See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/316889515

Piano Key Weir spillway standard design principles and flow induced
vibrations

Conference Paper · May 2017

CITATIONS READS

0 2,053

1 author:

Frank Denys
Stellenbosch University
14 PUBLICATIONS   5 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Fluid Structure Interaction of Piano Key Weirs View project

All content following this page was uploaded by Frank Denys on 13 May 2017.

The user has requested enhancement of the downloaded file.

PIANO KEY WEIR SPILLWAY STANDARD DESIGN PRINCIPLES
AND FLOW INDUCED VIBRATIONS
FJM Denys1
1. Department of Civil Engineering, University of Stellenbosch, Stellenbosch, South Africa

1. INTRODUCTION TO PIANO KEY WEIRS

Piano Key Weirs (often shortened to PKW or PK weir) are a relatively recent development in the field
of spillway hydraulics. They were first developed in the late 1990’s and early 2000’s as part of an
investigation by Hydrocoop (France) and the University of Biskra (Algeria), among others, into
improvements for the well-known labyrinth weir (Lempérière & Ouamane 2003). The typical labyrinth
weir has a zig-zag plan layout such that its total length is longer than that of a straight linear weir. PK
weirs take this concept one step further by having an alternating rectangular shape in plan.
Furthermore, these rectangular keys have sloping bases which guide flow over and away from the
crest of the weir, thus increasing the discharge capacity for a given overflow head. Incidentally, the
repeating staggered overhangs resemble piano keys, hence their name. A typical Type-A piano key
weir is shown in Figure 1 and Figure 2. Other types include Type-B, which incorporates only upstream
overhangs, and Type-C, which incorporates only downstream overhangs.

Figure 1. PKW schematic diagram (Machiels et Figure 2. Typical PKW cross section (Kabiri-
al. 2014) Samani & Javaheri 2012).)

Motivations for their use are varied but mainly revolve around their large specific discharge. As a
result, they are ideal for dam safety improvement programmes (in terms of inadequate spillway
capacity) as well as the rehabilitation of, or upgrades to, existing water resource schemes (raising of
dams). There are several technologies that have been developed which are capable of raising or
altering an existing spillway without increasing the risk of overtopping. However, many of these involve
mechanical gates, which are susceptible to failure if not maintained or operated correctly, and are thus
not favoured by many dam experts. Gate management is also of concern throughout Africa; so much
so that avoidance of their use is being recommended by many dam specialists (Lempérière et al.
2013). Piano key weirs, on the other hand, are permanent structures that are able to safely, efficiently
and economically achieve an increase in the full supply level or the discharge capacity of a spillway
without the need to raise the dam wall as a whole.

Several PK weir projects have been implemented worldwide despite historic shortcomings in a
comprehensive understanding of their behaviour. Countries where they have been successfully
implemented include Algeria, Australia, Canada, France, India, South Africa, Sri Lanka, Switzerland,
UK and Vietnam (see Figure 3 and Figure 4). In most instances where they have been previously
deployed, the designers made use of scaled physical models and generalised performance curves to
predict their behaviour (Schleiss 2011). Locally, the raising of the Hazelmere Dam using a PKW, which
is currently under construction, was iteratively tested using scaled physical models in the Department
of Water Affairs and Sanitation (DWS) hydraulics laboratory (see Figure 11). It is likely that the design
of the raising of the Tzaneen Dam using PKWs will undergo a similar process (Botha et al. 2013).

1
Figure 3. L’Etroit Dam, France (Erpicum et al. Figure 4. Malarce Dam, France (Erpicum et al.
2011) 2013)

2. PIANO KEY WEIR HYDRAULICS

2.1 Discharge efficiency

The piano key weir is a non-linear overflow structure which causes complex flow patterns in its vicinity.
As stated in Section 1, the design aims to provide a much longer overflow length than normal linear
weirs. In so doing, an increased specific discharge can be discharged at a given upstream head when
compared to linear weirs. At their peak efficiency these structures can allow specific discharges of up
to 100 m3/s/m (Lempérière & Ouamane 2003), although in practice the maximum discharge values are
usually in the order of 20 m3/s/m for a depth of 2 m (Laugier et al. 2017). This is typically between two
to four times higher than that for a linear weir at a similar hydraulic head.

As can be seen in Section 4, several quantitative methodologies have been developed to describe the
discharge behaviour of a PK weir. One such methodology is to assume that the PKW is a linear
structure meaning that the standard weir overflow equation applies, as follows:

𝑄 = 𝐶𝑑𝑊 𝑊√2𝑔𝐻3 (1)

The discharge coefficient, CdW, can then be estimated from the physically measured, or numerically
modelled discharge (Pfister & Schleiss 2013b). The W subscript of the discharge coefficient refers to
the notion that the discharge being estimated uses the linear width of the PK weir (as defined in
Section 2.2) and not that of the overflow crest length as a whole, which can be many multiples higher
than the width. The benefit of the longer crest is thus reflected in the discharge coefficient.

Typical discharge coefficients, as estimated using numerical modelling, for a number of overflow
structures are presented in Figure 5. They show that non-linear weirs are a great deal more efficient
than their linear counterparts. However, this behaviour only holds true for low upstream water heads.
As the water level increases, proportionate to the weir’s height, the relative benefit reduces
asymptotically. This is why PKWs should be designed to operate only at lower heads. Although PKWs
have been tested at a range of upstream heads (0 < Ht/P < 3), none have been built to operate at
larger than Ht/P = 0.66. Most prototypes are designed to discharge their maximum flows at around
Ht/P = 0.3 (Leite Ribeiro et al. 2013).

It should be noted that, under certain conditions, a labyrinth weir is capable of achieving a higher
discharge than a PKW at a lower cost. Therefore, each spillway type (linear, labyrinth and PKW)
should be assessed on its merits during the options analysis and preliminary design process.

2
 2.5
Ogee
PKW

Discharge coefficient (CdW)

2
Labyrinth

1.5

0.5

0
0 0.2 0.4 0.6 0.8 1 1.2
Ratio (H\P)

Figure 5. Total discharge coefficients for various weir types (Blancher et al. 2011)

2.2 Geometric parameters

Unlike a standard sharp crested weir, there are a large number of geometrical parameters which have
an influence on a PKW’s discharge behaviour. In order to assist in research efforts and prevent
confusion, a standardised terminology to describe these was proposed in Pralong et al (2011a). The
terminology is comprehensive, thus, for brevity, only a summary has been provided here. The key
parameters relating to physical dimensions are depicted in Figure 6 and Figure 7.

Figure 6: Fundamental parameters of a PKW – 3D view (Pralong et al. 2011)

Figure 7: Fundamental parameters of a PKW – plan view (left) and cross section (right)
(Pralong et al. 2011)

3
The above figures indicate most of the influential geometrical parameters that dictate PK weir
behaviour. The subscript “i” refers to an aspect of the inlet key and “o” the outlet key. Some of the
key parameters and related ratios are further described in Table 1.

Table 1. Fundamental parameter nomenclature (Pralong et al. 2011)

Parameter Definition
B Upstream-downstream length of PKW, B = Bb + Bi + Bo
Bh Sidewall overflowing crest length (between inlet and outlet key crest axes)
Bi, Bo Downstream (inlet key) / Upstream (outlet key) overhang length
CdW Discharge coefficient related to total width, where 𝑄 = 𝐶𝑑𝑊 𝑊√2𝑔𝐻3
Ht Total head over crest upstream of weir
L Total developed length along overflowing crest axis
Lu Developed length of one PKW unit, Lu = Wi + Wo + 2*Ts + 2*Bh
n Developed length ratio, aka magnification ratio, n = L / W
Nu Number of PKW units
Pi, Po Height of the inlet entrance / outlet exit measured from PKW crest
Q Discharge (m3/s)
q Specific discharge (m 3/s/m)
Ro Outlet key parapet wall height
Si, So Slope of the inlet/outlet key apron
Ts Sidewall thickness
W Total width of the PKW
Wi, Wo Width of the inlet/outlet key (wall to wall)
Wu Width of a PKW unit (one inlet + two half outlets), Wu = Wi + Wo + 2*Ts

Of this plethora of factors, it has been determined that knowing four dominant geometrical parameters,
namely the crest length L, the overflow head Ht, the upstream weir height Pi, and the total weir
(channel) width W, (encapsulated into the dimensionless ratios, L/W and Ht/Pi) the discharge can be
empirically determined with a maximum error of only ±17% (Leite Ribeiro et al. 2013). The additional
parameters also play a role but their individual effect on discharge is generally limited to less than 5%
(Lefebvre et al. 2013). Interesting parameters which have been determined to boost the discharge
efficiency are expanded upon in Section 5.

2.3 Hydraulic behaviour

As alluded to in the introduction, PK weirs establish a unique, complex and three-dimensional flow
field in their vicinity. At low relative water level heads, the weir acts in the same manner as a standard
linear weir with a very long crest length. Flow over each of the three portions of the piano key, namely
the upstream crest, the downstream crest and the side crest is directly perpendicular to each of the
respective crests.

At slightly higher overflow heads, the additional velocity and longitudinal momentum of the flow in the
inlet key causes flow over the side crest to deviate from this normal vector. As a result, the flow
efficiency over the side crest diminishes slightly. At even higher heads, the dimensions of the inlet
and outlet key come into play.

4
 Figure 8. PKW streamlines (Machiels, 2012)

Flow in the inlet key has been described as the “engine” of PK weir flow, whereas flow in the outlet key
is described as the “brakes”. This apt analogy alludes to the fact that a wider inlet boosts flow into the
weir. However, a wider inlet key coincides with a narrower outlet key which then restricts flow out of
the weir. A suitable balance must thus be attained between these two opposing effects (Machiels,
2012).

From a hydraulic perspective, the inlet key should be shaped and dimensioned such that the flow
velocity within it is as smooth and subcritical as possible. This is, of course, only achievable at
relatively low flows. At higher flows, the converging flow into the inlet key from the upstream water
body causes energy losses which decrease its efficiency. Furthermore, should the inlet key be too
narrow or the inlet area too small, there is the possibility that flow in the inlet key becomes partially
supercritical. In such an instance, the side crest length beyond this critical point is effectively cut off,
meaning that its discharge efficiency is dramatically reduced (Machiels, 2012).

The hydraulics of the outlet key revolves around the occurrence of local submergence in the key. The
width of this key and its slope dictate the local water level. When this water level exceeds the weir’s
crest level at a particular location, local submergence effects limit the discharge at this location.
Therefore, at higher discharges, a significant portion of the side crest suffers from submergence
effects, reducing the discharge efficiency of the weir as a whole. Increasing the volume of the outlet
key, by, for example, including a parapet wall on the upstream crest, can limit the onset of this local
submergence (Machiels et al. 2013).

3. UNIQUE DESIGN ASPECTS

PW weirs have several unique design aspects which enable them to be effective alternative solutions
for both dam and river applications.

3.1 Reduced structural footprint

One of the main advantages of a PK weir is its relatively narrow footprint when compared to normal
labyrinth weirs. This is not only beneficial from a cost perspective (with its much reduced volumes of
concrete) but also allows these structures to be placed on the narrow crests of gravity dams where
labyrinth weirs are not viable due to their large base widths. A cost comparative exercise conducted
by Paxson et al. (2013), which compared PK weirs, labyrinth weirs and gated spillways, concluded
that PKWs are an ideal and unique solution for increasing the spillway capacities of existing dams.

5
Furthermore, the internal slopes in each of the keys reduces the height of the transverse walls which
then leads to a reduction in the hydrostatic forces acting on these walls. This can lead to savings in
the structural costs. However, a typical PKW is generally designed with more reinforcement steel
(~90 kg/m3), when compared to classic structural loads, in order to withstand the appreciable thermal
loads which can develop in the structure (Laugier et al. 2013).

3.2 Submergence
Due to the PKW’s unique flow patterns, it is able to operate under submerged conditions at a higher
efficiency (i.e. a lower upstream head for a given discharge). This is despite the fact that, at such large
heads, the weir’s discharge efficiency is not much different to that of a normal linear weir (Belaabed &
Ouamane 2013, Cicero & Delisle 2013a). 
 This characteristic makes PK weirs effective in riparian
conditions where developments upstream of the weir may place limits on the water elevations in the
river. This concept was successfully tested in 2016 when the van Phong Barrage in Vietnam was
completely submerged.

3.3 Debris/driftwood
The sensitivity of a PK weir dam spillway to debris and driftwood has been investigated and found to
be relatively small. Although the presence of debris on the PK weir does marginally reduce the
discharge efficiency of the weir (Pfister et al. 2013), laboratory studies found that at larger flows, most
debris would be washed downstream. Should such debris not wash away, the PKW still retains
approximately 75-80% of its discharge capacity. This is due to the unique flow dynamics of the
structure which draws flow from below the surface of the upstream water level, thus bypassing any
debris that may be present (Laugier et al. 2013).

In riverine settings, debris may approach the PK weir from below the surface in addition to the floating
debris. Such debris may collect under the upstream overhangs of a typical Type-A PKW, which will
then need to be removed by hand. Type-C PK weirs, with only downstream overhangs, have been
shown to collect less driftwood (Belzner et al. 2017).

3.4 Modular construction

Due to its thin and repetitive nature, it is possible to construct a typical PKW off site in precast modular
units which are then assembled on site as was done at the Black Esk Reservoir, UK (Ackers et al.
2013). This not only allows for the easier construction of the thin-section concrete or steel members
but also allows for better quality control resulting in a more accurate and level crest profile. Pre-
production also leads to cost savings. Furthermore, in locations with very short available time
windows for construction (e.g. due to the regular occurrence of floods) much of the PKW can be
constructed off site and stockpiled prior to their installation.

4. DESIGN METHODOLOGIES

The hydraulic behaviour of piano key weirs, and the large number of variables which affect it, is now
reasonably well understood despite its complex nature. Several general rating curve formulas have
been compiled as described in the subsections below. It should be noted that each of these equations
is only applicable within the geometrical limits which were used to compile them. These limits should
be respected as some of the resultant rating curves are very sensitive to them (Pfister et al. 2012).
These limits are presented in Table 2.

Table 2. Application limits for the Type-A PKW capacity equations (Pfister & Schleiss 2013a,
Machiels et al. 2014)

L/W H/P Wi/Wo B/P Bi/B, Bo/B P/Wu Bi/P, Bo/P

Kabiri-Samani 0.33 – 0.00 –
2.5 – 7.0 0.1 – 0.6 1.0 – 2.5 ? ?
et al. (2012) 1.22 0.26
Leite Ribeiro et 0.50 – 0.20 – 0.25 –
3.0 – 7.0 0.1 – 2.8 1.5 – 4.6 0.29 - 0.65
al. (2012) 2.00 0.40 4.00
Machiels (2012) 0.50 – 0.00 –
4.2 – 5.0 0.1 – 5.0 1.0 – 6.0 ? 0.33 - 2.00
2.00 2.67

6
4.1 Proposed standard reference design
It should be noted that the methodologies described below are not definitive. There are a large
number of parameters which they do not take into account and furthermore, when used outside
realistic ranges for the various parameters, their results become erratic. A good estimate of what is
realistic concerning a PKW is the standard PKW weir shape which was proposed in 2010 (ICOLD
2010). This shape was stated as being close to the hydraulic optimum although it probably doesn’t
reflect an economic optimum (Lempérière et al. 2011). The reference shape is shown below in
Figure 9.

Figure 9: Reference design for PKW Type-A – plan view (left) cross sections (right) (ICOLD
2010)

Lempérière (2013) estimated that using the above standard geometry the unit discharge could be
estimated using the following relationship within the specified limits of: 0.4𝑃𝑚 ≤ 𝐻 ≤ 2 𝑃𝑚 .

𝑞 = 4.3𝐻 √𝑃𝑚 (2)

This linear relationship (𝑄 ∝ 𝐻) is in stark contrast with the standard on-linear weir equation (𝑄 ∝
𝐻 3⁄2 ). Furthermore, the constant, 4.3, represents a fixed-value discharge coefficient which typically
varies with the overflow depth. Despite this, Anderson & Tullis (2013) confirmed that the equation is
valid but that is only applicable to the specified reference geometry and not otherwise.

4.2 Kabiri-Samani & Javaheri 2012

Kabiri-Samani and Javaheri (2012), at the Isfahan University of Technology, Iran, conducted scaled
physical model tests incorporating PKW Types A, B and C with specific discharge ranges from 25 to
175 l/s/m. Both free flow and submerged flow were investigated although the equation below is only
applicable to free overflow. The methodology was developed using, and is thus generally applicable
to, sharp crested PK weirs. The method is based on the ogee crest overflow equation in the form of:
2
𝑄 = 𝐶𝑑 𝑊√2𝑔𝐻3 (3)
3

with the discharge coefficient, Cd, being defined as a function of the various relevant geometrical
parameters:
𝐵𝑖
𝐻 −0.675 𝐿 0.377 𝑊 0.426 𝐵 0.306 𝐵𝑜
𝐶𝑑 = [0.212 ( ) ( ) ( 𝑖) ( ) 𝑒 1.504 𝐵 +0.093 𝐵 ] + 0.606 (4)
𝑃 𝑊 𝑊𝑜 𝑃

7
4.3 Leite Ribeiro et al. 2012
Leite Ribeiro M, Pfister M, Schleiss AJ, and Boillat J-L (2012) also conducted systematic physical
model tests at the Laboratory of Hydraulic Constructions, Ecole Polytechnique Fédérale de Lausanne,
Switzerland. Tests were limited to PKW Type-A with free overflow conditions and ranged across
specific discharges of 26 to 440 l/s/m. The methodology was developed using, and is thus generally
applicable to, PK weirs with half-circular rounded crests. The resultant rating curve is expressed in
terms of the discharge enhancement ratio, which is defined as the discharge of the PKW, QPKW,
relative to that of a linear sharp crested weir, Qs.

𝑄𝑃𝐾𝑊 𝑄𝑃𝐾𝑊
𝑟= = (5)
𝑄𝑠 0.42𝑊√2𝑔𝐻 3

The ratio was found to vary primarily as a function of four key parameters, namely, the overflow length,
L, the weir width, W, the weir height, Pi and the overflow depth, H. Secondary parameters were also
incorporated to reflect the contributions of key widths, overhang lengths and parapet walls.

(𝐿−𝑊)𝑃𝑖 0.9
𝑟 = 1 + 0.24 ( ) (𝑤𝑝𝑏𝑎) (6)
𝑊𝐻

𝑊 0.05 𝑃 0.25
where 𝑤 = ( 𝑖 ) is used to define the influence of the inlet and outlet key ratio, 𝑝 = ( 𝑜 ) to reflect
𝑊𝑜 𝑃𝑖
𝐵𝑜 +𝐵𝑖 −0.5
any difference in the height of the inlet and outlet keys, 𝑏 = (0.3 + ) to specify the influence of
𝐵
𝑅𝑜 2
the overhang lengths and, 𝑎 = 1 + ( ) to stipulate the presence of any parapet walls.
𝑃𝑜

4.4 Machiels 2012

As part of his PhD dissertation, at the University of Liège, Belgium, Machiels (2012) conducted an
extensive parametric study limited to free overflow conditions and charged with specific discharges
ranging from 13 to 400 l/s/m. The methodology was developed using, and is thus generally applicable
to, flat topped PK weirs. The developed analytical discharge formula is based on a summation of the
calculated discharge over each of the three elements of a typical PKW, namely the downstream
overhang (d), the upstream overhang (u) and the lateral, side crests (s). The unit discharge is then
calculated as follows:

𝑄𝑃𝐾𝑊 𝑊𝑜 𝑊𝑖 2𝐵
𝑞= = 𝑞𝑢 + 𝑞𝑑 + 𝑞𝑠 (7)
𝑊 𝑊𝑢 𝑊𝑢 𝑊𝑢

with 𝑊𝑢 = 𝑊𝑖 + 𝑊𝑜 + 2𝑇 being defined as the width of an individual cycle. Each of the specific
discharges over each of the three PKW elements are estimated as follows:

1 𝐻 2
𝑞𝑢 = 0.374 (1 + ) (1 + 0.5 ( ) ) √2𝑔𝐻 3 (8)
1000𝐻+1.6 𝐻+𝑃𝑇

1 𝐻 2
𝑞𝑑 = 0.445 (1 + ) (1 + 0.5 ( ) ) √2𝑔𝐻3 (9)
1000𝐻+1.6 𝐻+𝑃

1 0.833𝐻 2 𝑃 𝛼 +𝛽
𝑒
𝑞𝑠 = 0.41 (1 + ) (1 + 0.5 ( ) ) ((0.833𝐻+𝑃 𝛼 ) 𝐾𝑊𝑖 𝐾𝑊𝑜 √2𝑔𝐻3 (10)
833𝐻+1.6 0.833𝐻+𝑃𝑒 𝑒 ) +𝛽

Note that each of the above three equations defines the weir height differently. P is the height of the
weir, the total upstream height (including that of the dam the PKW is built upon) is 𝑃𝑇 = 𝑃 + 𝑃𝑑 , and Pe
𝐵 𝑃 𝐵
is the mean weir height along the side wall, 𝑃𝑒 = 𝑃𝑇 𝑜 + (1 − 𝑜 ). The Greek symbols define the
𝐵 2 𝐵
−1.446
𝑃 0.7 3.58 𝑆𝑖
influence of the inlet key slope (𝑆𝑖 = ), where 𝛼 = − + 7.55 and 𝛽 = 0.029𝑒 .
𝐵−𝐵𝑜 𝑆𝑖 2 𝑆𝑖

The KWi coefficient defines the influence of the variation in flow velocity along the lateral crest. The KWo
parameter has been incorporated to describe how the effective overflow length over the side wall is
reduced as the overflow nappes in the outlet key interfere with one another and eventually cause local
submergence effects. They were refined after the original study (Machiels et al. 2014) and are defined
below.

8
 𝛾 𝑊𝑖
The inlet key width parameter is equal to 𝐾𝑊𝑖 = 1 − where 𝛾 = 0.0037 (1 − ). The outlet key
𝛾+𝑊𝑖 2 𝑊𝑜
width parameter is a stepwise function defined as follows:

𝐻
𝐾𝑊𝑜 = 1 for ≤ 𝛿1
𝑊𝑜

2 𝐻 3 3(𝛿2 +𝛿1 ) 𝐻 2 6𝛿2 𝛿1 𝐻

2
𝛿2 (𝛿2 −3𝛿1 ) 𝐻
𝐾𝑊𝑜 = (𝛿 3 ( ) − (𝛿 3 ( ) + (𝛿 3 ( )+ (𝛿2 −𝛿1 )3
for 𝛿1 ≤ ≤ 𝛿2 (11)
2 −𝛿1 ) 𝑊𝑜 2 −𝛿1 ) 𝑊𝑜 2 −𝛿1 ) 𝑊𝑜 𝑊𝑜

𝐻
𝐾𝑊𝑜 = 1 for 𝛿2 ≤
𝑊𝑜

where the function’s limits are defined by 𝛿1 = −0.788 𝑆𝑜 −1.88 + 5 and 𝛿2 = 0.236 𝑆𝑜 −1.94 + 5.

4.5 Comparison
Pfister & Schleiss (2013a) presented a comparative review of all three of the above methodologies.
Careful cognisance of each method’s application limitations was accounted for. The various PKW
rating curves were applied to an indicative example (with L/W = 5) and compared with those of a
standard, 100 m wide, ogee and a broad crested weir. The results are shown in Figure 10.

Figure 10. Comparison of rating curves for PKW, ogee crest and broad crested weirs (Pfister &
Schleiss 2013a)

The comparison reveals that the models deliver similar though not identical results. It has been
hypothesised that the differences in the results stem largely from the differences in the crest shape
profiles used in each of the methods. The Machiels (2012) methodology is the more conservative of
the three in that, for a given head, it results in the smallest discharge, or, for a given discharge, it
results in a larger head.

It should be noted that several of the above physical tests were conducted in channels and may thus
not reflect the different approach flow conditions that are present in open reservoirs (Pfister & Schleiss
2013a). Furthermore, none of the equations allow for the economic optimisation of the weir
dimensions, hence additional mathematical or physical modelling investigations are recommended.

4.6 Modelling
Due to their complex nature and the limitations placed on the developed design equations, it is
strongly recommended that either, or both, physical and numerical modelling are applied to any
proposed design. Examples of such modelling investigations are shown below in Figure 11 and
Figure 12. These tests can then also be used to optimize the design of the PKW and add assurance to
its hydraulic behaviour. The design formulas, however, provide a highly useful tool for preliminary
design purposes.

Any such design studies should preferably be published in relevant journals or conference
proceedings so that their findings may be incorporated into on-going research efforts.

9
 Figure 11. Hazelmere Dam Raising model Figure 12. Darthmouth Dam spillway 3D model
study (Philips & Lesleighter 2013)

5. IMPORTANT DESIGN ASPECTS

As explained previously (in Section 2.2) discharge over a PKW is largely defined by the L/W and Ht/Pi
ratios. There are a number of supplementary aspects which can be used to boost the discharge
efficiency of a typical PKW. A number of these were investigated and have thus been incorporated
into the various design equations. There are, however, a number of additional factors which should be
considered during the design process. Several of these are expanded upon below.

5.1 Inlet key vs outlet key width ratio

Numerous experiments have determined that an inlet key larger than the outlet key is generally
beneficial to discharge efficiency. This is primarily due to the fact that a wide inlet key leads to lower
energy losses there which thus boost discharge efficiency. The width cannot be too large, however,
as a point is reached where local submergence in the ever narrower outlet key becomes the limiting
factor.

The optimum value for this ratio (W i/W o) is not fixed, as it is dependent on other factors, but lies
anywhere from 1.0 to 1.5. Most studies assume a ratio equal to 1.2. It should be noted however, that
an economic optimization of the structure will often result in a W i/W o ratio equal to 1.0.

The ratio’s effects have largely been captured in the design equations described above.

5.2 Parapet walls

Parapet walls refer to a relatively short vertical wall positioned on the crest of the PKW. Their
presence has been shown to be beneficial under certain conditions as they increase the volume of the
individual keys in addition to giving the weir greater height. It has been found that there is only a
marginal benefit to a parapet on the downstream crest of the inlet key, but that there is a marked
benefit to placing one on the upstream crest of the outlet key (Machiels et al. 2013). This is because,
with its greater volume, the outlet key is able to operate under undrowned conditions for longer than
would otherwise be the case, allowing the side crest to operate at its free-flowing efficiency for longer.

5.3 Nose
The placement of a tapered nose under the upstream (outlet key) overhang to guide flow to the inlet
keys either side of it has been found to be a promising design feature (Philips & Lesleighter 2013). Its
presence leads to smoother flow lines, lower energy losses and less vortex shedding which thus leads
to increases in the PK weir’s discharge efficiency.

The impact a nose has on PKW behaviour has been investigated but has, to date, not been
incorporated in the design equations. It is in any case doubtful that its effect is large enough for its
inclusion into these formulae. Their use is generally recommended but their quantified effect can only
be determined by physical or numerical modelling.

10
Noses can take any convenient shape from rounded to triangular (extruded or added). The added or
lost mass has an effect on the overall stability of the structure especially Type-A PKWs which are self-
stable without these noses (Laugier et al. 2017).

5.4 PKW Type

As mentioned above, there are several types of PKW. The standard shape generally used is the
Type-A which has both upstream and downstream overhangs. A Type-B PK weir, which has only
upstream overhangs, has been shown to have a marginally larger discharge capacity (by ~3%) than a
standard Type-A geometry. This is because the longer upstream overhangs (without downstream
overhangs) allow for slightly lower energy losses in the inlet key. This benefit is marginal, however,
and the use of a Type-B PK weir does present additional structural challenges which are not
encountered by a Type-A PK weir (Pralong et al. 2011b).

The use of Type-B PKWs should thus only be considered in new dam projects where their unbalanced
structure can be incorporated into the main dam superstructure. Dam rehabilitation and dam raising
projects favour the use of Type-A PKWs since these are balanced structures (i.e. in a static force
analysis the force resultant is located within the base of the structure and is thus generally safe from
overturning).

Type-C PKWs, with only downstream overhangs, have shown no benefit from a discharge efficiency
perspective. Their use may, however, be warranted depending on site specific conditions such as the
management of debris.

5.5 Crest shape

The shape of the crest has a fairly significant role to play on the efficiency of any spillway, especially at
low water levels. Tests carried out on the influence of crest shapes on labyrinth weirs (Falvey, 2003)
should be entirely applicable to PKWs. Their effect on a PKW is definitive at low H/P values, but this
effect reduces significantly as H/P > 0.3. Cicero & Delisle (2013b) investigated the effect using scaled
physical models and determined that a half-round or quarter round crest on the lateral (side) crest had
the largest impact when compared to a flat topped crest.

It was also found that the crest shape on the upstream and downstream portions of the keys had little
influence, probably due to their short length. Nonetheless, in order to achieve smoother flow lines it is
recommended that the upstream crest be rounded (quarter round on its upstream side) and the
downstream crest be sharp crested. Radii of curvature generally range from 0.5 Ts to 1.0 Ts.
Considering construction aspects related to formwork and durability concerns there should be a
preference to quarter or half-rounded shapes (Laugier et al. 2017).

5.6 Economics
A PKW can be designed with a high hydraulic efficiency. Often such a solution is not economic and
thus an optimum balance between efficiency and cost needs to be found. The cost criterion is often
linked to a simple metric such as concrete volume. In this way an options analysis can be carried out
to compare the cost/benefit ratio of a PKW versus that of a labyrinth or even a gated spillway.

A general exercise was carried out in this light which determined that PKW are often the cheapest and
most efficient solution, but that the use of labyrinth spillways can, at times, offer a more ideal solution
based on site conditions and operational restraints (Paxson et al 2013).

5.7 Approach flow

As with most other dam spillways, whether they are uncontrolled linear overflows or gated structures,
the flow patterns of the approach can have a marked effect on their efficiency. Should a PKW be
located on or near an abutment, its effect on the flow patterns should be investigated via a numerical
or physical model.

5.8 Energy dissipation and scour

Due to their high discharge efficiency, energy dissipation downstream of the weir is of concern.
Special attention should be given to the design of energy dissipation structures downstream of a PKW
to prevent undesirable aspects such as scour and cavitation.

11
When placed on a gravity dam with a stepped discharge chute, a PK weir is able to induce fully
aerated flow on the steps much sooner than a standard ogee can (Silvestri et al. 2013). It is thus
thought that their use can increase the energy dissipation on these steps. Laboratory tests have,
however, determined that the residual energy in the flow on the steps downstream of a PKW is
actually higher than if it were downstream of a normal ogee crest (i.e. the energy dissipation is less).
The suspected reason for this is that the PKW focusses flow into jets which exit the outlet key which
do not spread evenly over the whole width of the stepped spillway.

In terms of scour downstream of a PKW in a riverine setting, physical model studies have shown that
the scour depth (and scour hole volume) is not only related to the sediment characteristics, the
discharge, the head differential and the tailwater depth but also to the PKW’s downstream overhang
length, Bi (Jüstrich et al. 2016). Literature comparisons also show that the scour characteristics are
similar to that of jet scour, meaning that the velocity of the jets in the outlet key as well as the angle of
this key are of relevance.

5.9 Constructability
The constructability of a PKW plays a significant role in its design. In many existing prototypes,
limitations on the extent of the overhang lengths were imposed due to practicality, construction
equipment, site access and occupational health and safety factors.

An additional constructability issue is that of timing. The occurrence of periodic or seasonal flooding
can limit access to a particular construction site. Modular construction (see Section 3.4) can assist in
this regards by shortening the amount of time needed for installation.

5.10 Sedimentation
It is relevant to note that the majority of the research into PKWs has been conducted by institutions
located in regions whose rivers have comparatively little sediment transport. As a result, the effect of
sedimentation on a PKW is not yet well defined but is actively being researched.

Labyrinth and oblique weirs are known to establish secondary flow fields in the upstream water body
as it approaches the weir. Under certain scenarios this flow field is capable of scouring the area
immediately upstream of the weir. It is expected that piano key weirs also exhibit this self-cleaning
behaviour (Tiwari & Sharma, 2015). The effect may, however, be too localised to allow the weir to
behave as it would in its normal, sediment free, condition.

5.11 Aeration
Most PKWs built to date do incorporate aeration of the nappe in some shape or form as it is relatively
easy and comparatively cheap to do so. Such aeration is believed to be beneficial overall, however,
its precise necessity is ambiguous. Some physical models of PK weirs have shown that the flow over
the weir is naturally aerated (Hien et al. 2006, Laugier et al. 2013). This is, however, only true for
relatively low flows. At higher flows the air pocket under the nappe becomes isolated and sub-
atmospheric pressure conditions may result. The influence this negative pressure has on the
hydraulic behaviour is thought to be beneficial to discharge efficiency, however, its influence on the
stability of the structure as a whole remains unclear.

An aeration network design have often been based on approximate rules-of-thumb, which consist of
an air flow rate in the order of 5 to 10% of the water flow combined with an air speed in the air supply
system not exceeding 50 to 100 m/s. This can however lead to very large aeration pipes, which may
be overly conservative. This possibly due to the use of these rules in the design of gauging weirs and
flap gates which are very sensitive to problems caused by the lack of aeration. More defined models of
air demand are now being used in PKWs which account for the relative resistance of these structures
to unaerated nappes (Vermeulen et al. 2017).

5.12 Hydrological impact

Due to a PK weir’s high specific discharge at relatively low heads, its response to an incoming
hydrograph can be faster than with a linear weir. Although this is very much dependent on the
characteristics of the upstream water body, the effect of a PKW on the downstream flood regime has
not been extensively investigated.

12
The PKW’s unique rapid response capability has been utilised in a number of hydropower installations
(Escouloubre, Emmenau) which can induce rapidly changing water levels during their emergency
rejection procedures. The use of PKWs then limits the height these water levels can attain (Laugier et
al. 2017).

6. FLUID STRUCTURE INTERACTION

Related to the abovementioned aeration concerns is the effect any lack of aeration has on the
structural behaviour of the weir. The sub-atmospheric region of pressure behind the nappe can
induce oscillations within this nappe which, together with other sources of vibrations such as vortex
shedding in the inlet key, can cause vibrations in the PKW walls.

These flow-induced vibrations are a known concern of overflow structures. In many dam spillways,
these vibrations are not specifically accounted for, since the natural frequencies of these structures
(related to their mass and their stiffness) are such that they do not resonate with those excitations
induced by the flow for significant periods of time. However, in thin-walled structures, such as PK
weirs, the possibility of resonance and its corollary, fatigue induced failure, are real risks. This is
especially the case because such vibrations chiefly occur at low heads, and it is at these low head
ranges that PKWs are expected to operate most of the time since it is at these levels that they are at
their most efficient.

Tests have shown that concrete weirs, with a typical wall thickness of 0.25 m or more, are stiff enough
to withstand such vibrations. Thin walled steel weirs (with a thickness in the order of 5 to 10 mm) may
however be susceptible to such vibrations (Denys, 2017).

7. REFERENCES

Ackers JC, Bennett FCJ, Scott TA, & Karunaratne G (2013). Raising the bellmouth spillway at Black
Esk reservoir using Piano Key weirs. Proceedings of the 2nd International Workshop on Labyrinth and
Piano Key Weirs (PKW 2013), pp. 235-242. CRC Press, Taylor & Francis Group.

Belaabed F & Ouamane A (2013). Submerged flow regimes of Piano Key Weir. Proceedings of the
2nd International Workshop on Labyrinth and Piano Key Weirs (PKW 2013), pp. 85-92. CRC Press,
Taylor & Francis Group.

Belzner F, Merkel J, Gebhardt M & Thorenz C (2017). Piano Key and Labyrinth Weirs at German
Waterways: recent and future research of the BAW. Proceedings of the 3rd International Workshop on
Labyrinth and Piano Key Weirs (PKW 2017), pp. 167-174. CRC Press, Taylor & Francis Group.

Blancher B, Montarros F & Laugier F (2011). Hydraulic Comparison between Piano Key Weirs and
Labyrinth Spillways. Proceedings of the International Conference on Labyrinth and Piano Key Weirs
(PKW 2011), pp. 141–150. CRC Press, Taylor & Francis Group.

Botha AJ, Fitz IP, Moore, AJ, Mulder FE & van Deventer NJ (2013). Application of the Piano Key Weir
spillway in the Republic of South Africa. Proceedings of the 2nd International Workshop on Labyrinth
and Piano Key Weirs (PKW 2013), pp. 185–194. CRC Press, Taylor & Francis Group.

Cicero G-M & Delisle JR (2013a). Discharge characteristics of Piano Key weirs under submerged flow.
Proceedings of the 2nd International Workshop on Labyrinth and Piano Key Weirs (PKW 2013), pp.
101–109. CRC Press, Taylor & Francis Group.

Cicero G-M & Delisle JR (2013b). Effects of the crest shape on the discharge efficiency of a Type-A
Piano Key weir. Proceedings of the 2nd International Workshop on Labyrinth and Piano Key Weirs
(PKW 2013), pp. 41–48. CRC Press, Taylor & Francis Group.

Denys FJM (2017). Fluid Structure Interaction of Piano Key Weirs. Proceedings of the 3rd
International Workshop on Labyrinth and Piano Key Weirs (PKW 2017), pp. 119-126. CRC Press,
Taylor & Francis Group.

13
Erpicum S, Laugier F, Boillat J-L, Pirotton M, Reverchon B & Schleiss AJ (eds.)(2011). Labyrinth and
Piano Key Weirs - PKW 2011. CRC Press/Balkema, 2011.

Erpicum S, Laugier F, Pfister M, Pirotton M, Cicero G-M & Schleiss AJ (eds.)(2013). Labyrinth and
Piano Key Weirs II - PKW 2013. CRC Press/Balkema, 2013.

Falvey HT (2003). Hydraulic design of labyrinth weirs. ASCE Press (American Society of Civil
Engineers).

Hien TC, Son HT, & Khanh MHT (2006). Results of some piano keys weir hydraulic model tests in
Vietnam. Proceedings of the 22nd International Congress of Large Dams (pp. 581-595).

ICOLD 2010. ICOLD Bulletin 144: Cost Savings in Dams. Appendix 2. International Commission on
Large Dams.

Jüstrich S, Pfister M & Schleiss AJ (2016). Mobile Riverbed Scour Downstream of a Piano Key Weir.
Journal of Hydraulic Engineering, 142(11), p.04016043:1-12. ASCE Press.

Kabiri-Samani A & Javaheri A (2012). Discharge coefficients for free and submerged flow over Piano
Key weirs. Journal of Hydraulic Research, vol. 50, no. 1, pp. 114–120, February 2012.

Laugier F, Vermeulen J & Lefebvre V (2013). Overview of Piano Key Weirs experience developed at
EDF during the past few years. Proceedings of the 2nd International Workshop on Labyrinth and
Piano Key Weirs (PKW 2013), pp. 213-226. CRC Press, Taylor & Francis Group.

Laugier F, Vermeulen J & Blancher B (2017). Overview of design and construction of 11 Piano Key
Weirs spillways developed in France by EDF from 2003 to 2016. Proceedings of the 3rd International
Workshop on Labyrinth and Piano Key Weirs (PKW 2017), pp. 37. CRC Press, Taylor & Francis
Group.

Lefebvre V, Vermeulen J & Blancher B (2013). Influence of geometrical parameters on PK-weirs

discharge with 3D numerical analysis. Proceedings of the 2nd International Workshop on Labyrinth
and Piano Key Weirs (PKW 2013), pp. 49–56. CRC Press, Taylor & Francis Group.

Leite Ribeiro M, Pfister M, Schleiss AJ, & Boillat J-L (2012). Hydraulic design of A-type Piano Key
Weirs. Journal of Hydraulic Research, Volume 50, Issue 4, pp. 400-408, 2012.

Leite Ribeiro M, Pfister M & Schleiss AJ (2013). Overview of Piano Key weir prototypes and scientific
model investigations. Proceedings of the 2nd International Workshop on Labyrinth and Piano Key
Weirs (PKW 2013), pp. 273–281. CRC Press, Taylor & Francis Group.

Lempérière F & Ouamane A (2003). The Piano Keys Weir: a new cost-effective solution for spillways.
The International Journal on Hydropower and Dams, vol. 10, no. 5, 2003.

Lempérière F, Vigny JP & Ouamane A (2011). General comments on Labyrinth and Piano Key Weirs:
The past and present. Proceedings of the International Conference on Labyrinth and Piano Key Weirs
(PKW 2011), pp. 17-24. CRC Press, Taylor & Francis Group.

Lempérière F, Ouamane A & Vigny J-P (2013). Piano Keys Weirs (PK weirs) could be used for most
African spillways. http://www.hydrocoop.org/. 


Lempérière, F. (2013). New labyrinth weirs triple the spillways discharge. http://www.hydrocoop.org/.

Machiels O (2012). Experimental study of the hydraulic behaviour of Piano Key Weirs, PhD Thesis
ULgetd-09252012- 224610, University of Liège (Belgium), 2012.

Machiels O, Erpicum S, Archambeau P, Dewals B & Pirotton M (2013). Parapet Wall Effect on Piano
Key Weir Efficiency. Journal of Irrigation and Drainage Engineering, vol. 139, no. 6, pp. 506–511, June
2013.

14
 Machiels O, Pirotton M, Archambeau P, Dewals BJ & Erpicum (2014). Experimental parametric study
and design of Piano Key Weirs. Journal of Hydraulic Research, vol. 52, no. 3, pp. 326–335, 2014. 


Paxson GS, Tullis BP, & Hertel DJ (2013). Comparison of Piano Key Weirs with labyrinth and gated
spillways: Hydraulics, cost, constructability and operations. Proceedings of the 2nd International
Workshop on Labyrinth and Piano Key Weirs (PKW 2013), pp. 123–130. CRC Press, Taylor & Francis
Group.

Pfister M & Schleiss AJ (2013a). Comparison of hydraulic design equations for A-Type Piano Key
Weirs. Proceedings of the International Conference on Water Storage and Hydropower Development
for Africa (AFRICA 2013). Addis Ababa, Ethiopia, 16 to 18 April 2013.

Pfister M & Schleiss AJ (2013b). Estimation of A-type Piano Key weir rating curve. Proceedings of the
2nd International Workshop on Labyrinth and Piano Key Weirs (PKW 2013), pp. 139–147. CRC Press,
Taylor & Francis Group.

Pfister M, Schleiss AJ, & Tullis BP (2013). Effect of driftwood on hydraulic head of Piano Key weirs.
Proceedings of the 2nd International Workshop on Labyrinth and Piano Key Weirs (PKW 2013), pp.
255–264. CRC Press, Taylor & Francis Group.

Pfister M, Erpicum S, Machiels O, Schleiss AJ & Pirotton M (2012). Discussion of ‘Discharge

coefficient for free and submerged flow over Piano Key weirs’. Journal of Hydraulic Research, vol. 50,
no. 6, pp. 642–643, Dec. 2012.

Philips MA & Lesleighter EJ (2013). Piano Key Weir spillway: Upgrade option for a major dam.
Proceedings of the 2nd International Workshop on Labyrinth and Piano Key Weirs (PKW 2013), pp.
159–168. CRC Press, Taylor & Francis Group.

Pralong J, Vermeulen J, Blancher B, Laugier F, Erpicum S, Machiels O, Pirotton M, Boillat J-L, Leite
Ribeiro M & Schleiss AJ (2011a). A naming convention for the Piano Key Weirs geometrical
parameters. Proceedings of the International Conference on Labyrinth and Piano Key Weirs (PKW
2011), pp. 271–278. CRC Press, Taylor & Francis Group.

Pralong J, Montarros F, Blancher B & Laugier F (2011b). A sensitivity analysis of Piano Key Weirs
geometrical parameters based on 3D numerical modelling. Proceedings of the International
Conference on Labyrinth and Piano Key Weirs (PKW 2011), pp. 133-139. CRC Press, Taylor &
Francis Group.

Anderson RM & Tullis BP (2013). Piano Key Weir hydraulics and labyrinth weir comparison. Journal of
Irrigation and Drainage Engineering, 139(3), pp.246–253. ASCE.

Schleiss AJ (2011). From Labyrinth to Piano Key Weirs - A historical review. Proceedings of the
International Conference on Labyrinth and Piano Key Weirs (PKW 2011), pp. 3–15. CRC Press,
Taylor & Francis Group.

Silvestri A, Archambeau, P, Pirotton M, Dewals B & Erpicum S (2013). Comparative analysis of the
energy dissipation on a stepped spillway downstream of a Piano Key Weir. Proceedings of the 2nd
International Workshop on Labyrinth and Piano Key Weirs (PKW 2013), pp. 111–120. CRC Press,
Taylor & Francis Group.

Tiwari H & Sharma N (2015). Turbulence study in the vicinity of piano key weir: relevance,
instrumentation, parameters and methods. Applied Water Science, pp. 1-10, March 2015.

Vermeulen J, Lassu C, Pinchard T (2017). Design of a Piano Key Weir aeration network. Proceedings
of the Third International Conference on Labyrinth and Piano Key Weirs (PKW 2017), pp. 127–133.
CRC Press, Taylor & Francis Group.

15

View publication stats