Hydraulic Structure for Water Flow Control

The present invention relates to hydraulic structures according to the preamble of claim 1. It further relates to arrangements of more than one such structure.

For controlling water flow, weirs and other obstacles are well known. US-3 593 527 proposes a hydraulic structure by which flow features can be converted, e.g. from a deep, narrow channel to a wide, flat bed, in reducing scour and maintaining surface levels. The principle of the layout is called MEL (Minimal Loss Structure) .

In prior constructions of drop structures is the problem that even though the water around weirs can often appear relatively calm, they can be extremely dangerous places to boat, swim, or wade as the circulation patterns on the downstream side — typically called "hydraulics" — can submerge a person indefinitely. This phenomenon is so well- known to canoeists, kayakers, and others who spend time on rivers that they even have a rueful name for weirs:

"drowning machines". The lack of such a horizontal eddy in this design solves also these problems, and many others like, the fact that these weir used before can become a point where garbage and other debris accumulate. Submerging as it is known in "drowning machines" will never occur with the invented structure.

Therefore it is an object of the present invention to propose a hydraulic structure with reduced erosion effects and a less dangerous flow dynamics.

Such a hydraulic structure is defined in claim 1. The further claims define preferred embodiments and arrangements of such hydraulic structures . The present invention relates to a drop structure including a Minimum Energy Loss (MEL) structure as shown in U.S. Pat. No. 3,593,527 to bring down the water level and generate controlled vortexes by a sudden change of a flow form after the structure. Thereby, the energy line is brought down in a way where the majority of energy is dissipated in eddies on the water away from the structure itself and also away from natural ground. A series of relatively calm eddies occurs near the riverbanks and this builds up a relatively calm counter current which largely prevents riverbank erosion, and therefore the banks can be constructed with easily erodible and thus more nature-like materials . The structure as described above has the fully functionality in the high flows as it is, and is very useful in places where the peak flow energy of water is the main problem, i.e. Spillways flood channels or chutes with a relative high head.

If the river is wanted to have balanced bed load

transportation also in the normal and low flow conditions, the energy control of these flows must be achieved with other ways, i.e. with turbines and/or fish ladder. These turbines must generally be able to function with relative low heads. This makes the river morphology to be "as it is in the nature". The nature tries always to correct the flow to a minimum energy loss situation (Froude Number Fr=l). This means that if the flow is subcritical, sedimentation occurs and if it's supercritical, erosion occurs. This balance is to be found in critical flow conditions,

This present design enables flow at maximum designed flood level without any increase on the water surface level, at the structure. In fact, there is normally a decrease of such a level. This structure can function safely without dramatic flooding also with much higher (20%-50%) flows than

designed. And the only possible damage to be awaited is that the surface erosion of the structure itself increases. If the determining flow occurs often, it is therefore wise to design the structure under the critical flow condition, also the useful range for Froude number will thus be i.e. 0.7 - 1.0

GENERAL PRINCIPLES

Gordon Mckay has well explained the general principle in "Introduction" of his book "Design of minimum energy culverts" (October 1971)

For over a hundred years, the basis of channel design has been some form of Chezy's formula, V = CslRS where v is the velocity of flow, R is the hydraulic radius, S is the slope of the energy line and C is a variant, depending on the channel boundary conditions and the state of the flow.

Considerable effort and time have since been spent attempting to define ' C . Much of this effort has been directed to introducing a constant 'n' which is dependent only on the channel boundary.

The overriding condition to the application of such formulae is that they apply only to long uniform channels - without defining 'long' or 'uniform' and steady flow. Computations relating to natural channels have been carried out on a similar basis, the length being restricted to some small length over which the energy slope is considered as constant and parallel to the bed and water surface. There is, obviously, immediate conflict as small lengths and non- uniformity do not attempt to satisfy the basic analytical requirement. In order to obtain sensible answers,

coefficients, somewhat arbitrarily chosen, are applied.

In 1932, Boris Bakmateff introduced the concept of specific energy - the internal energy associated with open channel flow. He showed that for a particular discharge, flow could occur at two depths except for one particular condition - the critical depth. In mild slopes, a small constriction led to a reduction in depth with a corresponding increase in velocity - such that the total internal energy of the flow remained constant. Bakmateff measured this energy from the original channel floor. This concept explained the "hydraulic jump", but apart from using the jump as a mixing facility or to reduce head at a particular place, very little practical use has been made of the concept.

It has always been presumed that the increase in velocity led to an increased 'friction' loss and consequently restrictions which caused such an increase in velocity must lead to an increase in the rate of energy loss and consequently, to be maintained, must require an increase in upstream water level. This would probably be true if no other changes resulted.

Natural streams are invariably non-uniform. It is extremely difficult to measure or to define 'slope' . The cross section changes, often quickly, from point to point. It is often difficult to distinguish a change in section-form and boundary roughness. It has been shown that 'n', originally introduced as a constant, varies radically with stage and from section to section in the same stream. Chow sets out possible variations of n with stage. Figure 1 is a plot of n against stage for the major coastal rivers of Queensland. The basic figures were obtained from the Irrigation and Water Supply Commission. Presumably, they were measured with reasonable accuracy. As the ' n' is a straight multiplier, any computations using a particular n value must be grossly in error .

Overall, the specific energy at a section is given by

±=

B gA ² where A is the cross sectional area of flow (and measured as

defined in fluid mechanics - perpendicular to the velocity) ;

B is the top water width; and

Q is the discharge.

Unless the specific energy is compatible section to section, energy dissipation - turbulence - must occur. Thus energy loss can arise purely from a change of section form - not boundary roughness. At the same time, the specific energy of a vertical element is always where y is the depth of the element and v the depth-average velocity, so that transverse flow is likely from points of deeper depth (high energy) to the shallower depths (low energy) . This transverse flow - again generating turbulence - will dissipate energy. Thus the cross sectional shape can significantly alter the rate of energy dissipation

irrespective of the boundary conditions.

The persistence of turbulence is inter alia a function of eddy size. Small eddies quickly disappear, but large eddies persist for a considerable time and hence distance. Thus small eddies generated at the boundary will dissipate quickly and will have a local effect only. Large eddies generated by form-change or cross section shape will persist downstream and will make the downstream reach apparently 'rough' . The size of eddies alone does not measure the rate of energy dissipation. The number of eddies is equally significant.

However, provided the specific energy is compatible from section to section, considerable changes of section can occur without turbulence developing - the reach can be

comparatively 'smooth'.

It is probable the conditions prevailing in natural streams are such that the form loss dominates the flow pattern and that the boundary 'friction' is a minor part of the total energy loss. The range of 'n' values is comparable to the variation of the drag coefficients of solid bodies in turbulent flow, e.g., a flat disc to a streamlined body of the same diameter (1.0 to 0.1) .

This invention relates to the fact, that the energy

dissipation can be made with a controlled change on a cross section, and therefore the energy is dissipated equal efficiently with various discharges and flow velocities, and always without big problems with the sedimentation or erosion. And therefore the bed load transportation can also be kept in balance in most of the flow situations. With the selected flow, which dimensions the structures, it is possible to have a FR=1 flow through out the channel/river. It should be noted, that the flow actually has virtual boundaries in the vortex area, and this means that the flow is actually critical outside the vortex area, even though the calculations made with real masses doesn't give such a result. This makes a big difference to the old structures, where the hydraulic jump needs a highly erodible FR>5 or eventually FR>9 to an efficient dissipation, or when

overflow dams are used, higher heads are needed to reach the same dissipation efficiency at high flows, and this can't of course be accepted because of the prevented fish migration.

Such a structure for controlling the energy of liquid flow is characterized by a particular relation between head, depth, width and total flow at every cross section

perpendicular to the flow, such as to give a minimum erosion but a balanced bed load transportation, and an efficient, but harmless way to dissipate the energy of water. More particularly, this relation is determined by the MEL principle . The invention will be further explained by preferred exemplary embodiments with reference to the Figures.

FIG 1. a plan of the structure;

FIG 2. a cross-section through the structure of Fig. 1; and FIG 3. a plan from an example of the principles of a

complete system with many different flow control phases .

Flow direction is from left to right.

In the Figures, the following symbols are used:

(Yc-1) water depth upstream the hydraulic structure

(Yc-2) water depth downstream the hydraulic structure (B-l) Width of the flow at water surface with water

depth (Yc-1)

(B-2) Width of the flow at water surface with water

depth (Yc-2)

(H) height of the drop; head

(L) Length between two structures. (L-l) Length of the convergence and the drop structure. (L-2) Length of the vortex area.

(L-3) Length of the riverbed which must take the impulse load of main vortexes [2].

(Yc-1) Critical water depths with (B-l)

(Yc-2) Critical water depths with (B-2)

1) Convergence made between (B-l) and (B-2), calculated with Equation 6, in U.S. Patent 3,593,527 (see

Appendix) ;

2) The approx. description of vortexes (34 pes of

circles ) .

3) Boundary of the vortex area. Notice that the water level will slightly sink in vortex area.

4) The sudden change on a flow form.

5) Represents how the riverbed should be constructed, with convergence 1.

6) Water surface

7) Example of a part of the structure simplified for the purpose of the description. In practice, a rounded transition between e.g. the bank of a river and the structure is preferred which significantly improves the performance of the structure (mainly with respect to maximum flow) . Improvement by a rounded transition may be 5 to 10 %.

8) The point of a possibly needed riverbed supporting structure .

9) "Overflow weir" -dissipation area

10) Height of the overflow weir constructed from 1 and 9.

11) Erosion preventive -counter current

12) "Hydraulic jump" -dissipation area

13) Fish ladder

14) Low/Mid flow channel 15) Fish screen/rake.

16) Turbine outflow guide

17) Turbine intake

18) Turbine

19) Turbine diffusor

PRINCIPLES OF THE DESIGN

The data required for the design are:

a: Discharge flow (Q)

b: Height of the drop, Head (H)

c: Limiting dimensions of the desired structures,

(B-l) Max width of the flow in water surface; and

(L) Distance form drop to drop, or

(I) Energy gradient.

Majority of the calculation is already explained in U.S. Patent 3,593,527 and thus it is not necessary to open it again here. The only new thing is the distance from drop to drop (L) or energy gradient (I=H/L) . Both of these gives the same information in practice. When the distance gets too small, also the energy gradient gets too high and there will be not enough space for the vortexes .

The principles are explained by three example calculations, which simultaneously show that there is always one unique solution to be found with each given data. The minimum length (L) for various heads and energy gradients must be found with detailed model tests. The 1:36 model tests made from a structure exactly as shown in FIG 1 and 2, and described in example I, has shown that the length needed for vortexes (L-2) is approximately at least 1.2 x ([B-1]-[B- 2] ) . Therefore it should be considered that the minimum length (L) for these systems is 25 m with 0.8 m head (H) , and thus the maximum gradient (I) is ca . 0.033 if more complicated dissipating solutions are not accepted. The impulse force of these vortexes must also be calculated or studied and be added to the loads of the structures and river bed. If the riverbed material is not stabile enough to carry this load, extra supporting structures 8 must be added.

Such other dissipating solution will automatically appear behind the (B-2) as a hydraulic jump 12. This jump itself does not either build erosion because it has no contact to riverbanks or structures. Therefore it might be useful in some cases to dimension the (B-2) -part of the structure with mid-flows FR=1 and thus with high flows where FR >1.0. This might give a really sustainable solution to pumped-storage hydroelectricity, where the extreme sudden mid-flow changes needs to be controlled to avoid ecological problems. In these cases the structure must be dimensioned to neutralize the negative effects of these sudden mid-flow changes and therefore the flooding situation must be solved by other ways. It is possible to dimension the height 10 of structure 1 in a way, where it functions as a conventional overflow weir 9. Notable is that this height 10 is not

constant /horizontal . This means that it is eventually possible to dimension this one simple structure 1 to have three phases of dissipation: First are always the eddies 2, second must be the hydraulic jump 12 occurring in area (L-2) between the left and right eddies 2, and the third are overflow weirs 1, 9 in area (L-l) .

It is strongly advised that the calculations and

functionality are proven by model tests, specifically when short (L-l), (L-2) and (L-3) are desired. There is no danger of flooding if these tests are not made and the lengths (L- 2) and (L-3) are later found to be too short to dissipate the energy of the designed flow. This leads only to a higher flow speed in the middle of the flow and thus to a drop on a water level 6. Of course the higher velocity causes surface erosion on the structures and thus the expected lifetime might not be reached. The situation is exactly the same as what happens if the designed flow is increased by a sudden catastrophic flood. When flow increases, the level is of course raised, but as the examples shows, the change is relative small. Example II is practically the same as

Example I but with a 33 % increase in flow. In the examples the water level 6 raises from 1.37 m to 1.65 m, that is 0.28 m, but in reality the raise will be even smaller because the water would flow with higher speed when the design is constructed as in example I.

EXAMPLES OF THIS HIGH FLOW DESIGN.

For ease and clarity of explanation, the examples are given for conditions of flow in rectangular cross sections. The same calculations can be made for any cross-sectional shape (geometric or otherwise), but they generally involve more complex calculations. These complex forms and calculations should eventually always be used, as this too simple example brings problems, for example on the point 7 in the Figures, because the flow velocity and the cross-sectional area (as known in hydraulics) are no more perpendicular. The

principle of dimensioning the structure remains nevertheless the same. An easy solution to solve this problem is to calculate the beginning of the structure instead of linear change of (Yc) with a linear change from (B), or a

combination of these two methods . EXAMPLE I

3

In a rectangular channel or river with Q = 100 ra /s, B = 20 m, head H = 0.8 m and energy gradient I = 0.02 is to be calculated:

3

Q/B= 5.0 m /s per m

The water depth on the beginning of the drop will be as calculated from U.S. Pat 3,593,527, equation (3), and is 1.37 m. The flow velocity at the same point is to be

calculated from equation (3A) and is thus 3.7 m/s .

Now the water level must be dropped by 0.8 m and it can be calculated from equation (4) etc. that the velocity at the end of the drop must be 5.4 m/s, from equation (3A) that the water depth must be 2.97 m, and then from equation (1) it can be calculated that (B) must be 6.25 m at the end of the drop. Now, when the edges are calculated, the change of the form 1 from the start to the end can be calculated with the equation ( 6 ) .

The distance between drop to drop (L) is 0.8/0.02= 40 m and cannot be changed.

EXAMPLE II

3

In a rectangular channel or river with Q= 100 m /s,

B-l = 15 m, head H = 0.8 m and energy gradient I = 0.02, it is calculated:

3

Q/B= 6.7 m /s per m

The water depth on the beginning of drop is 1.65 m. The flow velocity at the same point is 4.0 m/s. Now the water must be dropped by 0.8 m so the velocity at the end of the drop must be 5.7 m/s and the water depth must be 3.25 m, and then again it can be calculated that (B) must be 5.44 m at the end of the drop. When the edges are

calculated, the change of the form 1 from the start to the end can be calculated by equation (6) .

The distance between drop to drop (L) is 0.8/0.02= 40 m and is the same as in Example I, but as seen from Fig. 1, the relative length is higher, the energy dissipation can be awaited to be better than in Example I, and the same flow can be transferred in a smaller space. This is based on the fact that the water depth is higher and therefore the vortexes are also higher and the needed area is smaller.

EXAMPLE III (Two structures in the same cross-section)

3

In a rectangular channel or river with Q= 2 x 50 m /s, B-l = 2 x 10 m, head H = 1.0 m and energy gradient

I = 0.025, it is calculated:

3

Q/B = 5 m /s per m The water depth on the beginning of the drop is 1.37 m. The flow velocity at the same point is 3.7 m/s.

Now the water must be dropped by 1.0 m so the velocity at the end of the drop must be 5.7 m/s and the water depth must be 3.37 m. Then from equation (3) it can be calculated that (B-2) must be at the end of the drop 2 x 2.59 m. Again, when the edges are calculated, the change of the form 1 from the start to the end can be calculated with the Equation (6) . The distance between drop to drop (L) is 1.0/0.025= 40 m, i.e. the same as in Example I, but as seen from Fig. 1, the relative length is higher and the energy dissipation can be awaited to be better than in Example I, even though the amount of the dissipated energy is greater.

PRINCIPLES AND EXAMPLE OF LOW & MEDIUM FLOW DESIGN

The drop structures and the whole channel must be

dimensioned to function with the highest discharge without flooding and damages. These design flows can be 5 - 10 times higher than an average flow, and even 50 times higher than minimum flow. Of course there are also rivers that are complete dry a part of the year. As an example is taken a river in Switzerland, at Engstligen, a feeder river of

Kander, Aare and eventually Rhine. The discharge (Q) is at

3

minimum flow only 1.5 m /s and most of the year it varies

3 3

between 3 to 10 m /s. Median value is 6 m / s . As said, if we want to make the river morphology to be "as it is in the nature", we must have the flow in a Minimum Energy Loss

(MEL) situation, also as a critical flow FR =1. This is practically not possible to attain during minimum flow conditions, but that does not really matter because then erosion does not happen either. More important is that the flow condition meets the requirements in those over 200 days

3

a year when the flow is 3 - 10 m /s. As an Example (Fig 3.) we have two hydroelectric machines 15-19 which can work from 3 3

1 m /s - 3 m /s and therefore dissipate all the energy from

3 3

flows 2 m /s - 6 m /s, and a fish ladder 13 with

3

approximately 1 m / s .

In Figure 3 is shown a combination where this critical flow

3

is reached with a total flow of 5.7 m /s and a maximum water depth of 0.47 m, the fish ladder (fish pass) 13 is 1.2 m wide, average flow velocity is V= 1.8 m/s, B-l = 9.7 m, and the cross-section area is A= 3.172 m . 5.1 m /s is turbined to electricity (converted into electric energy by a

hydroelectric device, e.g. a turbine).

3

With 3 m /s the Froude number is 0.84, the max water depth

3 3

is -0.38 m, and 2.6 m /s is turbined. With 10 m /s the

Froude number is 1.2, the max. water depth is -0.57 m, and 6 3

m /s are turbined. If such a turbine combination is combined with a drop structure as explained in Example I, then in theory, the

3

maximal flow with FR=1 is approx. 107 m /s, and the flow will remain near Fr=l on all various discharges and

therefore the river morphology remains always "as it is in the nature", even though we have a fully constructed river with hydroelectricity and flood protection. To bring the best stability to the morphology at low and medium flows, there will be other structures needed to keep the flow in FR=1 conditions also between the structures in area (L-3). This is easily achieved by constructing a dividing

underwater structure 8 where openings to channels 14 are dimensioned for FR=1. This will render the (L-3) -part of the river safe from extreme morphology changes, as these changes concentrate to the (L2)-part of the river. Yet it is to be noted that behind the convergence 1 it is possible to build up a really calm water area if the convergence 1 is

constructed only as a relative thin wall and not filled with stones . This calm water area gives to the river life a shelter in all, even very extreme flow conditions. It also gives shelter against cold weather because it will quickly build up a layer of ice. SUMMARY

With the principles explained, it is possible to build a hydraulic structure in a river, where the water flows with near balanced energy in almost any flow situation, and therefore the river remains as a natural living environment even though the flow is forced to a narrow, fully build up space. In order to fully achieve this goal, the flow must be controlled in different phases. It must be separately decided if only the main phase is to be used, and what is the determining flow in this phase.

The main phase is the structure which builds these vertical vortexes. It can be dimensioned with only one particular flow. Flows smaller than this must be controlled with turbines and/or fish ladders, but it is also possible to build two or even more differently dimensioned structures side by side as a matrix to achieve this goal. If this structure itself is not optimized to the flooding situation, a part of the energy of this high flow can be controlled with a planned hydraulic jump and/or a combined drop

structure as described above.

Further aspects of the preferred embodiments are:

1. The approximately vertical vortexes created by the

structure by a controlled change on a cross section of flow efficiently dissipate energy and build up a calm counter current on the riverbank area, and therefore the riverbanks and the riverbed can be lighter and thus more nature-like constructed between the structures .

2. The structure can give a more balanced bed load

transportation condition, and when correctly

dimensioned regarding the lengths (L-2) and (L-3), the vortexes can build up even a riverbed area suitable for fish spawning. One such area is located near the transition from (L-2) to (L-3). Here, radical

morphology changes do not exist, yet the riverbed remains under a constant but small movement.

3. The structure's flow newer creates a powerful

submerging phenomenon known as a "Drowning machine" . Thereby, risks for people like swimmers and canoeists are effectively reduced. 4. The structure, when it is built only as a thin wall, can give a calm shelter to river life in all flow conditions. This shelter will never be covered with sediment, and it will never dry out. In cold season it will freeze easily and therefore it gives also

protection against extreme cold temperatures.

5. The structure is build up as a matrix of two or even more such structures set up side by side, and/or are overlapped, but only dimensioned differently and/or built to different heights to yield a wider ranging flow control than can be attained by only one of such a structure .

6. The structure can be dimensioned so that the highest occurring flows build up a hydraulic jump after the structure, i.e. another phase of energy dissipation. 7. The structure's height can be dimensioned so that if a greater than dimensioned flow occurs, the structure will function also as a conventional overflow weir providing still another phase of energy dissipation. 8. A structure, wherein also the low flows where these vortexes can't function any more can be kept in

critical conditions with a combination of hydroelectric machines dissipating the energy by conversion into electricity, and a fish ladder. Therefore, the bed load transportation can remain on balance in the drop structure area where it is transported with only a part of the water, and also between the drop structures where it is transported with the total water amount. 9. The trash rack of turbines are taking the water near the water surface in normal flow conditions, and these racks are flushed with a relatively high water amount preventing the sediment and thus also the riverbed-life to be sucked into the turbines. 10. The wall defining the water passage may be

provided only on one side, with the other side being constituted by e.g. a river bank, i.e. the water borders present without the hydraulic structure.

11. The wall defining the passage may be partly

constituted by other means than a wall in the strict sense, e.g. stones in a heap. Important is the presence of at least one edge at the end of the water passage where the vortices build up.

12. The edges for provoking the vertical vortices may be of a large range of angles. Most effectively, however, are angles of at least 90°. With this angles, and in particular with sharp angles, the vortices can build up more easily, in particular may extend on the back side of the wall. Sharper angles create more - In

effective vortices which dissipate the flow energy on a reduced length (L-2) . A very effective means is a blade like element, e.g. a steel blade, so that the angle even approaches 180°.

Appendix

Equations of US-3593527

Yc = q /g (3)

1/2

c = (g y _c ) (3a;

2

y + (v / 2g) = H _s (4)

1/2 3/2.

B >= Q / [g {(2/3) (H« ) } (6)

with

y: depth of flow

Yc: depth of flow according to MEL

q: rate of flow per unit width (Q/B)

v: velocity of the flow

v _c : velocity of the flow at y _c

B: width of the flow at the water surface

Q: rate of flow

g: gravitational acceleration

H _s : energy above the water level

h: height of structure above the bed.