General Turbine Flow Conditions

at a HYDROMATRIX®-Module
By Reinhard Prenner 1, Alois Nichtawitz 2, Volker Kienberger 2 and Helmut Drobir 1
1
University of Technology Vienna, Institute of Hydraulic Engineering
A-1040 Vienna, Karlsplatz 13, Austria/Europe
2
VA TECH HYDRO GmbH & Co
A –4031 Linz, Lunzerstraße 78, P.O.Box 28, Austria/Europe

ABSTRACT

This paper contains the results of series of experimental investigations of turbine

influx, efflux and draft tube pressure conditions at a " HYDROMATRIX®-module".
Such a module can be placed in the bay of a run-of-river station to generate
electrical energy. In case of flow separation (air-entrainment) in individual draft tubes
the efficiency of the turbines decreases significantly and the performance of the
turbine can not be guaranteed. In addition, turbine damages can occur easily. In
order to investigate stable turbine flow conditions three different hydraulic models
were built. The first model comprises one full and two adjoining half bays in a scale of
1: 40. This model should allow to appraise different approaching flow conditions to
the turbines and turbine outflow conditions as well. The second one presents a
sectional model with three bulb turbine-axes in span and two turbine-rows in height
(six pack). The third one was a copy of a set of a border bulb turbine row (pier unit)
including a half weir pier to test asymmetric flow conditions up- and downstream of
the module. Both models were built with a complete mechanical turbine system to a
scale of 1:12,5 in plexiglass. Measurements show the energy losses, the net-head
and the dynamic draft- head as a function over all operational gross-heads of the
plant. Operational measures taken to prevent flow separation (air-entrainment) in the
upper draft tube row during lowest tailwater conditions will also be described and
evaluated.

Introduction

HYDROMATRIX®-technology /1/, advanced by VA TECH HYDRO, is a new concept

in hydraulic energy production combining the advantages of proven technology and
low cost installation in existing hydraulic structures. An usage of this technology at
new hydroplants is generally possible and could offer a new way to reduce
construction costs significantly.
A HYDROMATRIX®-module realize the generation of electric energy by means of a
factory assembled module block (“grid”) of several small assembled turbine-
generator units (Figure 1) /2/. The module block consists of a stiff steel fabricated
structure which supports the TG-units including trash racks, draft tubes with control
gates, rooms for HPU, electric switchgear and control systems (Figure 2).

HydroVision 2004 - Copyright HCI Publications, 2004 - www.hcipub.com 1

 HWL
Spillway Gates

Electric Switchgear
& Control System
Intake Trashracks

HPU
TWL

TG-units Drafttube Gates

Figure 1: Upstream view on a Figure 2: HYDROMATRIX®-Module

HYDROMATRIX®-Module /1/ Typical Cross Section /2/

HYDROMATRIX®-modules can be installed in dams, usually in weirs, ship lock

sluices, intake towers for drinking water as well as in intake structures of irrigation
systems. In the usage of existing hydraulic structures these should be suitable or
adaptable for a HYDROMATRIX®-Module installation and a utility grid connection
should also be in close proximity.
In general the plant should have an available discharge of ~100 m3/s (3,500 cfs) and
the gross head should amount from 3 m up to 30 m (10-100 feet) with a minimum
submergence of 1.5 m (5 feet) below tailwater to achieve technically and
economically feasible applications.
However, HYDROMATRIX®-technology is a clean and environmentally friendly
energy production form (KYOTO-protocol). It is an opportunity to utilize unexploited
water resources at competitive costs. It represents a standardized modular concept
with a project schedule from 1.5 to 2 years. The HYDROMATRIX®-module is easy
removable for flood conditions and also for repair and maintenance work.
The investigation presented here deals with the installation of several
HYDROMATRIX®-modules in the tainter gate bays of an existing dam which currently
serves only navigation and flood control. In addition to the generation of electric
energy the modules have to fulfill the function of emergency stoplogs. Hydraulic
experiments /3/ to this project were recently carried out to determine the drag forces
occurring during underflow conditions. Submergence conditions of the draft tubes on
a simple pre-version of the HYDROMATRIX®-module were investigated as well /4/.
The modules are shifted into the stoplog slots of the existing bays (Figure 3). The
turbine-generator units (TG-unit) are arranged in double rows in a module.
Depending on the available module discharge conditions, single turbines can be
opened or closed by a draft tube gate at the two overlapping draft tube outlets. Each
gate is operated by a hydraulic hoist mechanism.
A TG-unit consists of a stay ring with fixed stay vanes, a fixed blade propeller type
runner and an induction type generator directly connected to the turbine runner
(Figure 4).

2
 MODULE CRANE

WINCH

LIFTED
MODULE

LOW STEEL
359ft

TAINTER GATE OPEN

UPPER POOL
324ft

FLOOD WATER

TAINTER GATE CLOSED

TWL
HYDOMATRIX- 302ft
MODULE

SILL CREST
290 ft

Figure 3: HYDROMATRIX®-module placed Figure 4:Turbine generator unit /2/

in a tainter gate bay

Total maximum efficiency is about 85%, this is approximately 8% lower than a

conventional bulb turbine can achieve (93%). In comparison to a "Compact- Hydro"
solution, energy production of unregulated turbine units is lower. On the one hand
this draw back is balanced to some extent by a low demand for revision and thus to a
high availability of the TG-sets due to a simple design not susceptible to failure. On
the other hand lower energy production can be partially compensated using a large
number of turbine units in a module, depending on discharge.

Aim of Investigation
The purpose of a diffusor formed draft tube is to recover a maximum amount of
kinetic energy from the water leaving the runner, thus decreasing the dynamic draft
head. The shape of the draft tube is normally formed by a transition of the circle
profile of the runner plain to a rectangular outlet profile. The vortex flow of the runner
has an essential influence on the velocity and pressure distribution and therefore on
the recovery of pressure head in the draft tube.

To prevent unfavorable flow conditions in the draft tube, a certain maximum angle of
the diffusor should not be exceeded and furthermore a sufficient submergence of the
draft tube exit is necessary. The last task was to find stable pressure conditions in
the superimposed draft tubes in the following investigations carried out at the
Laboratory (HYDROLAB) of the Institute of Hydraulic Engineering. In the case of very
low tailwater conditions full operation of both turbine rows can lead to air-entrainment
in the upper draft tube from downstream which decreases the performance of the
turbine significantly.

3
Experimental Investigations

In order to substantiate the hydraulic behavior of the HYDROMATRIX®-turbine under

various aspects three different models were built to investigate specific flow cases. A
full model was built to a scale of 1: 40 to show the approaching flow conditions of the
turbines during various operational cases. Further, the turbine outflow conditions – in
particular minimum submergence of the draft tube - with corresponding energy
losses were examined at a six-pack model and at a pier model to a scale of 1:12.5.
All tests were carried out according to the Froude´s law of similarity /5/.

Full Bay Model 1:40

In order to study the global flow approaching the modules, a full -model (Figure 5) to
a scale of 1: 40 has been defined, consisting of one full bay plus two halfs of a bay at
both side of it.

281,27
3
39.92
37.6
3

74.83

Figure 5: views of the full bay model, cross section and layout of the module

4
The main purpose of this test was to make sure that the water is approaching the
modules without detrimental disturbances and vortices. Nevertheless, random
surface vortices pulling dye for a short time to the turbine intakes were observed
(Figure 6). More detailed studies to this subject will be carried out on a larger model
size later.

Figure 6: Random coherent swirl throughout water column (dye core) as well as
random surfaces dimples upstream of the module

These tests were carried out under various inflow conditions, such as one single
module in operation, all units in operation and so far. Special attention has been paid
to the boundary units closed to the piers or to a closed turbine unit (Figure 7).

Figure 7: Approaching flow conditions to different operating turbine units

In principle no significantly negative consequences in terms of a reduced turbine

efficiency induced by different possible inflow conditions to the turbines was
observed. The adjacent to a closed section located turbines will have an estimated
average efficiency loss of about 5%.

5
Six-Pack Model 1:12.5

The six–pack model fabricated in plexiglass presents an inner section of the module
and consists of three pairs of turbine units. The module was fixed in the middle of a
44 cm wide flume where the flow could be observed through glass panels (Figure 8).

Figure 8: Test rig of the six-pack model

The water levels up- and downstream of the module were registered simultaneously
by inductive pressure gauges. Measurement of the flow discharges was taken by
means of a magnetic inductive flow meter. In contrast to a former investigation /4/ the
revolutions were held constant by means of an electrical motor (generator) at the top
of the module box where the turbine power was transferred by means of 6 gear
shafts. Measurements of revolutions were taken by a fibre optic sensor at the motor
(generator) shaft.

reservoir level 324 ft

Upper Side
flap gate
Left Middle Right

Left Right
Top Top
Upper Draft
Middle Tube Middle
cable line
Bottom Bottom
intake
apron
k
trash rac

TWL 302 ft
Top Top
draft tube gate

TG-unit Lower Draft

draft tube
Middle Tube Middle

Bottom Bottom
TG-unit
cable line draft tube
sill crest 290 ft Weir Crest

bottom seal

Figure 9: Cross section, measurement points view from downstream

6
Furthermore, pressure heads on the middle draft tube column were recorded at 15
measuring points on a cross section 2 cm close to the end section of the draft tubes
(Figure 9). These pressures were transmitted by means of plastic hoses to
piezometer tubes. Measurements should reveal the energy losses, the net head and
the dynamic draft head as a function of the over all operational gross heads of the
plant. The upstream water level was held constant (normal water level), tailwater was
varied to allow measurement of different flow conditions. The main investigations
concerns flow cases at different discharges as follows:
• All draft tube gates fully open
• All draft tube gates partially lowered 15 mm
• In addition, other specific but not critical cases were tested, but these results are
not presented below.

All draft tube gates fully open

The most important aspect of these tests was to determine the recovery of kinetic
energy downstream of the draft tube exit with a specific emphasis on flow stability.
These experiments were carried out by increasing tailwater what results in a turbine
discharge decrease. Flow separation (air-entrainment) in the draft tube at very low
tailwater levels decrease the turbine efficiency significantly. At increasing tailwater
levels (TWL) the turbine flow stabilize at a boundary tailwater level (Figure 10).

Figure 10: Flow separation in the upper draft tube, boundary submergence

It is obvious that the pertinent pressure zone in the upper draft tube is situated in the
middle of the upper draft tube edge. If the pressure drops in this region below the
atmospheric pressure (in this case below the upper edge of the draft tube), air can
back draft into the draft tube. To prevent flow separation with air- entrainment in the
upper draft tube row a submergence of approximately 1 ft (about 25 to 30 cm) is
desirable for a stable turbine operation (Figure 11).

Figure 11: Stable flow conditions at a minimum submergence

7
Increasing TWL lead to smoother tailwater surfaces and also to a reduced turbine
discharge. Correspondingly the reduction of the draft tube velocities is directly visible
in a smoother course of pressure heads. In this case a better average velocity profile
is developed, which reduces recirculation flows in the draft tube edges and improves
the draft tube efficiency as well.

305,00

304,50

304,00

303,13
303,50
pressure head [ft]

303,00

301,90
301,80

301,80

301,80
301,74

301,69
301,59
302,50

301,64

301,49
301,44
301,33
301,28

302,00

301,28
301,18

301,13

301,06
301,50

301,00

300,50

300,00
left

right

upper
middle

middle

middle

middle

middle

TWL
bottom

bottom

bottom

bottom
top

top

top

top

edge
upper side left side left side right side right side
upper draft upper draft lower draft upper draft lower draft
tube tube tube tube tube

Figure 12: Pressure heads on the draft tube outflow section (Q=109 l/s)

Figure 12 reveals that the net head of the upper turbine row is slightly higher than the
lower one. This difference will be compensated at higher TWL. A comparison of TWL
and draft tube heads in relation to different discharges is depicted in Figure 13.

310

309
left
308 tailwater
middle
307 right
pressure head [ft]

upper edge
306 TWL

305

304

303

302
upper edge draft tube
301

300
103 104 105 106 107 108 109
turbine discharge [l/s]

Figure 13: Draft tube heads and TWL in relation to turbine discharges

8
All draft tube gates partially lowered 15 mm
A slight lowering of the draft tube gates produce an increase of pressure in the upper
region of the upper draft tube. Therefore, it is possible to stabilize turbine flow in this
endangered flow separation zone by means of this measure (Figure 14).

Figure 14: Stable flow conditions of the upper draft tube row at TWL 302 ft

In this way the lowest TWL 302 ft of the project can be achieved. Figure 15
demonstrates the effect of a lowered draft tube gate on the pressure conditions in the
draft tubes. In comparison to Figure 12 a distinct increase of the pressure heads in
the upper draft tube region is evident. Consequently, this measure leads to an
enlargement of the lower turbine row net heads.

305,00

304,50

304,00

303,50
pressure head [ft]

303,00
302,10
302,00

302,00
301,90

302,50
301,55

302,00
301,15

301,05

301,06
301,00
300,80

301,50
300,65
300,60

300,60
300,60

300,55
300,50
300,40

301,00

300,50

300,00
left

right

upper
middle

middle

middle

middle

middle

TWL
bottom

bottom

bottom

bottom
top

top

top

top

edge

upper side left side left side right side right side
upper draft upper draft lower draft upper draft lower draft
tube tube tube tube tube

Figure 15: Pressure heads on the draft tubes outflow section,

draft tube gate 15 mm lowered (Q=109 l/s)

Correspondingly, the performance of the generator – especially of the lower TG-unit

row – must be dimensioned to these extreme conditions. In any case the transition
zone between the upper draft tube edge and gate must be tight and hinder an
entrainment of air which could lead to a drop in pressure.

9
Pier Model 1:12.5
The pier model (Figure 16) is a rebuilt section of the module with 1 vertical pair of
boundary units, the roller box of the module and one half of a pier dividing the bays of
the dam. The model was also fixed in the same flume where the six-pack model was
investigated. The mechanical and measurement equipment was fabricated and
mounted corresponding to the six-pack model above. The turbine model was
manufactured by laser lithography fully homologous to the proposed prototype unit.

weir pillar

15.2
44.0

flume roller box

trash rack

13.0
draft tube

Figure 16: Test rig and layout of the pier model

It is obvious that the boundary machines will behave differently compared to the units
in the middle of a module. To avoid flow separation at a minimum TWL of 302 ft it
was also necessary to lower the draft tube gate of about 15 mm (Figure 17). In
general the model tests revealed a similar behaviour of the turbine flow as on the six-
pack model (Figure 18 and Figure 19). Only impacts on the turbine outflow conditions
due to flow recirculation caused by the offset of the roller box was expected

Figure 17: Stable flow conditions at a minimum submergence with draft tube gate
fully open and draft tube gate 15 mm lowered (TWL 302 ft)

Another important objective of this investigation was to find out unsatisfactory

approaching flow conditions to the boundary turbine in terms of possible air
entrainment due to surface vortices and swirl effects. But no shortcomings to this
concern occurred in these model tests.

10
 305,00
Upper Side
304,50
Left Middl Right
304,00

302,87
303,50
Right
Top Side
pressure head [ft]

Upper
303,00 Draft Tube

302,05

301,90
Middle

301,95

301,79
301,69
302,50

301,8
301,49
301,44
Bottom
301,33

301,18
302,00

301,50
Top
301,00
Lower
Middle
300,50 Draft Tube
Bottom
300,00
left

right

upper
middle

middle

middle

TWL
bottom

bottom
top

top

edge
Weir Crest

upper side right side right side Glass Panel

of the Flume
upper draft tube upper draft tube lower draft tube

Figure 18: Pressure heads on the draft tubes outflow section (Q=36,5 l/s)

305,00
304,50
304,00
303,03

303,50
pressure head [ft]

302,34

302,34

303,00 301,89
302

302,50
301,49

301,49

301,18
302,00
300,86
300,92

300,75

301,50
301,00
300,50
300,00
left

right

upper
middle

middle

middle

TWL
bottom

bottom
top

top

edge

upper side right side right side

upper draft tube upper draft tube lower draft tube

Figure 19: Pressure heads on the draft tubes outflow section,

draft tube gate 15 mm lowered (Q=36,5 l/s)

Summary

In order to evaluate the hydraulic flow conditions of a Hydromatrix®-turbine specific

model tests were carried out at the University of Technology Vienna. General and
extraordinary flow conditions were investigated on three different model types. Under
full operation of all turbines the pressure heads in the draft tubes vary in correlation
to the TWL. If the TWL drops below 304 ft, flow separation in the upper draft tube row
is in all probability. Stable turbine flow conditions can be guaranteed at a TWL of 305
ft which corresponds to a necessary submergence of the upper draft tube row of
about 1 ft. The net heads of the upper and lower draft tube are in the same range
during full turbine operation. An operation of all turbines during lower tailwater
conditions than the above mentioned is possible by lowering of the draft tube gate of

11
about 20 cm into the upper draft tube section of the prototype. This leads to
moderate differences in the net head between both turbine rows.
Detrimental flow conditions in the vicinity of a pier or a closed turbine to an operating
unit have not a significant effect on the turbine efficiency. One or two boundary
turbine units will be influenced approximately 3%-7% of the calculated efficiency.
Stable air entraining surface vortices upstream of the module could not be observed
in these tests, but nevertheless there is a certain remaining risk. Further
investigations focusing on this concern will be carried out on a combined pier and six-
pack model at a later date.

References

/1/ www.vatech.at: Products & Services, HYDROMATRIX™, 2001

/2/ VA Tech Hydro: HYDROMATRIX™ – Product Information, pp. 1 - 27

/3/ Prenner, R. et.al.: A new Concept in Hydropower – Hydromatrix-Technology: Drag

Forces on an Advanced HYDROMATRIX®-Module, IAHR

/4/ Prenner, R. et.al.: Submergence of Draft Tubes in a HYDROMATRIX®-Module,

Waterpower XIII

/5/ Kobus, H.: Wasserbauliches Versuchswesen, Mitteilungsheft Nr. 4,

Deutscher Verband für Wasserwirtschaft, 1978

Authors

Reinhard Prenner, M.Sc., Ph.D., is Project Manager at the Institute of Hydraulic

Engineering, Vienna University of Technology, Austria. He has been involved
in hydraulics research and hydraulic constructions since his graduation from
the University in Vienna in 1981. After that he obtained a Ph.D. degree in
Fluid Mechanics and Hydraulic Structures. For the last few years he has
worked in the field of pressure surges, fluvial hydraulics, flow induced
vibrations, swirling flow problems and hydrodynamic forces.

Volker Kienberger, M.Sc., Ph.D., is a member of System Analysis, Engineering

Services of VA TECH HYDRO GmbH & Co, Linz, Austria. He works on the
design of penstocks and hydraulic steel structures, especially of
HYDROMATRIX®-Modules. In 1999 he obtained his Ph.D. degree at the
Institute of Hydraulic Engineering, Vienna University of Technology, in the
field of hydrodynamic forces at high-head leaf-gates.

Helmut Drobir M.Sc., Ph.D., is Professor at the Institute of Hydraulic Engineering,

Vienna University of Technology, Austria. Before his call to the University in
1992 he had been joined for more than 30 years in private companies
including VOEST and Tyrolean Water Power Company.

Alois Nichtawitz, M.Sc., is Head of Hydraulic Laboratory, research and developement

at the department for Large Hydro, VA TECH HYDRO GmbH & Co, Linz,
Austria.

12