is C.5 Pressure test methods
recommended here that the water loss rates calculated from
equation (C1) of this IGN for BS EN 545: 2010 Class 40 C.5.1 BS EN 805 tests
ductile iron pipes be taken as the acceptable measure of
water loss. These rates should then apply for all materials, BS EN 805 does not directly specify any given test method
based on nominal pipe diameters. for pipes. In Section 11 and in Appendices A26 and A27 of
BS EN 805: 2000, the details for mains testing and data
The allowable water loss rates for BS EN 545: 2010 Class analysis are given. It is always left to the discretion of the
40 Ductile Iron are shown in Table 2 of Section 6 and Figure ‘designer’ to choose the test method, and for PE pipes the
C1. However, it should be noted that actual pipe volumes method described is a recommendation rather than a
will vary with manufacturing tolerances, as these can requirement. The tests described are described in C.5.2 to
increase or decrease the actual internal diameter. Other C.5.4 below.
pressure classes are available and will also have different
water volumes. C.5.2 Limited Pressure Decay - Pressure Loss
Method (BS EN 805 11.3.3.4.3)
For an exact calculation, the following equation may be
used where D is the nominal diameter in mm Following raising of pressure to STP, the main is closed and
it is stated that any changes in pressure are monitored over
Allowable water loss (litre/hr/km) a one hour period (or longer if specified by the designer).
= 0.000018*(ID)2 (C3)
BS EN 805 states that the rate at which the pressure is lost
C.4 Acceptable pressure loss rates should not exceed the specified limit of 0.2 bar/hr. for
ductile iron, steel or ‘plastic’ pipes.
A value of pressure loss may be an easier parameter to
measure in any test, so it is necessary to calculate acceptable This test is appropriate for the assessment of ductile iron
loss rates for different pipe materials and stiffness. and steel pipes (see 6.3 of this IGN) but the simple analysis
recommended is inappropriate for plastic pipes which suffer
from creep stress relaxation.
26  Water UK 2015
IGN 4-01-03
(Page 27 of 35)
It is likely that a result which can be accepted with C.6 Use of BS EN 805 Test Methods for Plastic Pipes
confidence will only be obtained after at least 10 hr. of
pressure decay for PVC and some 36 hr. for PE pipes. It is The wording in BS EN 805 is very confusing with respect
only at these times that the pressure drop from creep stress to different materials under test. The standard does not
relaxation has decayed to the same level as the ‘allowable mention PVC or GRP pipes specifically at all, but
water loss’. Further information is given in Appendix D of recommends that “plastics pipes” be treated the same as
this IGN. ductile iron.
C.5.3 Limited Pressure Decay – Water Loss Method There is only mention of PE and PP pipes as being special
(BS EN 805 11.3.3.4.2 a) cases of materials with viscoelastic properties and thereby
in need of different assessment. This is quite wrong. PVC
In this procedure (long used for ductile iron evaluation in will suffer from creep stress relaxation – although not to the
the UK), the pressure is raised to STP, maintained for one same degree as with PE and PP.
hour, and is then allowed to decay over a 1 hr. - 2 hr. period
to a pressure P1. The pressure is then returned to STP. The pressure in PVC pipes will naturally decay without
Immediately, water is drained from the system to return the leakage and the effects are significant for times up to 10
pressure to P1. The volume drained in this final process is hours. Thus pressure decay test data (or water volume
deemed to be the amount lost due to leakage in the decay measurements) on PVC and GRP are difficult to analyse in
phase. There are two fundamental problems with this test the first 10 hours since stress relaxation effects could be
when applied to short lengths of small diameter pipe: confused with leakage.
 The BS EN 805: 2000 allowable water loss rates are so BS EN 805 does recognise that all PE (and PP) pipes will
small for small diameter pipes that if the pressure is not suffer from the problem that the pressure will still decay in
returned to STP and subsequently drained to P1 with a main without any leakage because of stress relaxation.
great precision, then the result will be meaningless. This will totally confuse the assessment of leakage with any
Very small errors in pressure control (e.g. of the order test method relying on a simple pressure decay rate or the
+/- 0.05 bar) cause significant variations in the water measurement of water added or removed.
volumes lost.
A special test is described in Section A27 of Annex A of BS
 If even small volumes of air are present, the water EN 805: 2000. This was long specified by WRc in their
volume measurements will be meaningless as a true MDPE manuals (Type I test) in the 1980s, but was seldom
measure of leakage, if pressures are not controlled with used in the UK for the reasons outlined below.
great precision.
The new demands for a more sensitive measure of leakage BS EN 805 outlines a primitive method of opening a valve
means that water loss assessment is likely to lead to to allow the pressure to drop swiftly after STP is achieved.
confusing results and this test is not recommended. The valve is closed and if the pressure then increases as the
Instead the Pressure Loss Method (C.5.2) is preferred for pipe diameter contracts, this is taken to be a sign that the
short lengths of small diameter pipe. main is secure.
C.5.4 Constant Pressure Test – Water Added Method There are many problems with this test – primarily:
(BS EN 805 11.3.3.4.2 b)
 There is no possible method of relating any of the
In this procedure, it is required to hold the pressure at STP pressure measurements to an absolute leakage rate. It is
and then measure any water needed to be added to hold thus a highly empirical analysis procedure.
pressure. There are two items to note with this method:
 Tests have shown that mains with leaks greater than
those outlined in Table 2 will still show pressure
 The control equipment to hold the pressure absolutely rebound after valve closure.
constant needs to be sophisticated. Again, small
pressure fluctuations can cause large errors in assigning This test is not recommended.
water leakage rates.
 The water losses allowed can be so small that it is
simply not feasible to measure the flow rates of any
water added to maintain pressure. Instead, the
equipment should be capable of measuring accurately
the volume of water added to maintain the STP.
27  Water UK 2015
IGN 4-01-03
(Page 28 of 35)
APPENDIX D SUPPLEMENTARY INFORMATION Table D1: Examples of end forces generated by 15 bar
ON PRESSURE TESTING OF UK PIPELINES pressure on PE pipes to illustrate the level of danger
D.1 Introduction External End Force

Diameter (mm) (tonnes)
The purpose of this appendix is to provide additional 80 0.8
background information on the testing methods and their 180 3.9
applications for all new water and sewerage pressure 250 7.5
pipelines laid in the U.K. 315 11.9
400 19.2
Section D.5 provides the mathematical background to the 500 30
Type 2 test used in Section 5. 610 44.7
720 62.3
D.2 Choice of System Test Pressure (STP) 800 76.9
[Supplementary to 3.3 of this IGN]
1000 120.2
Historically, different Water Companies have used an ad
D.3 Allowable Pressure Drop Rates
hoc range of test pressures. These have usually resulted
[Supplementary to 5.4 and 6.4 of this IGN]
from the whim of individuals. Some engineers have used
the mains working pressure as the basis for the test pressure;
D.3.1 Allowable Pressure Drop Rates for Single
some have used the pipe rating.
Materials
It is now strongly recommended that all companies
The corresponding ‘allowable pressure decay rates’ which
universally follow the guidelines specified in BS EN 805
would cause the allowable volume changes for pipes made
and establish common levels for the System Test Pressure
from PE, GRP and PVC materials are given in Table D2.
(STP). This has the benefit of greatly simplifying data
Values of moduli used are those measured at 23°C.
analysis and increasing the ability to detect leakage which
is sensitive to pressure.
Small pressure changes applied to the less stiff pipes cause
considerable volume changes, so the allowable pressure
D.3 End Loads
drop rates are always lower for the lower stiffness systems.
[Supplementary to 3.4 of this IGN]
For plastics, creep stress relaxation alone will cause higher
Where socket and spigot joints have been used (e.g. for
pressure drop rates than those given in Table D2, for up to
PVC, ductile iron and GRP mains) sufficient thrust blocks
10 hours with PVC and for at least 36 hours with PE pipes.
or other anchorages should be in position and any concrete
This obscures leakage assessment in simple pressure drop
used should have been adequately cured.
tests and so the analysis for plastics needs to differ from
elastic materials such as ductile iron – see Section D.4.
The ends of the pipeline should be securely anchored and
any temporary strutting should be properly designed.
Table D2: Allowable Pressure Decay Rates (APDR) for
different pipe materials arising from the “allowable”
It is the responsibility of the site engineer to ensure that all
leakage given in Table 2 (Section 6)
end fittings can safely withstand the end forces generated
by high test pressures.
Material Allowable Pressure
These forces can be very high and examples of the end loads Loss (bar/hour)
on PE pipes raised to a test pressure of 15 bar are given in Ductile Iron 0.200
Table D1 below. The forces given should be multiplied pro PE 100 SDR 11 0.030
rata for other pressures. PE 100 SDR 17 0.016
PE 100 SDR 21 0.012
It is essential that the end fittings themselves are watertight. PVC SDR 17 0.024
PVC SDR 26 0.023
End fittings or struts installed to resist pressure forces GRP 0.023
should not be removed until all pressure applied during
testing has been removed from the main.
28  Water UK 2015
IGN 4-01-03
(Page 29 of 35)
Note 1: The values in Table D2 have been corrected from D.4.2 Analysis of Pressure Above Static Head
Issue 1 to align with BS EN 805. Where there is a significant static head on the pipeline,
analysis can be affected. This head is always present and
Note 2: The values for allowable pressure loss for additional cannot decay. Therefore, all pressure data should be
SDR have been calculated using equation (C1) and analysed using the pressure applied by pumping above the
maintaining the same value for ∆𝑉max , the allowable water static head.
loss in litres, for all SDRs.
Raised Pressure (PA) = (Actual Pressure at time t1, t2 etc –
D.3.2 Allowable Pressure Drop Rates for Pipelines with Static Head P0) (D2)
Multiple Materials
D.4.3 Data to be Used for Analysis - Filtering Short
Sometimes, pipes of different types may be accommodated Term Decay Pressures
within a system under test. If there are pipelines with
sections of different materials (A and B), a law of mixtures There are frequently odd pressure variations in the time
may be used to obtain the allowable pressure loss rate (in immediately after closing the pipeline via the isolating
bar/hr) for the whole system. The formula for allowable valve, so no data for either PVC or PE should be analysed
pressure drop (∆p) is: until a time period equal to the rise time has passed.
∆p = LfA * (APDR)A + LfB *(APDR)B (D1) With PVC, the initial decay data will not settle to a constant
logarithmic decay rate for at least 1 hour. No data at times
Where LfA and LfB are the respective length fractions of less than one hour should be used for PVC pipes.
materials A and B.
Data may be analysed by a graphical method (see D.4.4)
It would be most unusual to find more than two materials or and also by calculation (see 5.4).
a large multiplicity of SDR ratings in a test length, but if
this is the case, the Law of Mixtures principle expounded in D.4.4 Trendline Analysis
equation D1 may be extended as necessary. However, it is
recommended that this only be done where the ratio of the A graph of the Raised Pressure (PA) value vs the corrected
two different pipe materials is not greater than 3:1. decay time should be plotted on logarithmic axes.
For small lengths of relatively stiff (metal) pipe mixed with For PE pipes, it is expected that for corrected times greater
pipe made from viscoelastic (plastic pipe) material, it is than the rise time, the Raised Pressure decay results should
necessary to check the metal pipe locally for leaks that are lie on a straight line with a slope between -0.07 and -0.09.
likely to be masked by the viscoelastic nature of the plastic
pipe. If pressure decay results are plotted using a spreadsheet
program, power law trendlines can be applied to all the data
D.4 Data Analysis for PVC and PE pipelines in two separate time intervals. Note: the interval times are
[Supplementary to 5.4 of this IGN] approximate – take nearest readings to those recommended.
D.4.1 Correction for creep during pressurization time The time intervals for PE pipes: (t1 to t2) between tL and 8
tL and (t2 to t3) between 8tL and 20 tL, dependent on the
The rate of linear decay of pressure with time caused by volume of air present. See Table 1 in section 5. The
creep stress relaxation will give rates of pressure drop that equivalent pressure readings should be taken at times (t1 to
greatly exceed the BS EN 805 allowed rate for times up to t2) between tP and 8 tP and (t2 to t3) between 8tP and 20 tP,
10 hours for PVC pipes and 36 hours for PE pipes.
Where tP = time taken to achieve test pressure (also known
To check whether there is evidence that the rate of decay is as “rise time”) and tL is the corrected decay time for creep.
predicted to be unacceptably high, it is necessary to analyse
the logarithmic decay of pressure with time and compare For PVC pipe, the intervals are: (t1 to t2) between 1 hr. and
this with known material creep relaxation behaviour. 3 hr. and (t2 to t3) between 3 hr. and 10 hr., dependent on
volume of air present.
Equation (1) in section 5.4 provides a method of calculating
a correction to the decay time which is needed to account The resulting equations will be of the form:
for the time spent in raising the pressure.
PA = (pressure measured at Decay Time (t1, t2, t3 etc. ))n (D3)
29  Water UK 2015
IGN 4-01-03
(Page 30 of 35)
The negative slopes between the two time intervals should The assessment criteria are:
be determined as the power law exponent (n). The absolute
value of n is dependent on numerous factors and cannot be  For both PE and PVC pipes, the value for the longer
used to determine leak-tightness – this can only be done time period should not increase above the short term
through change of slope (see 5.4). However, the values are value by more than 25%.
given below as a guide.
PE pipes: The expected slope (n) for most PE materials is  For PE pipes: if the longer term slope exceeds -0.13,
between -0.07 and -0.09. the rate of pressure decay is unacceptably high.
 For PVC pipes: if the longer term slope exceeds -0.08,
Note 1: Special ‘barrier layer’ PE pipes with polypropylene the rate of pressure decay is unacceptable high.
or aluminium layers, will have lower slopes. Reference
should be made to the pipe supplier for the expected creep It is to be noted that the simple analysis in Section 5.4 relies
relaxation exponents. heavily on single point data and if there is an indication that
the slope has increased by more than 10%, results should be
Note 2: Values of n may occasionally range between -0.06 plotted and a computer trendline determined (as in Section
and -0.1, depending to some degree on compaction and also D.4.4) using all data in the specified time ranges.
the air content. (A slope below -0.06 may be indicative of
excessive pre-pressurisation). If pipelines fail to meet the acceptance criteria, the test
should be stopped and the excess water bled carefully from
PVC pipes: The expected slope (n) for most PVC materials the system. A search for potential leaks should be initiated.
is between -0.03 and -0.07.
After leaks are found and repaired, the test should be
Note 3: When pipelines utilise different materials, equation repeated – but only after a time greater than four times the
(D1), seeD.3.2 above, may be modified to calculate total original test time has elapsed to allow for complete
expected power law pressure decay exponents for lines with creep deformation recovery.
multiple materials (A and B) by replacing APDR with
values of ‘n’. Example – steadily increasing pressure decay
The formula for allowable value of n is: An example of a set of pressure decay data is shown in
graph below for a PE pipe with a leak generating a pressure
n = LfA*(n)A + LfB*(n)B (D4) loss (in excess of that expected for creep alone) which is
just at the BS EN 805 limit given in Table 2.
where LfA and LfB are the respective length fractions of
materials A and B. It is recommended that this only be done PE80 450mm SDR17 500m Pressure Decay: Table 2 Leak
where the ratio of the two different pipe materials is not 10
greater than 3:1. expected slope = -0.075
Applied Pressure PA (bar)
slope n1 = -0.076
D.4.5 Pass/Fail Criteria for PE and PVC Pipes slope n2 = -0.094
slope n3 = -0.12
For either PE or PVC pipes, no great significance should be Ratio n2 /n1 =1.24 - on limit
placed on the absolute value of n within the ranges quoted. Ratio n3/n1 = 1.6 - indicates leak
leak causes 0.015 bar/hr loss
It is important to note that for a secure pipeline, the pressure 8.1546
1 Power (8.1546)
will decay with a constant power law slope and it is any 10 100 Power (7.7836)
1,000 10,000
increase in slope that is important in assessing whether a Corrected Time (min)
main is suffering leakage. Any slope change is directly
proportional to the leakage rate and computer analysis has Figure D1: Example
shown that assuming up to 4% air content, a 25% increase
in slope corresponds to the allowable BS EN 805 leakage D.4.6 Checking Absolute Rate of Pressure Decay for
rate shown in Table 2. PVC and PE Pipes
Small slope changes or occasional decreases in value at As a final confirmation that the rate of pressure decay is in
longer decay times are not an issue of concern. It is where excess of the allowed limits, it is possible to calculate the
there is a consistent increase in slope that leakage is current pressure decay rate at the end of the test period. This
indicated. is only valid if the test has been running for more than 36
30  Water UK 2015
IGN 4-01-03
(Page 31 of 35)
hours since creep stress relaxation effects will confuse  The test gives more accurate results the longer the
results at shorter times. pressure is allowed to decay. However, leaks of the
order specified in Table 2 in Section 6 can first be
The decay rate should be calculated by use of the time detected in relatively short timescales.
interval over which the pressure decayed by more than
0.05 bar in the last phase of the test.  The basic test method may be used for all pipe
materials.
If the pressure decay rate is in excess of 2*APDR (see Table  The test actually measures the rate of pressure decay,
D2), the pipeline has unacceptable leakage. which is the main criterion specified by BS EN 805.
Note: The factor 2 allows for the residual creep stress  The analysis of results for thermoplastics varies
relaxation which will still occur at 36 hours. depending on the material, but in all cases there are
relatively simple procedures that can be performed
D.4.7 Estimating the Rate of Leakage quickly to give answers that can be directly interpreted
as water volume losses, using the principles specified
Finally, when a pipeline has been shown to have an by BS EN 805.
unacceptable pressure decay rate, it is frequently of use to
the installation contractor to know the extent of the actual D.5.2 Modified U.K. pressure decay (Type 2) test for
leakage. Knowing the leak rate may indicate whether, for metal pipes
example, a single joint is leaking or whether there are
multiple leak paths in various places. The analysis for metal pipes is simple:
To estimate the leak rate, it is necessary to know the current  The results of pressure decay with time can be plotted
pressure decay rate in the test. This can be calculated by to derive a rate of pressure loss; or
taking the pressure decay over the last time increment and
dividing by the time interval - let this be ∆P/∆t.  The rate of pressure loss with time can be calculated
after any initial pressure fluctuations have settled.
Note: if the logger is setup to log the time for a defined
With metal pipes, there should be no loss in pressure if there
pressure drop, or if the logger is setup to log at specific time
has been a pre-test to allow for joint settlement and
intervals, the time taken for the pressure to decay by the last
saturation of any cement mortar lining (see 6.3.2) and so
0.05 bar should be used as the last time increment.
any continuous decay is a sign of leakage. The measured
rate of decay can be compared with the BS EN 805 level of
The pressure drop due to leakage can be converted to a
0.2 bar/hr. If it is above this level, the pipeline has an
water volume loss (litres/hr/km) using equation (D8) in
unacceptable leakage level. However, the accuracy of
D.5.4.
pressure drop measurement required is exacting and can be
affected by temperature change.
D.5 Pressure Decay (Type 2) test
D.5.3 Modified U.K. pressure decay (Type 2) test for
D.5.1 Advantages of pressure decay test
PVC and GRP pipes
 The test method itself is extremely simple. The pipeline
With PVC, pressure will decay naturally without leakage
is raised to pressure with a pump and when the test
for at least 10 hours after achievement of STP. Thereafter
pressure (STP) is achieved, the pump is switched off, a
the pressure drop caused by stress relaxation alone becomes
valve is closed and the pressure is monitored.
of less and less significance.
 The equipment needed to raise the pressure is
unsophisticated. Thus, where there are leaks in PVC pipelines, the sign of a
leak can be discerned by plotting a linear graph of the
 The pressure measurements may be made using either Pressure vs Decay Time. Two typical tests on PVC-A pipes,
pressure gauges or transducers. There is no need for one known to be leaking, the other secure, are shown in
highly sophisticated or highly sensitive monitoring Figure D2.
equipment.
 Data may be recorded by simple loggers so that For the case with the leaking pipe, the pressure was
constant monitoring by staff is not necessary. There is approximately 10 times in excess of the allowable rate
also a permanent record for QA purposes. (Table 2 in Section 6) and was detectable after less than 2
31  Water UK 2015
IGN 4-01-03
(Page 32 of 35)
hours on test. This was caused by a single leaking socket

and spigot joint. n = creep stress relaxation exponent
After 4 hours there was a constant rate of pressure drop with  Leakage is a linear decay of pressure with time and thus
time when the creep stress relaxation rate became a small any leak will cause the pressure to deviate continuously
fraction of the rate of loss caused by the leakage. from the log-log power law.
For much smaller leaks, the pressure decay rate caused by  A steady increase in the negative slope of the
the leak becomes dominant in comparison to the creep stress logarithmic plot of PA∆ vs corrected time is a sure sign
relaxation rate after 7 hr. - 10 hr. Thus tests on PVC of leakage.
materials need to be continued for 8 hours to be certain that As an example, Figure D3 shows that on a 7.7 km length of
the pipeline is secure. 500mm PE pipe, the leak generated a pressure decay rate
that was some four times in excess of the allowable water
D.5.4 Modified U.K. pressure decay (Type 2) test for PE loss rate given in Table 2 of Section 6. It was obvious from
pipes the slope changes that there were serious leakage problems
after approximately 12 hours – this was at about twice the
Background pressure rise time.
Historically, the U.K. Water Industry has used a pressure The change in slope of the log-log plot of pressure change
decay test to assess PE pipes for leakage. This is in accord with time is a very sensitive measure of leakage.
with BS EN 805. The test procedure itself is in full accord
with that outlined simply in Clause 11.3.3.4.3 of BS EN Historically, the absolute value of gauge pressure was
805: 2000. required to be plotted as a function of time.
Use of Change in Pressure for Analysis A ‘correction’ of 0.4*time taken to reach STP was required
to be added to all the decay times – to account for the creep
It is now known that the use of an absolute value of pressure deformation that had accumulated during the pressure rise
led to confusion in the rates of creep stress relaxation when phase.
high initial static head existed. The BS EN 805 analysis is
actually calculated in terms of the change in pressure (e.g. Derivation of the Pressure Loss Rate due to Leakage
equation (D5)) and so the new requirement also requires
measurement of a pressure change above the initial static If equation (D5) is differentiated, we have:
level. Examination of some hundreds of historic tests has
shown that this greatly enhances the consistency of data dPA/dt = -n*∆P1 * (Corrected Time) –(1+n) (D6)
expected for creep deformation in tests which showed no
signs of leakage. This rate can be predicted from known creep stress
relaxation data.
The basis of the decay data analysis is as follows:
If there is a leak which causes a pressure change (dPA/dt)leak,
 Pressure will fall continuously with time because of the total rate of negative pressure change will be:
creep stress relaxation.
Measured Rate of Pressure Drop = - (n*∆P1 * Time –(1+n) ) +
 The decay is linear on logarithmic axes, because creep
(dPA/dt)leak, (D7)
stress relaxation follows a power law, i.e.
The leak rate is the only unknown and can thus be derived.
PA = P1* (Corrected Decay Time)-n (D5)
This relationship can be plotted on logarithmic axes (where
Where: the creep relaxation component will follow a linear power
law) and the leak component will be a constant.
PA is the pressure applied above static head, see
equation (D2) in D.4.2 As an example, a PE pipe leaking at the ‘allowable’ rate
would be as shown in Figure D4.
Corrected Decay Time is given by equation (1) in 5.4.
P1 is the value of PA when time = 1
32  Water UK 2015
IGN 4-01-03
(Page 33 of 35)
The test data shown were calculated by simply taking the If the absolute level of leakage is known, it can be of great
incremental changes in pressure ∆P over logged time assistance to the installation contractor in finding and
intervals ∆t as the test continued. repairing the source of the leak.
Note: If transducers are set at too sensitive levels, the D.6 Visual Inspection for Leakage
pressures will oscillate. Therefore, it is recommended that
the minimum pressure change ∆P should be at least 0.05 bar With most pipes, there is seldom any leakage through the
- 0.1 bar to prevent flutter of logged pressures being too pipe wall. All pressure pipes in use in the UK will have been
significant. tested at the factory to much higher pressure levels than
It can be seen that in this example after 30 hours on test (800 STP.
mm diameter, 3.8 km long pipe giving long test time), there
is a significant deviation from the expected linear decay due With PE pipes, 25 years of testing have rarely detected a
to creep stress relaxation and the leak rate falls to a constant failure of a butt fusion weld with systems welded with
value after approximately 72 hours. This is beyond the time modern equipment.
when creep stress relaxation causes any major pressure
change and can thus be considered to be the rate of pressure Leaks will occur generally at mechanical joints or
loss caused by a leak. electrofusion welded pipes/fittings in PE systems.
If this value is in excess of the values given in Table 2 in It is recommended that the pipeline be raised in pressure
Section 6, then the pipeline has failed to pass the required back to STP and that the contractor carry out a visual
criteria. inspection of all joints that are visible.
It is to be noted that sometimes leaks are quite sensitive to If there are no signs of visible water loss, a leak-noise
the pressure level, in which case the final pressure decay correlator may be used and the ground inspected for damp
rate may not settle at a constant value. The rate will still be patches.
well in excess of the ever decreasing creep rate but in this
case it would be necessary to introduce a pressure factor to If there is no success, the pipeline should be divided into
calculate a leakage rate. shorter sections and further pressure tests conducted.
This type of analysis requires that the PE pressure test be

continued for at least 36 hours, to be certain that the pressure
drop caused by the water leakage rate is in excess of the
pressure drop due to creep. However, the early signs that
there is a problem can be discerned when the test data
deviate from the power law line. The analysis of the slope
changes given in D.4.4 give a much more sensitive early
warning that there are unusual pressure changes occurring
and this analysis is preferred as the first stage in test data
interpretation.
The pressure change with time analysis here is to be used

when an absolute value of the pressure change due to
leakage is required.
Conversion of Pressure Decay Rate to Volume Loss
The derived pressure drop due to leakage can be converted

to a water volume loss (litres/hr/km) by the following
calculation involving scaling via equation (C2):
Water loss = (Measured Pressure Decay Rate/ APDR) *

0.000018* (ID)2 (D8)
Values of the Allowable Pressure Decay Rates (APDR) are

given in Table D2.
33  Water UK 2015
IGN 4-01-03
(Page 34 of 35)
Figure D2: Pressure Decay Curves for PVC-A Pipe: with and without leaks
Figure D3: Raised Pressure (PA) v Corrected Time (on logarithmic axes) for a PE Pressure Test
34  Water UK 2015
IGN 4-01-03
(Page 35 of 35)
D4: Rate of Pressure Decay as a function of Time for a PE Pipe (log axes)
35  Water UK 2015

IGN 4-01-03 31 October 2015 PDF

Uploaded by

Copyright:

Available Formats

IGN 4-01-03 31 October 2015 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

IGN 4-01-03 31 October 2015 PDF

Uploaded by

Copyright:

Available Formats

IGN 4-01-03

WATER INDUSTRY October 2015: Issue 2

GUIDE TO PRESSURE TESTING OF PRESSURE PIPES AND

MDP is defined by BS EN 805: 2000 (clause 3.1.5) as the

Tap for air Calibrated Tap for air

Figure 1: General layout for testing equipment

5.4.3 Calculating rates of decay 5.4.4 Pass / Fail Criteria

Two rates of decay should be calculated using the pressure 𝑛2

returned in a controlled manner to STP. Water is then 6.5 Post-test procedure

Mains Renewal (10 Minute Test)

8 Test for service connections 9 References

8.1 Introduction BS EN 805: Water supply. Requirements for systems and

A hand pump capable of raising and maintaining a pressure

8.3 Test procedure and reporting

8.3.1 Choice of System Test Pressure (STP)

The test pressure is 18 bar.

Note: This value is specific to SDR 11 PE80 service pipes

8.3.2 Preparation and procedure

The service should not be tapped prior to this test being

Fill the section to be tested. Raise the pressure to 18 bar and

Inspect the service visually during the test. Release the

If a leak has been identified then the service should be

The observations should be recorded together with:

 The start time and date of the test.

A.3 Example - calculation showing volume of water to APPENDIX B: ANALYSIS TO DETERMINE

Pipe specification: Diameter 180mm; Material PE100 B.1 Introduction

than expected at 9 * the pressurisation time. When the 𝑛3

Using the example in A3, where there is an initial air

C.2 Rate of leakage – general issues

Water Loss litres/hr/km

 For a rigid / semi-rigid pipe such as ductile iron,

D.1 Introduction External End Force

hours on test. This was caused by a single leaking socket

P1 is the value of PA when time = 1

This type of analysis requires that the PE pressure test be

The pressure change with time analysis here is to be used

Conversion of Pressure Decay Rate to Volume Loss

The derived pressure drop due to leakage can be converted

Water loss = (Measured Pressure Decay Rate/ APDR) *

Values of the Allowable Pressure Decay Rates (APDR) are

You might also like