Innovations in Modulated Flow and Self Configuring
Siphonic Roof Drainage Systems
D.P.Campbell (1)
d.p.campbell@hw.ac.uk
Drainage Research Group
School of the Built Environment,
Heriot Watt University,
Edinburgh, Scotland U.K..
Abstract
Siphonic roof drainage (SRD) systems are an efficient method of removing rainwater
rapidly from roofs. Such systems are designed to run full-bore, resulting in subatmospheric system pressures, higher driving heads and higher system flow velocities.
Hence, SRD systems normally require far fewer downpipes, and the depressurised
conditions also mean that much of the collection pipework can be routed at high level,
thus reducing the extent of any underground pipework. But, they work properly at only
one roof run-off rate and therefore suffer from sizing and operational problems which
limit their performance. Climate change is creating situations where normal ranges of
rainfall intensity are being frequently exceeded, and this may have an impact on the
performance of SRD systems.
A multiple parallel pipe SRD system appears to offer benefits and avoids sizing
problems associated with current SRD systems. A movable cap covering the inlet to a
small bank of parallel pipes has the potential to avoid noise associated with making and
breaking siphonic action through flow modulation. Laboratory scale tests demonstrate
the basic feasibility of the multiple parallel pipe system and indicate that handover of
flow between pipes occurs smoothly, and that the flow modulation cap functions
reliably.
Keywords
siphonic roof drainage, SRD, parallel pipe, flow modulation cap, rainfall, climate
change.
1.0 Introduction
Siphonic roof drainage (SRD) systems are an efficient method of removing rainwater
rapidly from roofs, and this is gaining significance as climate change progresses.
Current SRD’s tend to work properly at only one roof run-off rate and therefore suffer
from sizing and operational problems which limit their performance. The aim of this
paper is to describe the development of an innovative development of SRD’s,
including multiple parallel pipes, and a flow modulating cap which also reduces
noise. Laboratory scale tests demonstrate the basic feasibility of the multiple parallel
pipe system and indicate that handover of flow between pipes occurs smoothly, and
that the flow modulation cap functions reliably.
Climate change is imposing direct threats to the built environment, and these threats are
evolving at an unprecedented rate. Within the built environment water cycle, these
changes include increasing rainfall intensities and frequencies1,2 which challenge the
performance of current roof drainage systems3. There are basically two different types
of roof drainage system; conventional and siphonic. Conventional systems operate at
atmospheric pressure, and the driving head is thus limited to the gutter flow depths.
Consequently, conventional roof drainage systems normally require a considerable
number of relatively large diameter vertical downpipes, all of which have to connect
into some form of underground collection network before discharging to the surface
drain.
In contrast, siphonic roof drainage (SRD) systems are designed to run full-bore,
resulting in sub-atmospheric system pressures, higher driving heads and higher system
flow velocities. Hence, siphonic systems normally require far fewer downpipes, and the
depressurised conditions also means that much of the collection pipework can be routed
at high level, thus reducing the extent of any underground pipework. This also leads to
a reduction in the extent of costly underground drainage networks . This is particularly
attractive where contaminated land may be encountered. The flexibility afforded by
depressurised flow can free up space, providing greater layout flexibility[3]. But, they
work properly at only one roof run-off rate and therefore suffer from sizing and
operational problems which limit their performance4. This paper covers an innovative
re-design of the SRD system, which has the potential to avoid these problems while
offering increased responsiveness, flexibility and safety. Increased responsiveness will
mean that the system will operate at several different design capacities over wider
rainfall rates and the increased flexibility will mean that the system will cope with the
future challenges posed by climate change more readily, avoiding flooding and
ponding.
In most developed cities, roofs account for a large proportion of the impermeable urban
surface area. Effective roof drainage helps ensure economic growth by protecting
buildings from rainwater intrusion and directing any excess flow away from the
building perimeter. Rainwater falling on roofs arrives in a range of time-dependant
rates, from “typical” through to “storm”. For each climatic region globally, the ratio
between these rates will have a statistically derived range, defined here as the
typical:storm ratio, or rTS. Tropical locations will have an rTS of approximately 6, while
temperate locations such as the UK have an rTS of approximately 3[4,5]. Since current
SRD’s work at design capacity at only one rainfall rate, a single pipe system can only
be designed to operate in it’s design state within a very narrow rTS band: it is up to the
designer to select this band, or fit overflow or secondary SRD’s, to minimise the risks
of flooding and ponding[6]. Where secondary systems are used, rather than using a
single systems which operates at the design storm intensity, two smaller systems are
used which together can drain the design storm. Commonly, the primary system will be
sized to accommodate the events with a 3 year return period. The secondary system
then caters for the remaining intensity range (say 2-500 year return period). For an
installation in a large UK city, for example, this will mean that the primary and
secondary systems will have a capacity of 22% and 78% of the system’s total capacity
respectively. However, secondary systems essentially double the work involved at both
the design and installation stages, and at less than the storm capacity the single working
system will hunt, creating noise and leaving a increased water depth on the roof during
the event. These two deficiencies demonstrate the need for a revolution on the way
siphonic rainwater drainage systems as the current approach lacks flexibility.
At precipitation rates below the typical design capacity, which are common, a
conventional SRD will hunt between siphonic and non-siphonic flow which may cause
vibration, noise and a reduction in the carrying capacity. This arises as flows may
contain substantial quantities of entrained air and exhibit pulsing or cyclical flow
phases[7]; a result of greatly varying gutter water levels and an indication of truly
unsteady, transient flow conditions. Such problems are exacerbated when the system
incorporates more than one outlet connected to a single downpipe (multi-outlet
system)[8], as the breaking of full-bore conditions at one of the outlets (due to low
gutter depths and air entry) is transmitted throughout the system and, irrespective of the
gutter depths above the remaining outlet(s), results in cessation of fully siphonic
conditions. As this condition is common, it has become clear that current design
methods may not be suitable for assessing the day-to-day performance characteristics
of SRD’s. It is also clear that adding extra capacity to a system to accommodate climate
change as it alters regional rTS values, or to provide a factor-of-safety against
construction errors, will amplify the problem. At rainfall rates above the storm design
capacity a conventional SRD is at risk of flooding. The current practice of adding an
overflow or secondary system is a reaction to this problem, confirming the need for the
innovative approach outlined here.
SRD’s are a popular method of transferring rainwater from roof level to the ground,
with an estimated 10,000 systems now installed in the UK. Their use is particularly
well established for large steel-framed buildings with restricted numbers of internal
columns, such as shopping centres, football stadia, airport terminals and railway
stations[9]. A practical example of these advantages is illustrated by the Millennium
Stadium, the largest project of its type in the UK to date, seating up to 75,000
spectators.
SRD’s were developed in the 1960s in Scandinavia however a significant body of
published research has been carried out at Heriot-Watt University (Edinburgh,
Scotland) in the 1990s and 2000s. 2004 saw the publication of a DTI funded draft
standard for Siphonic Roof Drainage, and the formation of the Siphonic Roof Drainage
Association. The draft standard has since been published by BSI[10]. Together, the
operating characteristics of SRD’s enable significant savings, in terms of time, space
and money. The need for vertical rainwater pipes inside buildings can be eliminated,
saving approximately 0.5m2 per absent down-pipe[11].
1.1 SRD Operation
For the siphonic action to start the system must be airtight, this requires special
rainwater outlets, correctly dimensioned pipes and fully sealed joints[9]. Gutter outlets
are designed to be submerged, to allow full-bore flow conditions to develop and be
sustained. The determination of outlet depth is therefore complicated as conditions are
dependent on both gutter inflows and the downstream pipework[12]. To restrict the entry
of air into the system a baffle plate is typically fitted over the orifice of the outlet, this
prevents the formation of vortices that could draw air into the system.
Priming in SRD’S is the process in which the air in the system is replaced with
rainwater. SRD’S are typically designed on the assumption that the rain water entering
the system is not accompanied by air, consequently any air within the system will result
in a reduction in system capacity[11]. Turbulent gutter flow conditions will invariably
lead to quantities of air being entrained in the rainwater flow, typically up to 10%.
The priming process is complex and key to system performance. Experimental work[13]
has shown that this process follows an identifiable sequence of events and the timing,
interaction and precise form of these events depend to a significant degree on the size
and complexity of the system[13].
The transition from annular flow to full-bore flow in the vertical stack is the least well
understood part of the priming process. Flow velocities must be sufficient to overcome
the buoyancy of the air in the stack during initial priming, and mixed within the inflow
during operation. Experimental work indicates the terminal velocity of a rising air
bubble of equivalent diameter of 0.03 – 0.04 m, is in the range 0.3 – 0.45 m/s[14],
therefore the system flow velocities must exceed this level[13].
In order to prime all air must be purged from the system; it has been observed that this
occurs through the formation of hydraulic jumps. Studies have shown that flow in
SRD’S is initially sub-critical, changing to supercritical as the tail pipes prime[13]
A hydraulic jump is a phenomenon in fluid dynamics which occurs in free surface flow,
when liquid at high velocity discharges into a zone of lower velocity. At this point an
abrupt rise or hydraulic jump occurs in the liquid surface. This occurs as the rapidly
flowing (super-critical) liquid is abruptly slowed, and some of the kinetic energy it
possesses is converted to potential energy which manifests as a hydraulic jump[15]. It is
this increase in water height in the system pipework, which ultimately chokes the free
surface flow creating flow bore flow, and thus depressurisation of the system.
The phenomenon is dependent upon the initial fluid speed; if the initial speed is below
the critical speed then no jump is possible. In fluid dynamics supercritical flow occurs
when the flow velocity is larger than the wave velocity, the analogous condition in gas
dynamics is supersonic. For initial flow speeds which are not significantly above the
critical speed, the transition appears as an undulating wave. If the initial flow speed
increases the transition becomes more abrupt[15]. Flow velocities during priming must
therefore be sufficient to allow the formation of hydraulic jumps.
In fluid dynamics, the change from sub-critical to super-critical flow is described by a
dimensionless quantity called the Froude number (FR):
FR < 1 = subcritical flow
FR > 1 = supercritical flow
Studies have suggested that rather than specifying a minimum flow velocity in SRD’S,
a minimum Froude number (FR) of 1.50 is achieved, as this accounts for pipe diameter.
The Froude number can be calculated as follows:
FR  B.V 2 / g. A
The shallow gradients associated with siphonic roof drainage systems mean that
hydraulic jumps form readily when the inflow is supercritical[13].
An alternative approach to minimum system velocity is described by [16]:
V  6.0 ( g. A / p)
Essentially is if the flow velocity is too low, it will not be able to entrain air into the
flow, or preclude bubbles rising to form air pockets along the pipe soffit[16]. This will
reduce the level of depressurisation and increase the risk of slug flows developing,
which create large pressure fluctuations. These fluctuations then propagate though the
system as pressure transients.
Transient propagation is experienced in fluid-carrying systems when the running
condition is subject to change. This is irrespective of whether the fluid is liquid, gas or
a mixed two phase flow. Within SRD’S transient propagation occurs as an inevitable
consequence of any change in the system operating condition. The magnitude of any
pressure transient depends upon the properties of the fluid, conduit and the rate of
change of the system conditions[17].
Theory suggests that the sudden blockage of a siphonic outlet, by detritus such as
leaves, can lead to the generation of large pressure transients. Also the surcharge of the
sewer can cut off the water path; in this situation flow may be rapidly brought to rest.
In these conditions significant pressure transients are generated [18]. Less extreme
blockage events have been simulated in laboratory studies, during which significant
pressure transients were recorded for partial and relatively slow blockages [19].
1.2 Single-outlet SRD’s
At rainfall intensities less than 40% of the fully primed system capacity, single-outlet
siphonic systems act in a similar manner to conventional roof drainage systems. Above
40% of system capacity, within the mid to high range of rainfall intensities, unsteady
flow conditions occur. When subjected to constant rates of such inflow, conditions
include cyclical variations in gutter water levels and system pressures; these result in
large quantities of air being drawn into the system[16].
Air entering via reduced flow depths causes the severest problems, if a large pocket of
air is drawn into the system, it can cause a sudden local de-pressurisation which then
propagates through the entire system when it reaches the main vertical stack[11]
Due to the high negative pressures that can occur in SRD’S, it is important that a
resilient material is used with high resistance to implosion. In addition the system must
be securely fixed with strong brackets to maintain their integrity through periods of
vibration, which occur during both priming and at inflow rates between 40 and 100% of
system capacity.
The maximum capacity for the system is determined by the available head between the
outlets and the point of discharge. When the system is primed this available head is
fully used by the flow to overcoming the hydraulic resistance of the system elements.
At this point the system is running at maximum capacity and any additional rainfall
will lead to overtopping of the gutters[11].
The use of smaller diameter pipework than in CRDS, and the used of full bore flow
results in very little spare capacity. As such, debris entering the outlets and pipework
may significantly reduce system capacity. This risk of failure is accounted for by
applying a factor of safety to the design. To allow for some limited blockage of the
outlet[10] requires that flows should be increased by at least 10%.
Increasing system capacity by 10% equates to an increase in the rainfall return period
of 20–40%. Whilst allowing additional capacity is prudent, this reduces the frequency
during which the SRD’S is likely to operate in a fully primed state[13]. By increasing the
system capacity the frequency of operation at sub-design conditions is also increased;
this in turn increases the occurrence of unsteady flow conditions, which lead to
problems of noise and structural vibration.
In order to prevent flooding at roof level, siphonic systems need to prime quickly and
attain their design flow rate within the duration of the design storm event should take a
maximum of 60 seconds to prime a system[10]. However, other than for idealised
installations, there is no analytical method available which can determine how long a
system will take to prime, or even if a system will prime[13].
As analytical methods to allow the priming process to be considered in detail during the
design phase, are not available, designers should use their understanding of the priming
process to inform decisions.
Current design practice assumes that, for a specified design storm, a siphonic system
fills and primes rapidly with 100% water. This assumption allows siphonic systems to
be designed utilizing steady state hydraulic theory[16].
1.3 Multi-outlet SRD’s
Frequently SRD’S are installed in a system with more than one outlet connected to a
single downpipe known as a multi-outlet system.
Studies into the performance of multi-outlet SRD’S broadly confirm that priming
characteristics are similar to, though more complex, than the priming of a single outlet
system[16]. The most significant differences occur due to the inter-relationship between
outlets, which create more complex flow conditions. This increases the occurrence of
mixed water-air (two phase) flow in the system[19].
Indeed the problems encountered with SRD’S are exacerbated when the system
incorporates more than one outlet connected to a single downpipe[12].
Flow division between outlets depends on the relative losses associated with each
branch of the system[12]. Local gutter water levels may mean that one or more of the
outlets act as an air inlet to the system. This breaks the full bore conditions, leading to
re-pressurisation, transient propagation throughout the system and cessation of siphonic
action[19].
A key benefit of installing a multi-outlet SRD’S is their ability to redistribute flows
between outlets if one becomes blocked[13]. Studies into systems with two outlets have
shown that when a single outlet is blocked prior to the rainfall event, the system acted
as a single outlet siphonic system, with a lower, fully primed capacity but with
considerably lower system pressures[19].
2.0 Sizing SRD’S
BS8490:2007[10] suggests the return period used to determine the design rainfall
intensity should be based upon the same categories of risk as described in BS EN
12056-3[10] for CRDS. It also suggests that SRD’S are designed to cater for the steadystate flow conditions corresponding to the design rainfall intensity, with no allowance
made for storage effects within either the gutters or the pipework.
This approach neglects the small quantities of entrained air that enter a siphonic roof
drainage system due to turbulent gutter conditions[20]. It has however been reported to
yield characteristics similar to those observed in laboratory test rigs[12]. However, one
European installer is known make allowance for two phase flow by using reduced flow
density[16].
Bernoulli’s energy equation is used to determine the change in flow conditions between
any two points in the system:
2
2

u  
u 
 h1  1    h2  2   h12  z12

2 g  
2 g 

The terms on the right hand side of the equation indicate the changes in the total energy
of the flow, attributable to the pressure energy, (h1 - h2), potential energy (z1 – z2) and
the corresponding kinetic energy, where point 1 is upstream of point 2. The terms on
the left hand side of the equation determine the loss of total energy between the two
points. This accounts for the losses at bends, fittings, changes in cross-sectional area
and pipe frictional losses.
BS8490:2007[10] suggests the use of the Colebrook-White relationship to calculate
energy losses due to hydraulic resistance. However, these relationships are known to
perform poorly in applications where the water contains air[21, 22]:
 k
 u 2 
P
log 10 
i F  

 3710d i
 8 g .di 

2

1.775v  

3
g .i F .d i  
This equation involves the pipe diameter, surface roughness and viscosity of the liquid.
These two equations are applied across the whole piping network to obtain the SRD’S
design. Due to the complex nature of the equations and the large number of calculations
needed, a purpose designed computer analysis package is required to solve the
problems accurately and efficiently. The design of SRD’S is therefore an analytical
process of matching the resistance of that network, to the height of the building at the
design flow capacity.


However, steady-state design methods are not applicable when a SRD’S system is
exposed to either, a rainfall event below the design criteria, or an event of time-varying
intensity. During these periods the flow may contain substantial quantities of air. As
such events are the norm; it is clear that current design methods may not be suitable for
determining the day-to-day performance characteristics of siphonic roof drainage
systems. This is a major disadvantage, as it is during these events that the majority of
operational problems (noise and vibration) tend to occur. There is also no published
method to assess ability to prime[16].
The design methods described cannot account for occurring operational problems. The
pressures in SRD’S can fluctuate[11] due to:

Interaction with underground system

Sudden partial or total blockage of outlets.

Changes in pipework configuration during installation from that designed.

Volumes of air entering the system
RDS are typically designed to drain the runoff rates associated with short duration, long
return period events typically in excess of 30 years[13]. However sewers may surcharge
at a frequency in excess of once in every five years, therefore when a storm occurs
which approaches the design storm, it is likely that the connecting sewer will be
surcharging[11].
Studies have shown that the use of any configuration other than a freely discharging
vertical downpipe results lower system capacity[19]. Submerged terminal connections
result in significantly longer priming times and higher gutter flow depths than those
discharging freely. This is a result of the increased energy required to purge the air
from the system[13].
Current design approaches do not consider how these fluctuations influence system
performance[11].
2.1 Computer modelling
To augment the understanding of siphonic roof drainage systems and in response to
their perceived shortcomings, a series of research projects have been undertaken, to
develop numerical models for use as diagnostic design aids[12].
Transient propagation in SRD’S is an example unsteady flow which has been defined
through the St Venant equations of continuity and motion. These are a pair of quasilinear hyperbolic partial differential equations that may be solved numerically by use of
the method of characterises (MoC) to transform the partial into total derivatives. The
MoC is a mathematical technique suited to the solution of pairs of quasi-linear
hyperbolic partial differential equations, linking two dependent and two independent
variables. The equations defining pressure transient propagation within SRD’S belong
to this family of problems[17].
A numerical model has been developed (SIPHONET) that can accurately simulate
priming and steady-state conditions is single outlet SRD’S[23]. This model has been
enhanced to simulate multi-outlet SRD’S (SIPHONET2)[19].
The principal uncertainties in these models derive from the following error sources:
1. The flow is not genuinely steady-state
2. The flow is two phase and contains unknown quantities of air
Although the performance of these models can accurately simulate SRD’S (minor
discrepancies are considered to be as a result of turbulent gutter conditions), problems
with numerical stability and extended computational run times restrict their general
applicability[19].
2.2 Field Studies
Siphonic drainage test rigs, under laboratory conditions, have been recorded as priming
in less than 60 seconds. Studies on full scale systems installed on buildings in
Edinburgh were noted as having slightly longer priming times[16]. The modest size of
these systems indicate that current design methods may lead to gross underestimation
of the actual priming time for larger facilities[13].
A number of serious problems are known to have occurred with siphonic roof drainage
systems. However it is considered that the number of failures is only a small percentage
of system installed throughout the world. Furthermore there is no evidence to suggest
that SRD’S are more susceptible than CRDS to failure[16].
Where failures have occurred in single or multi-outlet systems they have invariably
been the result of one or more of the following[12]: 
a lack of understanding of operational characteristics

material specification

installation defects

poor maintenance
Reports indicate that the majority of operational problems relate to noise/structural
vibration and tend to occur at sub-design conditions. This is a major disadvantage as
RDS predominately operate in this region[12].
Proposed solution
This paper focuses on a fundamental reconsideration of SRD’s, to yield a system which
will cope over a wider range of rainfall rates. By examining the equations above, it is
clear that there are no significant restrictions to be expected in reducing the size of the
pipes involved. Furthermore, there are no factors which indicate that horizontal priming
legs are required. Therefore, the solution described in this paper is entirely consistent
with accepted SRD theory. The proposed solution comprises two separate concepts:
(i) A group of several individual identical system pipes, where the number of pipes
used, and the diameter chosen is the primary mechanism for dealing with different
rainfall intensities. The siphon diameters are related to roof area, and with upper
terminations staggered one above the other.
(ii) A small mobile cap is suspended above the inlet pipes, with a mass or force
balance system designed to eliminate noise by controlling siphonic action. This
modulates or ‘digitises’ the flow, leading to significant noise reduction.
These together have the potential to provide the following advantages over the typical
siphonic system:
• any rainfall range can be accommodated, ensuring siphonic flow at typical and
severe storm intensities over greater ranges
• smoothly transitioning between flow regimes automatically
• flood risk and noise problems are minimised
• the new system is immune to over sizing problems
Such a system will cope with any range of rainfall intensities and will become more
useful as climate change brings more diverse and intense storms to the UK and other
locations.
3.0 Modified System Description
Essentially, the multiple pipe system is simply a group of small siphon pipes. The basic
design is illustrated schematically in Figure 1.
The lowest (first to prime) pipe is flush with the roof level, so that standing water on
the roof is eliminated. Bends in the pipework are accommodated with combinations of
rigid fittings. The pipes are fixed to the wall as with any other siphonic pipe. In use, the
siphon with the lowest opening operates first at the smallest capacity, then each
successive siphon primes and the summed total of all primed pipes accommodates
progressively greater flow rates: the maximum capacity is achieved when all pipes are
primed. Horizontal legs were not found to be essential for priming in laboratory trials,
and diameters as small as 19mm gave flow rates of approximately 1ls-1. The siphon
pipes could protrude at different heights from the bottom of a gutter. At the lower
termination, the siphons would have a free air discharge into a sump or another suitable
fitting that articulates with the local drain. It is not necessary for all pipes to terminate
at the same place.
3.1 Modulation Cap Description
A mobile cap is suggested as a means of eliminating the transitory noise that results
from the individual pipes priming. This cap, shown in Figure 2, is suspended above the
pipes with a counterweight or force balance, so that priming is induced in a fraction of
a second, thereby limiting system noise. Similarly, disengaging of the system is also
induced in less than a second, therefore effectively turning the analogue system into a
modulated, or digitised, 1-bit system.
The cap is positioned so that in its resting state, when drainage is not required, its edge
overlaps the top of the drainage pipe as illustrated in Figure 4. The counterweight is
sized with respect to the cap so that it will raise the cap and maintain this degree of
overlap when the system is dry. In the laboratory demonstration system, this distance of
overlap is 1 centimetre and the counterweight is attached through a pivoted weight
system. The cap diameter is 5cm and the height is 8cm. Drainage into the pipe begins
when water rises to and over the top edge of the drainage pipe.
Prior to this, at the point at which water reaches and passes the base of the cap but
before the water passes the top of the drainage pipe, the counterweight causes the cap
to remain at rest in its upper position. The water level inside the cap is free to rise as the
drainage pipe, which at this point is empty, provides an air path. As the water passes
over the top of the drainage pipe and begins to flow down it, the system generates
siphonic flow.
This in turn creates the suction force normal in siphonic operation, which acts against
the weight of the counterweighting system and the cap is pulled down to the lower
position which has the effect of suddenly increasing the water depth around the cap,
thereby ensuring noiseless siphonic operation.
The cap contains 'guide vanes’, which are placed on the inside of the cap so that they
come into contact with the top edge of the pipe when the 'suction' force of the rising
water pulls the cap down. These 'guide vanes' are also placed in such a manner as to
allow the flow of water to continue into the drainage pipe when the 'suction' force has
pulled down the cap so that the 'guide vanes' are in contact with the top of the drainage
pipe. The 'guide vanes' therefore maintain the correct minimum clearance for water
flow by preventing the cap colliding with the top of the drainage pipe and thereby
shutting off flow. They also prevent lateral or sideways movement and therefore
rattling of the cap. The lower operating position of the cap is therefore set by the depth
of the 'guide vanes' placed on the inside of the cap, which touch the top of the drainage
pipe.
3.2 Results
A basic system comprised of 3 siphon pipes, with modulation caps fitted to the lowest
two pipes, has been set up in the DRG laboratory. This has been tested with a 19mm
ABS plastic pipe and a range of flow rates designed to determine the limits of this
system. A header tank was set up to represent the roof gutter and a simple recirculation
system was established with an available head of 7.5m, shown schematically in Figure
4.
For this system, preliminary flow measurements are as follows. Figure 5 illustrates the
dependence of siphonic flow rate with length of down pipe, with a minimum length of
about 4m to develop full siphonic flow. The same results were obtained with the cap
present or absent.
Figure 6 illustrates the performance of a similar system, but with en external sleeve
fitted to the 19mm pipe. This was an attempt to determine whether the same maximum
flow rate could be achieved while running a second pipe internally to the first: initially,
the approach was to have concentric system pipes rather than parallel. Unfortunately, as
can be seen from Figure 6, the high velocities achieved simply increased the friction
due to the extra surface area of the internal pipe which quickly overcame the benefits of
the second concentric pipe.
Figure 7 illustrates the minimum water depth required for air-free induction of full
siphonic flow both with and without a cap fitted. To provide a range of data points in
the small test system used, some throttling was used to simulate extra friction due to
length, as terminal velocity had been reached. There is certainly some benefit to the cap
in terms of reducing the water depth required to initiate siphonage, however main
benefit of the cap is in reducing noise. It is also worth pointing out that, as soon as the
cap is drawn down into the operating position, the water level lowers below that of the
starting depth due to the fact that the cap bottom has been submerged.
In the laboratory demonstration system, the cap moves to the operating state in
approximately 0.75 seconds. Whilst the flow of water into the drainage system
continues, the suction force keeps the 'guide vanes' in contact with the top of the
drainage pipe, and the water flows into the drainage pipe.
When the flow of water ceases, or substantially decreases, to the point that the level of
water outside the cap falls below the bottom of the cap, the 'suction' force is therefore
decreased to the point that it is insufficient to keep the cap in the lower position. The
suction force is suddenly reduced as air is drawn into the system, reducing the water
flow rate and stopping siphonic action, allowing the counterweight to raise the cap
against the remaining reduced suction level.
In the laboratory demonstration system, this raising of the cap occurs within 0.5
seconds. Consequently the flow of water into the drainage pipe ceases completely.
Therefore, through the use of the present invention in a drainage system, and in
particular a siphonic drainage system, flow of water into a drainage pipe is either
constant or non-existent, avoiding resultant noise as outlined above, created when a
relatively small amount of water is flowing into a drainage pipe. Furthermore, by fitting
caps to a set of side-by-side pipes, each will operate noiselessly, and a wide range of
flow rates can be handled.
Importantly, the overlap of the cap and the drainage pipe causes siphonage to begin at a
higher water level that that at which it is stopped, thereby preventing ‘hunting’ of the
system. Hunting occurs when a system attempts to choose between two operating
states, when the selection criteria are identical. This causes the system to oscillate
between the two states.
This flow modulation cap is easy to fit to each pipe in a side-by-side drainage system or
a conventional siphonic system. By stopping siphonage through breaking the suction
rather than sealing of the cap, water hammer is avoided. The device is fail-safe.
Siphonage will be induced at any position of travel of the cap, and an immobile cap in
any position will result in noise due to air entrainment at the same level as a system not
fitted with the cap.
Figure 8 illustrates recorded water depth against time in a water tank fitted with three
19mm diameter siphons, supplied with 1, 2 or 3 l/s, measured with respect to the top of
the relevant siphon pipe. This indicates stable operation and an uneventful handover
between each of the siphons: two or three stages are just visible in the top two curves,
indicating successful sequential operation and suggesting that a very extended, stable,
design curve can be attributed to the system, rather than a single design flow rate.
4.0 Conclusion and Further Work
A parallel pipe siphon system appears to offer benefits and avoids sizing problems
associated with current siphonic systems. Laboratory scale tests demonstrate the basic
feasibility of the system and indicate that handover of flow between pipes occurs
smoothly, and that the flow modulation cap appears to function reliably.
A primary requirement of this work is the need to develop a sizing procedure for
choosing the number and diameters of siphons in a proposed multi-siphonic system.
For example, a low rTS of 3 or less may require just two or three pipe siphons, while
higher rTS values require more. Figure 9 illustrates the anticipated format of the sizing
nomogram. Note that if only one storm intensity was ever encountered, only 1 pipe is
required. Higher rTS ratios force the user to the appropriate curve, from which the
required number of downpipes are selected according to roof area. The inherent safety
of the system is revealed when oversizing a system is considered: the extra siphons
simply remain empty, rather than contributing to system hunting between the primed
and unprimed state. If a conventional siphonic system is oversized, it will fail to prime
and the roof would be in danger of flooding.
The aim of proposed development work would be to develop a siphonic roof drainage
system which works as a siphonic system with a greater operational range of rainfall
rates than current systems. It is anticipated that the following stages would be involved
in the development process:
(i) Determine the relationship between inlet vertical stagger, pipe diameter and
number of pipes for a suitable range of rainfall rates, effectively establishing the rTS
value limits
(ii) Develop a sizing technique and make it available to those concerned with
building design and water use.
(iii) Conduct site trials and validate the sizing technique and an acceptable degree of
pipe flexibility.
(iv) Quantify the nationwide and worldwide application potential for a system with
wide rainfall rate tolerance.
References
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
“An Introduction to Simple Climate Models used in the IPCC Second Assessment
Report - IPCC Technical Paper II”. February 1997. JT Houghton, LG Meira Filho,
DJ Griggs and K Maskell (Eds). IPCC, Geneva, Switzerland. pp 51: available
from: http://www.ipcc.ch/pub/online.htm
“IPCC Third Assessment Report: Climate Change 2001: Synthesis Report”.
Stand-alone edition - Watson, R.T. and the Core Writing Team (Eds.) IPCC,
Geneva, Switzerland. pp 184: available from: http://www.ipcc.ch/pub/online.htm
Demetri, G. (1995), Product Review: Vacuum systems - siphonic drainage;
Architects Journal, May 1995, pp.31-40
"Utilization of Vertical Profile of DSD Into Building an Algorithm for Estimating
Ground Rainfall Amount Using RADAR," III International Symposium on
Hydrologic Applications of Weather RADAR, Aug. 20-23, 1995, Sao Paulo,
Brazil.
Rattenbury, J.M., Sommerhein, P., Smith, W.M., Arthur, S., Duren, G.S.,
McDanal, S.J., Moore (Jr), J.S., Sargeant, J.D., Wearing, M., & Yap, K.L., (2005),
A112.6.9-2005 Siphonic Roof Drains, American Society of Mechanical
Engineers, ISBN 0-7918-2970-7
May RWP. Design criteria for siphonic roof drainage systems, Report SR654.
England: HR Wallingford Ltd., 2004.
Gargano R, Hager WH. Undular hydraulic jumps in circular conduits. Journal of
Hydraulic Engineering 2002; 128:1008-1013.
Wright GB, Swaffield JA, Arthur S. Investigation into the performance
characteristics of multioutlet siphonic rainwater systems. Building Services
Engineering Research & Technology 2002; 23(3): 127-141
Roberts, K. (2001), Theme: roofing – siphonic roof drainage; Architects Journal,
November 2001, pp.22-40.
BS 8490:2007, Guide to siphonic roof drainage systems
Arthur, S., Swaffield, J.A. (2001), Siphonic roof drainage: current understanding;
Urban Water Vol.3, 2001, pp.43-52.
Arthur, S., Wright, G.B. (2005), Recent and future advances in roof drainage
design and performance; Building Service Engineering Research and Technology,
26, 4 pp.337-348.
Arthur, S., Wright, G.B. (2007), Siphonic roof drainage systems – priming focused
design; Building and Environment Vol.42, pp.2421-2431.
Wallis, GB. (1969), One-dimensional two-phase flow; New York, McGraw and
Hill
Acheson, D.J. (1990), Elementary Fluid Dynamics; New York, Oxford University
Press
Swaffield, J.A., Wright, G.B., Jack, L.B. & Arthur, S., (2004), Pressure transient
analysis to inform system design for building and roof drainage systems, Proc The
Practical Application of Surge Analysis For Design And Operation 9th
International Conference, ISBN 1-85598-051-7. Junior Author – Peer Reviewed
Swaffield, J.A., Ballanco, J., McDougall, J.A. (2002), Pressure surge in building
utility services exacerbated by the presence of trapped or entrained air; Building
Service Engineering Research and Technology, 23,3 pp.179-196.
18 Swaffield, J.A., Jack, L.B., Campbell, D.P., Gormley, M. (2004), Positive air
pressure transient propagation in building drainage and vent systems; Building
Service Engineering Research and Technology, 25,2 pp.77-88.
19 Wright, G.B., Jack, L.B., Swaffield, J.A. (2006), Investigation and numerical
modelling of roof drainage systems under extreme events; Building and
Environment Vol.41, pp.126-135.
20 Arthur, S., Swaffield, J.A. (2001), Siphonic roof drainage system analysis utilising
unsteady flow theory; Building and Environment Vol.36, pp.939-948.
21 Idelchik, I.E., (1987) Handbook of Hydraulic Resistance; 3rd edition; Begell
House, Inc. 1996;
22 Miller, D.S. (1978) Internal Flow Systems; Bedford, BHR Group Limited
23 Arthur, S., & Swaffield, J.A., (1999), Numerical modelling of the priming of a
siphonic rainwater drainage system, The Proceedings of Chartered Institute of
Building Services Engineers: Building Services Engineering Research and
Technology - Proceedings of the Chartered Institute of Building Services
Engineers, Vol 20, No. 2. Senior Author