RD 2011

CFRD 02

The Second International Symposium on Rockfill Dams
1




BASTELOS DAM. A SELF-SPILLWAY ROCKFILL DAM

Antnio Veiga PINTO
1
, Rui MARTINS
2

1, Executive Director Mecasolos, R. Xavier Arajo, Ncleo 8-6B, 1600-226 Lisboa, Portugal,
vpinto@mecasolos.com
2, Research Officer, Retired LNEC, R. Cidade de Benguela, 66, 1 Esq, 1800-073 Lisboa, Portugal,


ABSTRACT

The rockfill dam with self-spillway built, under certain conditions, may be a good solution in economic terms,
since one can dispense with the conventional spillway usually costly in embankment dams. This
communication presents various aspects of Bastelos Dam, built in Portugal, which has a self-spillway built
within the rockfill dam. Some questions are concerning to the design of the spillway, the mechanisms that could
lead to instability of the embankment body, the monitoring plan and to the observed behaviour. From the studies
carried out to Bastelos Dam, it may be emphasized that self-spillway rockfill dams seem to be a good solution
when well designed and constructed.

Key-words: Self-Spillway; Rockfill; Hydraulic Behaviour; Flow Discharge; Through Flow Zone.


1. INTRODUCTION

One of the characteristics of the construction materials referred to as rockfill is the high values of
permeability. This fact leads to design solutions without conventional spillway, since its construction
significantly increases the total cost. The success of this kind of solution which takes advantage of the
drainage capacity of rockfill depends on whether the stability of the structure under seepage forces can
be guaranteed at the same time.

In this paper an attempt is made towards the design approaches for through flow rockfill dams. A
solution of this type has just been used in Portugal (Bastelos Dam). Finally an analysis of the design of
this dam is presented.

The design of this type of dam is done in such a way that excessive head of water in the reservoir is
relieved totally by flow through the rockfill.

Since the beginning and with several cases of failure, the technology of through flow or overflow
rockfill has reached a level where discharge capacities can be guaranteed in temporary or permanent
situations.


2. TYPE OF DAM

The concept of the through flow rockfill dams, or also called self-spillway dams, is attributed to
Wilkins [1].

The self-spillway dams not only dispense the use of a conventional spillway but they also practically
solve the problem of energy dissipation in the restitution zone, given the energy losses of the flow
during seepage and due to geometry of the slope.

Special selection and placement criteria are applied for the rockfill conditioned by hydraulic and
geotechnical requirements, causing the cost to increase.


The Second International Symposium on Rockfill Dams
2
Furthermore measures must be taken for the porosity of the through flow zone not to decrease during
the operational life of the structure.

The failure of this kind of structure, with through flow spillway, is very likely to occur suddenly,
which makes it essential to know the downstream flow discharge.


3. BASTELOS DAM

The study of overflow and self-spillway dams was initiated at LNEC in seventies [2].The necessity for
detailed hydraulic studies through physical modelling of the prototypes was very early appreciated.

Various studies of an experimental nature followed, aiming at establishment of the laws of hydraulic
behaviour [3].

It was also proposed by LNEC, that Monte de Cr Dam, with 9 m of maximum height should be used
as a large scale model for the observation of the behaviour of overflow rockfill dams [4].

In 1989, LNEC was invited to collaborate in the observation plan of Bastelos Dam which was at the
time beginning to be constructed. The quantities to be measured during observation were evaluated by
considering all the elements of the project and the models so far available for the behaviour of this
type of structure. The uncertainty of hydraulic behaviour existing for this type of solution was pointed
out by LNEC. The dam was concluded in September 1993. Some of the recommendations of LNEC
were ignored.

Next, an analysis of the adopted solution will be attempted with the methodology already presented.

3.1 Type of Profile

In agreement with the hydrological data the project estimates a maximum discharge of 130 m
3
/s with a
return period of 1000 years. The catchment area is 26 km
2
with a maximum storage capacity of
1.1 x10
6
m
3
.

The dam is located in Mogadouro Municipality (Portugal) and water supply is the main purpose. The
dam profile is presented in Figure 1. The dam has a maximum height of 22 m. The element of
imperviousness is an upstream bituminous concrete face which extends as far as the base of the
spillway.

Figure 1. Cross-section of Bastelos Dam.

The Second International Symposium on Rockfill Dams
3

The central zone of the dam is supposed to function as a free-fall zone for the flow through the rockfill
which for this purpose has a relatively uniform particle size distribution from 0.70 to 1.25 m.

The exit zone of the self-spillway is situated near the downstream toe with a height of about 7 m. It
consists of rockfill of relatively uniform dimensions between 1.25 to 2.50 m.

In Figures 2 to 5 are presented aspects of Bastelos Dam.


Figure 2. Upstream face of Bastelos Dam.

Figure 3. Downstream view of Bastelos Dam.

The Second International Symposium on Rockfill Dams
4

Figure 4. Reservoir, intake tower and crest of Bastelos Dam.

Figure 5. View of the self-spillway rockfill in the crest of Bastelos Dam.

3.2 Analysis of Hydraulic Behaviour
3.2.1 Crest Discharge Capacity

According to the project, the required value of discharge corresponds to a maximum value of
hydraulic head, H
d
, of 1.39 m, thus discharge capacity being given by:

in which = 0.40 for Q = 130.6 m
3
/s, width of the crest, W = 47.5 m and g = 9,8 m/s
2
.
Thus is not necessarily an unrealistic value but it needs experimental confirmation owing to the
unconventional character of the threshold spillway [5].
H
2g W = Q
3
d
2 /
(1)

The Second International Symposium on Rockfill Dams
5
3.2.2 Discharge Capacity of the Through Flow Zone

The discharge capacity through the rockfill obviously depends on the coefficient of permeability and
consequently the void ratio. The project assumes a value of porosity of 0.35 for zone 2 of the profile
(Figure 1). This value was not verified during construction.
For porosity, n, equal 0.35, the void ratio (e) is 0.54. The hydraulic gradient (i) can be obtained from
the relation H/L.
The vertical length is about 48.5 m. For a head water of (625.5+1.40)-(609+602)/2=19.9 m, and so,
i = 0.41. 1.40 is the difference between MWL and NWL.
The total area is 47.5x(609-602) = 332.5 m
2
, and the active area (S) will be 315x0.35 = 116.4 m
2
.
Considering d = (0.7+1.25)/2 = 0.975 m, will be:

Q = SV
v
(3)

Q = 116.4x1.15 = 133.9 m
3
/s
So the discharge capacity is from the same order of the estimated maximum flow.
Blocks with smaller dimensions exist in the zone of the crest, which may lead to a reduced efficiency
of the spillway.
The obstruction of the sill zone by detritus must be also considered. The metal slab installed along the
upstream of the sill may lead to a certain void ratio reduction, especially when the reservoir is full and
service operations are impossible.
Deficient behaviour of the sill may also occur, owing to differential settlements and consequent
concentrations of the discharge flow.

3.3 Stability analysis

As already mentioned, Bastelos Dam was not designed to function as overflow rockfill. Therefore a
relatively high possibility of failure due to hydraulic instability exists.

The stability of the dam under conditions of steady flow should be evaluated.

In a study carried out at LNEC [5] a series of hypotheses for the shear strength in terms of effective
stresses (since no experimental data were available) and for the phreatic line were examined.

In Figure 6 the results of limit equilibrium analysis with a rather plausible hypothesis (' = 50 and the
phreatic line parallel to the downstream slope) in case of overflow are presented, indicating a clearly
not insufficient value of the factor of safety, F = 0.5.

Figure 7 shows some of the mechanisms that may lead to instability of Bastelos Dam.
(2)
2gedi 0,56 =
Vv
m /s = . x . x . x . , =
Vv
15 1 41 0 975 0 54 0 6 19 56 0 .

The Second International Symposium on Rockfill Dams
6

Figure 6. Stability analysis of Bastelos Dam [5].


Figure 7. Eventual mechanisms of failure of through and overflow rockfill dams.

The Second International Symposium on Rockfill Dams
7
3.4 Monitoring Plan

The observation plan sets up an automatic system measuring the following quantities, Figure 8, [6]:
1.Water level in reservoir;
2.Flow discharged;
3.Piezometric levels in six piezometers installed within the dam for drawing the seepage line;
4.Level upwelling off the spillway;
5.Flow downstream.

In the system installed, the measurement of the flow is done indirectly by measuring the water level in
relevant sections.

To obtain these quantities were used for pressure transducers to atmospheric pressure, submersible.
The conversion levels at flow rates require knowledge of the flow curves in the relevant sections.


Figure 8. Monitoring system of the dam [6].

The discharged flow measurement is made, indirectly, by measuring the water level in the reservoir
that will be reached above the spillway sill. According to the designer, that level will reach 1.39 m to a
maximum flow of 130 m
3
/s.

Associated to the values measured for flows discharged there are four levels of alarm. The heaviest
level corresponds to have an upstream flow discharge higher than estimated, or the average value of
piezometric levels or water levels on downstream of the dam exceeds 95% of the maximum specified.

3.5 Hydraulic Behaviour

The monitoring system has been in test for about half a year in 1998, and led to satisfactory results.
The diagram corresponding to the reading of water levels in the reservoir during this period is
indicated in Figure 9.

The Second International Symposium on Rockfill Dams
8

Figure 9. Reservoir water level [6].

In July 1998 the water level upstream reached an approximately value of 0.25 m. The Figure 10 is an
aspect of the flow discharge during this period.

Figure 10. Flow of water through the spillway [6].
Recently the observation system is not functioning. Still, visual inspections carried out during the
discharge of flood flows, there has been ever observed a water level above the sill of the spillway not
exceeding 0.5 m, what means, less than a quarter of the design flow.





The Second International Symposium on Rockfill Dams
9
4. LESSONS FROM BASTELOS DAM

In case of a situation of overtopping due to insufficient capacity of flood discharge flow inside the
dam happens, there could be overtopping that a state limit, may lead to a sudden failure of the dam. So
while there is no greater experience and check the hydraulic behaviour in this type of dam, only must
be adopted these solutions in the class III dams, ie, with minimal risk.

Note also that the dam has no freeboard and was not experimentally tested.

One aspect to consider is that there is no guarantee that the porosity of the draining rockfill of Bastelos
Dam is not equal to 0.35. It is further accepted that the porosity should decrease over time due to
accumulation of debris in the voids of rockfill.

For example, Laughing Jack Dam on the evolution of voids was of the order of magnitude of the
amounts sated in Table 1 [7]. The owner of the dam referred to [7]: the measured discharges
through the fill suggest that only about 30% of the design flow would be passed at the design
maximum flood lake-level. We attribute this tentatively to the significant quantities of twigs and other
debris trapped in the rockfill during flooding, but it may also reflect a lower than expected fill void
ratio.

Table 1. Evolution of void ratio in the self-spillway of Laughing Jack Dam.

Period Design As built 30 years life
Void ratio 0.85 0.55 to 0.63 0.30 to 0.35

As it was said, there is still not the observed hydraulic behaviour of Bastelos Dam. It could be
important correlate water reservoir levels with discharge, as to estimate V
v
and and the active area for
this dam and for future designs.


5. CONCLUSIONS

Experience shows that the use of rockfill leads to sufficiently stable structures even when they are
subjected to high seepage forces, with low cost proving to be an alternative solution for diversion
works.

In the case of non-protected rockfill the phreatic line must be guaranteed not to reach the downstream
slope surface.

The biggest number of applications is for cofferdams and rockfill bodies during the construction stage.

Failure is probably due to the displacement of blocks near the downstream slope toe resulting in the
formation of a shear surface.

Bastelos Dam, a through flow rockfill, has been built 20 years ago in Portugal. As regards the
hydraulic and structural behaviour and safety of this dam, an analysis of the design characteristics is
presented. From this analysis with a new solution, but not supported in theoretical or experimental
results, some doubts are raised as to whether a good structural behaviour can be expected of this dam.
Nevertheless, the hydraulic and structural behaviour of Bastelos Dam, 20 years working life, has been
satisfactory.

From the studies carried out to Bastelos Dam, it may be emphasized that self-spillway rockfill dams
seem to be a good solution when well designed and constructed.

The Second International Symposium on Rockfill Dams
10
References
[1] WILKINS, J. K. (1956) - "Flow of Water Through Rockfill and its Application to the Design of
Dams", Proc. 2
nd
Australia - New Zealand Conf. on Soil Mechanics and Foundation Engineering,
Canterbury, 141-149.
[2] MARTINS dos SANTOS, L. (1973) - "Overflow Rockfill Dams", (in Portuguese), LNEC thesis,
Lisbon.
[3] MARTINS, R. (1991) - "Principles of Rockfill Hydraulics", in Advances in Rockfill Structures,
Kluwer Pub., 523-570.
[4] MARANHA das NEVES and MARTINS, R. (1980) - "Study of an Experimental Overflow
Rockfill Dam", (in Portuguese), LNEC, 1-25.
[5] VEIGA PINTO, A. and MARTINS, R. (1993) - "Bastelos Dam. Revision of the Design and Safety
Analysis", (in Portuguese), LNEC Report, Lisbon, 1-26.
[6] ALMEIDA GARRETT, J. and TOCO EMLIO, F. (2000) - "Bastelos Dam. Monitoring System.
(in Portuguese), LNEC Report, Lisbon, 1-29.
[7] MARTINS, R. (1988) - "Personal Communication, from The Hydro-Electric Commission,
Tasmania.