DE69400296T2 - Method and device for measuring the flow in partially or completely filled pipes and in open channels - Google Patents

Description

The invention relates to a method and a device for determining the volume flow (flow) of a liquid that flows turbulently through a channel or a channel that can be completely or only partially filled.

Certain methods and devices for measuring volume flow are described in the following documents:

EP-A-0278671 describes a system in which the flow velocity is measured at a single point in the flow of the return system of a drilling rig, simultaneously with a height measurement of the material. These signals are sent by means of microprocessors 41, 45 via an interface 47 to a central computer 14, which stores the various parameters of the line and calculates the flow based on the measurements and the stored parameters. In the case of a floating drilling rig, the measurements must be repeated to obtain instantaneous values of the flow and these are averaged over time (see page 15, left column).

The abstract of Japanese Patent Vol 15, no 462 (P-1279) 22.Nov.1991 & JP-A-031 97822 shows a method in which measured values of the velocity at many points in a pipe are entered into a computer programmed with a model of the pipe. The computer takes into account the turbulence in the pipe to calculate individual flow velocities at the points for display.

The abstract of the Japanese patent Vol 8, no 72 (P-265) 4 April 1984 & JP- A-58218616 applies to a speed measurement at a point of the flow and a water level measurement. DE-A-293501 5 applies to a water level measurement and the storage of a characteristic curve of a river and the water level in a computer. Speeds do not appear to be measured.

Other previously known methods and systems are described below in Chapter 3.

The object of the invention is to provide an improved method and device which avoids extensive field calibrations, which allows a determination based on a minimum of field measurements, but which at the same time provides increased accuracy.

Subject of the invention

This object is achieved by the method according to claim 1 and the device according to claim 21.

The invention is described with reference to the following figures:

FIG. 1: Schematic representation of the flow measuring system CSLME

FIG. 2: Velocity distributions in a trapezoidal channel for two different flows (Source: Varwick, F.: On the flow formula for open artificial channels, dissertation TH Dresden, Dresden 1944)

FIG. 3: Example sensor arrangement in a circular profile in longitudinal and cross section

FIG. 4: Alternative sensor arrangement in open channels

One embodiment of the invention comprises the CSLME measuring system ("Central Simulation - Local Measurement and Evaluation") for determining the flow of fluids in partially or fully filled pipelines and in open channels (hereinafter referred to as flow when partially filled). The CSLME system is a technique for precise flow measurement using line-averaged or point-like velocity measurements. No calibration is required using velocity measurements with network-like measuring points; the CSLME system is based on the numerical simulation of turbulent velocity distributions in any cross-section. The CSLME system is shown schematically in FIG. 1 and consists of

a central simulation system CS, which contains a numerical model and a powerful computer

local measurement and evaluation systems LME, which contain devices for measuring the fluid level or pressure head and speeds, including the devices for evaluating measured values and

Data storage devices as an interface between CS and LME systems.

Details on the technical implementation of these components are given in Chapter 5. The term "fluid level" defines the phase boundary between the fluid whose flow is to be measured and an overlying fluid with a much lower density. For the sake of simplicity, the term fluid level is replaced in the following by the more descriptive term "water level". The term "water level" is also used when the flow is not to be determined for water.

uses moving parts not installed in the flow and

requires speed measurements.

A further classification is possible with regard to the location of the speed measurements:

2.2.1. Measurement of the average cross-sectional velocity

These are integral measuring methods that directly deliver the area-averaged flow velocity in the cross-section as a measuring signal. Electromagnetic flow measuring systems are used here, in which a precisely defined magnetic field is generated in the discharge cross-section. The area-wide velocity and thus also the flow can be calculated directly from the measuring voltage.

The advantage of this measuring method is that the flow measurement is largely independent of the speed distribution in the cross-section. Therefore, the requirements for uniform flow and positioning longitudinally in the pipe or channel are very low. The achievable accuracy of this flow measuring method is high. The prerequisite is a minimum conductivity of the fluid, which is not the case with petrol or petroleum, for example. The disadvantages are the high measurement complexity of the device and the strong dependence on the profile geometry. For each profile type and each profile size, the design and construction of separate measuring tubes or measuring channels is necessary. For large profiles, especially for open channels, the technical complexity for generating the magnetic field and installing the measuring device increases significantly.

2.2.2. Line-averaged or point-based speed measurement

In these flow measurement methods, the flow velocity is measured either as an average along a defined measuring section within the flow cross-section (“line method”) or as a point value at a precisely defined location within the cross-section (small control volume). The average cross-sectional velocity is determined from this line-averaged or point-like velocity measurement. In contrast to the previously mentioned measurement methods, these methods basically require a suitable evaluation method and the velocity distribution in the cross-section must be known.

The disadvantage of these methods is that accurate flow measurements are only possible if the velocity distribution in the cross-section is known with sufficient accuracy. If there is little knowledge or none at all that can be determined from the line-averaged or point-like velocity measurements, the accuracy is very low. The advantage, however, is that the measurement effort required for a line-averaged and especially a point-like velocity measurement is much lower compared to methods in which measurements of the average cross-sectional velocities are used. In particular, with point-like velocity measurements, the effort required for the device is almost independent of the profile size and it is easy to apply the method to different profile types and profile sizes.

3. Differences between the CSLME system and existing flow measurement methods with line-averaged or point-based velocity measurement. 3.1. Methods limited to completely filled channels and using measurements of the velocity distribution

Up to now, flow measurement methods based on line-averaged or point-shaped velocity measurements have mainly been used only in full pipes. For example, a measuring method is known from US 5,186,150 which is only applicable to full pipes and in which point-shaped velocity measurements are used. The method according to US 5,186,150 requires that for each calculation of the flow the entire velocity distribution within the cross-section is measured, which is achieved by repeating many point-shaped velocity measurements at different locations in the cross-section. In contrast, the new flow measurement method CSLME is not limited to full pipes and the determination of the velocity distribution in the cross-section does not require repeating velocity measurements at different locations in the cross-section. In contrast, the CSLME method only requires the measurement of the water level and a single local velocity at any location in the cross-section. These two measurements are sufficient to calculate the discharge very accurately when using the CSLME method. This will be explained later.

3.2. Methods not restricted to fully filled pipes and not using velocity distribution measurements

Some other flow measurement methods are known that are only limited to fully filled pipes, but do not require velocity measurements at many different locations in the cross-section. Instead of measuring the velocity distribution, these methods assume a velocity distribution that corresponds to a stationary, uniform flow. Therefore, when using this method, the measurement of the velocity distribution within the cross-section is replaced by the assumption of a well-defined flow state. However, the assumed approximately stationary, uniform flow only exists in pipe sections in which changes in the velocity distribution along the flow direction are very small. In particular, immediately upstream of the measuring station, disturbances to the velocity distribution (caused, for example, by bends, drops, inlets or outlets, etc.) should be avoided. The disadvantage of these methods is that longer straight pipe or channel sections are required upstream and downstream of the measuring stations, in which no disturbances of the velocity distributions are permitted. Even for stationary uniform flow at full filling, knowledge of the turbulent velocity distributions is usually only available with sufficient accuracy for circular profiles. In contrast, very little is known about the turbulent velocity distributions, especially for non-circular cross-sections of complex geometry, even when it is a stationary uniform flow at full filling. Therefore, the flow measurement methods known to date, which use line-averaged or point-like velocity measurements, are almost completely limited to circular profiles at full filling, unless the entire velocity distribution is directly determined by velocity measurements at different locations in the cross-section.

Flow measurement methods are known for circular pipes that are fully filled. The accuracy of these methods is good enough because in circular profiles the position of the line measurement in the line-averaged velocity measurement method can be selected with the special condition that the value of the line-averaged velocity, independent of the flow, approximately corresponds to the average velocity of the circular pipe cross-section when the flow is approximately stationary and uniform. There are methods that use these effects and select special measuring paths or special locations in the cross-section to measure local velocities. In these special cases, a further evaluation method is unnecessary. Therefore, the methods used so far, which do not use the measurement of the entire velocity distribution in the cross-section, are characterized by the following properties:

Line-averaged speed measurements are preferred.

The position of the speed measurement is precisely defined and is fixed before the measurement.

Only a minimal evaluation of the measured values is carried out.

It is hardly possible to transfer this measuring principle to partially filled channels, since the position of the measuring section along which the average line velocity corresponds to the average cross-sectional velocity is no longer constant, but changes with the water level.

3.3. Methods applicable to flow in partially filled channels using velocity distribution measurements

Flow measurement methods that use line-averaged or point-like velocity measurements are also known for use in open channels. These techniques are based on the ratio of the mean cross-sectional velocity vm to the local velocity vL. This ratio is determined using one or more calibrations. A measurement method of this type is known, for example, from DE4016529C1. During calibration, the velocity distribution is determined by measuring the local velocities at many locations. These locations define the measurement network. This technique is known as "calibration using network-like measuring points".

3.4. Methods applicable to flow in partially filled channels and not using velocity distribution measurements

These methods also require knowledge of the ratio vm/vL, but in contrast to the methods explained in Chapter 3.3, the value Vm/VL is not determined by calibration. For example, a method is known from US4.202,211 in which vm=vL is assumed everywhere, which means vm/vL=1.0. However, as FIG. 2 shows, local velocities vL and mean cross-sectional velocities vm are only equal for certain locations in the cross-section, even if the assumption of a stationary uniform flow is met. Moreover, in contrast to the fully filled circular profile, these locations are not stationary (see Chapter 3.2), but change in particular depending on the water level. In contrast, if the locations at which velocity measurements are carried out remain unchanged, the ratio vm/vL will depend on the water level. However, since in U54,202,211 a constant ratio of vm/vL=1.0 is assumed independent of the water level and no recommendation is given for variable positions of the local velocity measurements, the accuracy of this flow measurement method is only low.

The new flow measurement method CSLME does not have this disadvantage. CSLME is a new flow measurement method that can be applied to fully filled pipelines and to partially filled channels with the same accuracy using line-averaged or point-shaped velocity measurements. CSLME does not require measurements of velocity distributions. The key characteristic of CSLME is that neither calibration nor measurement with grid-like measuring points is required. The measurement of a single local velocity at any location in the cross-section is sufficient to determine the flow. With CSLME, the ratio vmlvl is not determined by calibrations using grid-like measurements, but by a numerical simulation of the turbulent velocity distribution, which depends on the characteristic hydraulic parameters water depth, profile type, gradient, etc.

Like the methods explained in Chapter 3.2, the new CSLME method assumes almost uniform flow and achieves comparable accuracies. However, while the high accuracies of the previous methods according to 3.2 are only limited to completely filled, circular pipes, the CSLME method guarantees the same level of accuracy for profiles of any shape. and CSLME is applicable both to flow in pipelines at full fill and in channels at partial fill.

The new CSLME flow measurement method is characterized by the following properties:

Local speeds are measured either at a point or along a line.

The location of the speed measurement does not have to be known before the measurement and can be variable.

The location of the speed measurement can be subject to random fluctuations. In this case, the location is an additional measurement value.

The assignment of the local measured value VL to the mean cross-sectional velocity vm is carried out by numerical simulation.

Numerical simulation, the use of a computer and the digital storage of calculated velocity distributions on a data storage device are additional components of the CSLME flow measurement method.

The free choice of location for local velocity measurement enables a simple transfer of the measuring method to partially filled pipes and channels.

The CSLME method is largely independent of the cross-sectional shape.

By using numerical simulation in the CSLME method, not only is the achievable accuracy of the flow measurement increased, but also enables a significant reduction in the technical effort and at the same time the costs of the flow measuring device are reduced.

The methods used to date normally prefer line-averaged measurements. In some of these cases, a measurement evaluation can be dispensed with if the measuring sections are specially arranged. Without the use of a measurement evaluation, higher accuracy of flow measurement could be achieved with the known line-averaged speed measuring methods than with the known point-based speed measuring methods. However, measuring the average speed along a measuring section requires the installation of a transmitter and a receiver. In the case of partially filled pipes and channels, these must be arranged on different sides of the cross-section, since the measuring section must not intersect the water level.

Particularly in the case of wide channels, the measuring section must be approximately parallel to the water level. If the water level fluctuates greatly, this immediately leads to the next measurement-related problem: if the measuring section is too high, it will dry out at low water levels and thus become unusable. If it is too low, however, there will be large deviations between the average flow velocity along this measuring section and the average cross-sectional velocity at high water levels. These speed differences lead to larger errors in the flow determination if they are not corrected by an evaluation procedure. Alternatively, not just one but several measuring sections, which are installed at different heights, are used. However, this means a further increase in the measurement-related effort and thus an increase in the cost of the measurement procedure.

In the CSLME measuring method with point-based speed measurement, however, the transmitting and receiving sensors can be integrated in one housing. The measuring method does not work in transmission but rather backscattering. In open channels, the sensors do not need to be arranged laterally, but can be installed on the bottom. The measuring beam does not run almost parallel, but approximately perpendicular to the water level. This means that when the channel is partially filled, only one measuring sensor is required to record point-based speeds for the entire range of water level fluctuations. The measuring effort for a point measurement from the channel bottom is therefore less than that for line-averaged measurements. In contrast to other methods, however, the CSLME method does not lead to a reduction in the accuracy of the flow measurement when this simplified point-based measuring method is used. The reason for the high accuracy of the CSLME method is the complex evaluation method, with which it is possible to calculate the average cross-sectional speed from any local speed with high accuracy.

However, point-shaped speed measurement is not a prerequisite for the CSLME method. Instead, line-averaged speed measurements can be carried out along defined measuring sections as an alternative or in combination. For this purpose, transmitters and receivers operating in transmission are installed on opposite sides of the cross-section.

These sensors work in transmission. The average cross-sectional speed is calculated using the CSLME method from the measured line-averaged speed, similar to point-based speed measurements.

4. Basics of the invention

As shown in Chapter 3, the accuracy of flow measurement in the case of partial fillings can be increased if the measurement data acquisition is supplemented by a high-performance evaluation system. This even makes it possible to reduce the measurement effort. The evaluation system has the task of determining the ratio of local velocity Vp to the average cross-sectional velocity vm for a measuring point or a measuring section P. vp can be a point-like or a line-averaged velocity.

Using the simulation and evaluation systems integrated in the CSLME process (see FIG. 1), this ratio can be determined with high accuracy for any water level and any profile shape. The following principles are used:

4.1. The numerical model

Stationary uniform turbulent velocity distributions are calculated numerically with high accuracy for any profile with partial filling (see FIG. 2) for given water level positions and longitudinal gradients. The simulation is based on a fine discretization of the discharge cross-section with a numerical mesh and the application of the finite element method or a finite difference method. On the basis of this numerical mesh, the velocity distribution is calculated in such a way that the balances for continuity and momentum are fulfilled with high accuracy in all three coordinate directions. The distribution of the turbulent viscosity required for this is determined numerically using a suitable turbulence model. In the numerical simulation, the secondary flows in particular must be taken into account.

The numerical model is used to solve the differential equations for the flow numerically at partial filling. This procedure leads to large systems of equations for the variables of the differential equations, which are defined for each node of the numerical network. As a solution to these systems of equations, which contain several hundred or even several thousand unknowns, the desired velocities are calculated for each node of the numerical network. Based on the values at the nodes, velocities can be calculated at any position within the cross-section. In particular, the ratio vp/vm between any local velocity and the average cross-sectional velocity can be calculated for any position of the location P.

The new flow measurement method CSLME therefore replaces the calibration based on measurement networks with a simulation based on a numerical network.

4.2. The accuracy of the flow measurement

The numerical model is characterized by the fact that the accuracy of the simulation can be increased by refining the numerical network, which increases the computational effort. With sufficient network refinement, the accuracy of the CSLME flow measurement method, even in the case of partial filling, will only depend on the accuracy with which water levels and local velocities can be measured.

4.3. Measuring the water level

At least one, but usually two, sensors SW are used to measure the water level yw (see FIG. 3). For symmetrical profiles (circle, egg, trapezoid...) the sensors should also be arranged symmetrically.

The SW sensors are characterized by the fact that they

do not cause any disturbance to the flow,

have a high measurement accuracy,

allow measurement values to be recorded in a very short time (several measurements per second) and

the entire range of values between minimum and maximum water level can be recorded.

4.4. Measuring local speeds

At least one, usually two, or more SV sensors are used to measure the local speeds. For symmetrical profiles, these sensors should also be arranged symmetrically. The SV sensors are characterized by the fact that they

do not cause any disturbance to the flow,

measure the speed vp either in a very small control volume (point measurement at location P) or as a line average along a defined measurement distance (line measurement along line P),

not only measure the local velocity vp at point P, but also the cross-sectional coordinates xp, yp of the measuring point P for each measurement (see FIG. 3), if the point-based measuring method is used and not the special method according to Chapter 4.5,

for point measurements, the location of the measurement can vary for each measurement if the point measurement method is used

have a high measurement accuracy,

allow measurement values to be recorded in a very short time (several measurements per second) and

Can measure speeds in the entire value range between minimum and maximum water level.

4.5 Special method of point-shaped speed measurement with measurement of the maximum speed

With this special procedure, a measurement of the cross-sectional coordinates xp, yp of the measuring point P of the local speed measurement is not necessary. However, the other requirements according to Chapter 4.4 also apply to this special procedure. In addition, with this special procedure, it must be required that the speed sensors are installed in such a way that they can be used to measure speeds for all variations in the water level in areas of the profile in which the maximum profile speed is only very slightly exceeded. These areas are determined using numerical flow simulation before the sensors are installed.

A large number of local speeds are measured using speed sensors. The maximum speed vmax results from the evaluation of the signals received. Using the hydrodynamic equations contained in the CSLME method, the average cross-sectional speed vm is calculated on the basis of the maximum speed Vmax. It is therefore characteristic of this special method to replace the information about the local coordinates xp, yp of the measurement location with the corresponding property of the maximum speed vmax.

5. The technical design of the CSLME flow measurement system

The flow measurement system CSLME (see FIG. 1) consists of a central simulation system CS and local measurement and evaluation systems LME. The interface Suitable data carriers are formed between the two subsystems, which can be written to by the CS subsystem and read by the LME subsystem. The central simulation system CS only exists once and can be installed separately from the local LME measurement and evaluation systems. The LME subsystems, on the other hand, are installed at each measuring point on site and exist in multiples. The individual components of the system are explained below.

5.1. The central simulation system CS

The CS is the core of the flow measurement system. It consists of an extremely complex numerical model and a very powerful computer. Stationary uniform velocity distributions are simulated, which are characterized by the following four features:

1. Profile type

The profile type characterizes the type of geometric shape. Possible profile types include circle, rectangle, trapezoid, egg or mouth. In addition, it is also possible to define a profile only by its polygonal perimeter. Such an application can be interesting for natural rivers or canals, for example. For hydraulically rough profiles, it may be necessary to define different profile types for different roughnesses.

2. Profile size

The geometric dimensions are varied for each profile type. For example, the diameter is changed for a circle or the base width is changed for a rectangle.

3. Water level yw

The velocity distributions are calculated for each profile type and the different profile sizes for different water levels yw.

4. Longitudinal gradient I

The gradient of the pipe or channel in the direction of flow is decisive for the fluid acceleration and thus the magnitude of the velocities in the cross-section. Therefore, the longitudinal gradient is also varied.

For the very large number of possible combinations of the four characteristics mentioned, the speed distributions are simulated once each using the numerical flow and turbulence model. The speed distributions are then known across the entire area at every point in a cross-section and are stored on suitable data carriers together with the flow rate that can be determined from them and the four characteristics. A separate data carrier can be written for each profile type or even for each profile type and each profile size. The content of the data carriers is profile-dependent in this case. Magnetic tapes, floppy disks or compact disks can be used as data carriers. The complex simulation system, which requires a very powerful computer, results in the following advantages:

The costs of measurement in the local measurement and evaluation systems LME can be reduced considerably and

In contrast to the data carriers, the LME are largely profile-independent.

5.2. The local measurement data acquisition of the LME system

The local measurement and evaluation system LME (see FIG. 1) consists of

SV sensors for measuring local speeds,

Sensors SW for measuring the water level (or the pressure head when fully filled and

a microprocessor that converts the measurement signals from the sensors SV and SW into the required value of flow Q using the velocity distributions stored on the data carrier.

FIG. 3 shows a possible sensor arrangement for point-shaped speed measurements using a circular profile as an example, which meets the requirements of Chapter 4. One SV sensor and one SW sensor are arranged in a common housing. This combination sensor can be designed to be screwed in and in this case only requires a suitable profile hole on site. Two of these combination sensors are installed in a symmetrical arrangement on the profile base. On the one hand, the sensors should be installed as deep as possible in order to be able to detect to be able to carry out flow measurements at low water levels. On the other hand, they should be installed high enough so that they cannot become silted up.

5.2.1. SV sensors for measuring local speeds

Local speeds can be measured advantageously with ultrasonic sensors. For point-based speed measurements, sensors are used that work according to the pulse Doppler method. This method requires that the smallest sound-scattering particles are carried along in the fluid. This requirement is usually met. The SV sensors work in backscattering, so the transmitting and receiving oscillators can be arranged in the same housing. The transmitting oscillator sends out ultrasonic beams that are received by the scattering particles and scattered back to the receiving oscillator of the probe. The measurable frequency difference (Doppler frequency) between the transmitting and receiving frequencies is a measure of the local speed vp. The method according to Chapter 4.4 also requires the measurement of the location P of the speed measurement.

In order to measure the location P, the pulse transit time from the sensor to the scattered particle and back is measured using the pulse Doppler method. This transit time is a measure of the distance of the scattered particle from the sensor. Measuring this distance is sufficient to calculate the location of the measuring point P(xp, yp) (see FIG. 3), since the direction of the ultrasound beam is known.

The ultrasonic pulses sent by the transmitting transducer usually encounter a large number of sound-scattering particles on their way towards the water surface, which partially scatter this beam. Therefore, many ultrasonic signals also reach the receiving transducer. However, since these signals are received with a time delay, each of these signals can be assigned a local flow velocity vp and also a distance of the measuring point P from the ultrasonic sensor. Each ultrasonic signal therefore provides the cross-sectional coordinates xp, yp of the respective points P. By emitting just one ultrasonic pulse, the local velocities can be measured and differentiated at many positions P on the beam.

Nevertheless, the pulse Doppler method is in principle a method for point-based velocity measurement, because from a pulse evaluation one obtains many discrete measurement triplets xp, yp, vp along a line. In contrast, the evaluation of an ultrasonic pulse in the line-averaged speed measurement method only produces one measured value, which represents the average speed along the measuring section. Since the entire pulse evaluation takes place at the speed of sound, it is possible to repeat the pulse Doppler measurement very frequently in a very short time. The result of this process is a measured speed profile along the ultrasonic beam between the SV sensor and the water level with high local resolution.

However, with the point and line-averaged measuring methods used to date, only a small fraction of the extensive measurement data that the pulse Doppler method can deliver within a very short time could be used. This limitation of the known methods is based on the fact that a sufficiently accurate assignment of the local speed vp at any point in the cross-section to the average speed vm when partially filled was not possible. For the CSLME flow measurement method, however, this sensor is ideal because the large amount of measurement data can be fully utilized due to the simulated speed distributions. It follows that the use of the CSLME method enables a much more accurate determination of the flow with the same measurement effort.

An alternative option for ultrasonic measurement is the laser Doppler anemometer (LDA) using backscattering. The requirements of Chapter 4 can also be met using this optical method. The particularly advantageous features of this method are the very high measurement accuracy, high spatial resolution and the ability to record measured values in a very short time. The LDA method requires the presence of light-scattering particles in the fluid and requires the fluid to be optically transparent. Disadvantages are currently the high cost of purchasing and operating this measuring system.

In line-averaged speed measurements, the average speed is measured along a defined measuring path. Ultrasonic sensors are also advantageously used for this. However, these methods work according to the time-of-flight method and are installed as transmitting and receiving sensors at the opposite ends of the measuring path.

5.2.2. Sensors SW for measuring the water level or pressure head

Known pressure cells that work according to a piezoresistive, capacitive or inductive method can be used as SW sensors for measuring the water level or pressure head, both when the tank is partially or fully filled. Pressure cells are technically sophisticated sensors that have proven themselves in a wide range of practical applications as accurate, fast and robust measuring sensors and do not cause any disruption to the flow.

For open channels or only partially filled pipes, the water level can alternatively be measured using ultrasonic sensors. In principle, the sensor can be mounted above the maximum possible water level or on the profile base.

FIG.4 shows an example of a sensor arrangement for a trapezoidal channel, in which the water level is measured using ultrasonic sensors SW. The sensors SW are mounted on a measuring bridge and remain dry. The water level is measured from above. These sensors therefore work without contact and cannot be damaged by the fluid flow. The local speeds are measured by separate speed sensors SV, which are installed on the channel bottom as in FIG. 3. This arrangement of the ultrasonic sensors is particularly suitable when no separate measuring bridge has to be erected, but when there are system components on site to which the sensors SW can be attached.

A water level measurement using ultrasound can be carried out particularly easily if, as recommended in Chapter 5.2.1, the local velocities vp are measured using the pulse Doppler method. If the speed sensors SV are installed on the profile base and their ultrasonic beams are directed almost perpendicularly to the water level, a separate sensor SW for measuring the water level can be completely dispensed with. Instead, the water level is measured when measuring the local velocities using the pulse Doppler method. This is another advantage of the CSLME flow measurement method, which enables an effective evaluation of the measurement signals of the pulse Doppler method to determine the flow (see Chapter 5.3).

As explained in Chapter 5.2.1, the velocity measurement of the pulse Doppler method is based on the backscattering of ultrasonic pulses from sound-scattering Particles that must be contained in the fluid flow. However, such backscattering also occurs in an extreme manner at the free water surface due to the large density differences of the fluids at the phase boundary. The water surface therefore causes a particularly clear backscattering of the ultrasonic waves. The distance of the water surface from the speed sensor SV and thus the height of the water surface is determined from the measured pulse travel times from the sensor SV to the water surface and back again.

5.3. The local measurement value evaluation of the LME system

With the help of the local measurement value evaluation of the LME system, the flow rate Q is calculated from the measurement signals of the SV and SW sensors (see Chapter 5.2). For this purpose, the data carriers written by the CS system and a microprocessor are used (see FIG. 1). Depending on the type of local speed measurement used, the evaluation is carried out according to one of the following methods or a combination of these:

5.3.1. Point-shaped speed measurement with location measurement xp, yp

1. At least one sensor SW is assigned to each sensor SV in such a way that for each point-like local speed measurement at the measuring point P the four measured values yw, xp, yp and vp are available for further measured value conversion.

2. The flow measuring point and thus also the LME are defined by a certain profile type and a profile size. Together with the measured water level (or the pressure head when fully filled) yw, three of the four characteristics of the velocity distribution explained in Chapter 5.1 are thus known. The fourth characteristic is the gradient 1 in the direction of flow. However, this gradient cannot be measured with reasonable effort and sufficient accuracy. Therefore, the value for 1 is replaced by the local velocity vp at the measuring point P, which can be measured much more easily and more accurately. One of the velocity distributions stored on the data carrier is therefore selected with the help of the microprocessor. This selected velocity distribution fulfills the first three characteristics mentioned and its velocity at the measuring point xp, yp corresponds to the measured value Vp. The flow Q assigned to this selected velocity distribution and also stored on the data carrier can therefore be assigned to the local Speed measurement at measuring point P and is designated Qp. The search algorithm shown here in a somewhat abbreviated form can be refined using a suitable interpolation algorithm. The determination of Qp requires only a small amount of computing power from the microprocessor and can be carried out in a very short time.

3. Each point-like local speed measurement at the measuring point P is assigned a flow rate Qp in the manner shown using the data carrier and the microprocessor. This measurement value conversion is repeated for each measuring point P and each speed sensor. The required flow rate Q is ultimately the mean value of the individual flows Qp, whereby a suitable weighting of the individual flows can be taken into account. It is particularly important to note that only a single local speed at any location P of the measuring cross-section is required to determine the flow rate Q. Repeating the measurement is not a prerequisite for the CSLME method. However, repetition is a simple method for significantly increasing the accuracy of the measurement overall. Since the accuracy of the flow measurement can be increased by increasing the number of measuring points, the fastest possible sensors SV should be used in the CSLME method. This requirement is met by acoustic or optical sensors.

5.3.2 Point-shaped speed measurement with measurement of the maximum speed according to the special procedure in chapter 4.5

1. The measurement signals from the SW sensors are averaged and give the average measured water level yw (or the pressure head in pipes when fully filled). In this special procedure, this value Yw and the measured maximum speed Vmax form the basis for determining the flow rate Q.

2. Analogous to the procedure described in Chapter 5.3.1, one of the velocity distributions stored on the data carrier is selected using the microprocessor. The profile type and profile size of the selected velocity distribution correspond to the current flow measuring point and its water level yw (or pressure head in pipes when full) and maximum velocity vmax correspond best to the measured values. The value assigned to this velocity distribution and also The flow rate Q stored on the data carrier is simultaneously the flow rate Q sought. In this special procedure, the multiple measurement of the location of the velocity measurement is therefore replaced by the characteristic property of the velocity maximum, which in an analogous manner enables a clear identification of the relevant velocity profile of the numerical simulation, compared to the procedure according to Chapter 5.3.1.

5.3.3. Line-averaged speed measurement

1. If local speed measurements are carried out that provide average values along one or more measuring sections, then, in contrast to the procedure according to 5.3.1, the location of the measurement is determined by the positions of the sensors. Therefore, these positions are known and no longer need to be measured. At least one sensor SW is assigned to each sensor SV in such a way that for each line-averaged local speed measurement along the measuring section P, the two measured values yw and vp are available for further measurement value conversion.

2. Analogous to the procedure described in chapters 5.3.1 and 5.3.2, one of the velocity distributions stored on the data carrier is selected using the microprocessor. The profile type and profile size of the selected velocity distribution correspond to the current flow measuring point and its water level (or pressure head in pipes when fully filled) and average line velocity along the same measuring section P correspond best to the measured values. The flow Q assigned to this selected velocity distribution and also stored on the data carrier can thus be assigned to the local velocity measurement along the measuring section P and is designated Qp.

3. Each line-averaged local speed measurement along the measuring section P is assigned a flow rate Qp using the data carrier and the microprocessor in the manner shown. This measurement value conversion is repeated for each local measurement and each measuring section. The required flow rate Q is obtained analogously to the procedure described in Chapter 5.3.1 as the mean value of the individual flow rates Qp, whereby a suitable weighting of the individual flow rates can be taken into account. Regarding the repetition of the measurements, please refer to Chapter 5.3.1.

6. Main features and preferred applications of the CSLME flow measurement system

Finally, the essential features of the invention are briefly summarized and preferred applications are mentioned. The flow measurement system CSLME uses numerical fluid mechanics methods to simulate fluid flows and uses devices to measure water levels (or pressure heads) and local velocities. These methods and devices are considered in isolation and are known for the application purpose mentioned. However, the CSLME combines them in the manner shown for the purpose of determining the flow rate when the system is partially filled. In this special combination and its application to partial filling in open channels, it represents a new method for flow measurement. The key features of the CSLME are as follows:

Division into a central simulation system (CS), local measurement and evaluation systems (LME) and data storage devices

Loss-free flow measurement method without flow fittings or mechanically moving parts

Local velocity measurements using fast sensors that can be installed on the pipe or channel wall

Possibility of assigning local flow velocity at any position within the cross-section and flow velocity averaged over the cross-section by using numerical fluid mechanics

Possibility of assigning maximum flow velocity within the cross-section and flow velocity averaged over the cross-section by using numerical fluid mechanics

Reduction of (continuous) measurement effort in the LME by increasing (one-time) simulation effort in the CS

Especially when using the pulse Doppler method to measure local velocities, a significant increase in accuracy can be achieved compared to other methods, despite significantly reduced measurement effort.

The LME system is largely independent of the profile and therefore a high level of standardization can be achieved.

The CSLME flow measurement system can be used for many applications of flow measurement in partial filling. Due to the above-mentioned process characteristics of the CSLME, there are particular advantages over other processes under the following conditions:

for non-circular profiles and channels

for large profiles (e.g. natural waters and canals)

as a temporary or mobile measuring device (e.g. for control measurements)

for measuring points that are difficult to access or remote (e.g. existing sewer systems, pipelines)

for fluids with low electrical conductivity (e.g. petrol, crude oil)

for fluids containing solids (e.g. wastewater, streams and rivers during flooding)

Claims (32)

1. A method for determining the volume flow of a liquid flowing turbulently in a channel with a known cross-section, which includes measuring the liquid level and measuring an axial flow velocity in the channel and determining the volume flow of the liquid in the channel based on the measurement of the liquid level and the measurement of the axial flow velocity, said method comprising a) simulating a series of turbulent velocity distributions for the said channel on the basis of a numerical mesh of the said cross-section for one or more different liquid levels selected in such a way that the whole range of liquid level fluctuations occurring in the said profile during actual flow conditions is uniformly covered by the said selected liquid levels, and a numerical model simulating turbulence-driven secondary flows in the said cross-section, (b) measuring an axial flow velocity (vp) within said known cross-section to provide a measured axial flow velocity (vp) which is either - (i) an axial velocity within a very small discharge volume located at a measured spatial position (P) within said profile or - (ii) the line-averaged axial velocity (vp) along a known line (P) between two points within said profile or - (iii) the maximum axial velocity (vmax) occurring anywhere within the mentioned profile, is, c) measuring the liquid level (yw) in the channel and d) determining the volume flow rate (Q) in said channel on the basis of said measured axial flow velocity (vp) and a selected one of said simulated velocity distributions which is uniquely selected from said series of simulated velocity distributions on the basis of said measured liquid level (yw). 2. A method as in claim 1, wherein said volume flow is determined using a combination of measurements (i), (ii) and/or (iii). 3. Method as in one of claims 1 and 2, comprising improving the accuracy in said determination of the volume flow (Q) by repeating steps (b), (c), (d) of the method defined in claim 1 and calculating the resulting volume flow (Q) as a weighted average of the individual determinations. 4. Method as in one of claims 1 to 3, wherein the volume flow is determined by an arrangement of a central simulation system (CS), local measurement and evaluation systems (LME) and data storage devices (FIG. 1). 5. A method as in any one of claims 1 to 4, wherein said simulation step is performed in said central simulation system (CS). 6. A method as in any one of claims 1 to 5, wherein said data storage devices store data provided by said central simulation system (CS). 7. Method as in any one of claims 1 to 6, wherein said measurements are provided by said local measurement and evaluation system (LME) which comprises at least one sensor (SV) providing at least one said measured axial velocity, at least one sensor (SW) providing at least one measurement of the liquid level in the channel and a microprocessor which transforms the signals measured by means of the sensors (SV and SW) into a value of the volume flow (Q) using data provided by said central siren system (CS). 8. A method as in claim 7, wherein the measurement of the liquid level in the channel is provided by two sensors (SW). 9. A method as in claim 8, wherein said sensors (SW) provide measurements over the entire range of liquid levels between the lowest and highest values. 10. A method as in claim 7, wherein said measured axial flow velocities are provided by a plurality of sensors (SV). 11. Method as in claim 7, wherein each sensor (SV) measures a local flow velocity (vp) within a very small discharge volume located at a position (P) within the cross section of said profile and the cross-sectional coordinates (xp, yp) of said position (P) are measured simultaneously. 12. A method as in any one of claims 7, 10 and 11, wherein the installation of said sensors (SV) allows a variation of said position (P) for each measured axial velocity and wherein said sensors (SV) provide velocity measurements over the entire range of liquid levels between the lowest and highest values. 13. Method as in one of claims 7 to 12, comprising the assignment of at least one sensor for measuring the liquid level (SW) to each sensor for measuring the speed (SV). 14. A method as in any one of claims 1 to 13, comprising ultrasonic sensors installed at the channel bottom and providing measurements of axial flow velocities by applying the Doppler method based on the backscatter principle. 15. A method as in any one of claims 1 to 13, comprising laser Doppler anemometers (LDA) that provide measurements of axial flow velocities using backscatter. 16. Method as in one of claims 7 to 13, comprising the assignment of at least one liquid level sensor (SW) to each speed sensor (SV) in such a way that for each local speed measurement at each sensor (SV) the four measured values: mean value of the measured liquid level (yw), cross-sectional coordinates (xp, yp) of said position (P) and said local flow velocity (vp) at said position (P) are provided and transformed into a volume flow (Q) with the aid of said microprocessor. 17. A method as in claim 1, comprising measuring the maximum axial velocity (vmax) as defined in step (b)(iii) of claim 1 without measuring the position at which said maximum axial velocity (vmax) occurs and comparing this measured maximum velocity with the simulated maximum velocity of the simulated series of velocity distributions as defined in step (a) of claim 1 and on the basis of this comparison determining the volumetric flow rate (Q). 18. Method as in claim 17, comprising speed measurements based on the ultrasonic Doppler method and comprising the selection of the largest value of all measured speeds and comprising a sensor installation which allows, independently of the liquid level, measurements of local speeds in the areas of the cross-section in which the maximum speed (vmax) approximately occurs. 19. A method as in claim 1, wherein the line-averaged axial velocity (vp) defined in step (b) (ii) of claim 1 is measured along a predefined line (P) within the cross-section and this measured line-averaged axial velocity is compared with the simulated line-averaged value along the same line as defined in step (d) of claim 1. 20. A method as in claim 19, wherein said line-averaged axial velocity (vp) is measured using the ultrasonic time-of-flight method. 21. Apparatus for determining the volume flow of a liquid flowing turbulently in a channel with a known cross-section, comprising a central simulation system (CS) in which a numerical model is programmed and which simulates a series of turbulent velocity distributions for said channel on the basis of data such as the cross-sectional size and shape of said known profile and the hydraulic configurations of said liquid and said channel and on the basis of a numerical network of said cross-section, said programmed central simulation system (CS) simulating turbulence-driven secondary flows within said known cross-section, said apparatus further comprising a data storage device storing said series of simulated velocity distributions generated by said central simulation system (CS) and a local measurement and evaluation system (LME) comprising devices of measuring sensors which enable a measurement of the liquid level (yw) and provide a measurement of the axial flow velocity (vp) in the channel, said measured axial flow velocity (vp) being one of the alternatives defined in step (b) of claim 1, and comprising a microprocessor which determines the volume flow of the liquid in the channel on the basis of the measured liquid level (yw) and the measured axial flow velocity (yp) provided by the means of measuring sensors and a selected one of said velocity distributions stored in said data storage device which forms the interface between said central simulation system (CS) and said local measuring and evaluation system (LME). 22. Apparatus as in claim 21, comprising means for associating said measured axial flow velocity (vp) with the area averaged axial flow velocity (Vm) in said known channel. 23. Device as in one of claims 21 to 22, wherein the central simulation system (CS) is installed separately from the local measurement and evaluation systems (LME) which are installed in the measuring stations. 24. Device as in one of claims 21 to 23, wherein the local measuring and evaluation system (LME - FIG. 1) comprises at least one sensor (SV) providing at least one measurement of an axial flow velocity, at least one sensor (SW) providing at least one measurement of the liquid level and a microprocessor which transforms the signals measured by the sensors (SV and SW) into a volume flow (Q) with the aid of data provided by said central simulation system (CS). 25. Device as in one of claims 21 to 24, wherein a sensor for measuring the liquid level (SW) and a sensor for measuring the axial flow velocity (SV) are installed in a common housing as part of said local measuring and evaluation system (LME). 26. Device as in any one of claims 21 to 25, wherein said sensors are installed near the channel bottom to measure volumetric flows even at low levels of the liquid surface, but sufficiently high to avoid siltation. 27. Device as in claim 26, wherein said sensors are ultrasonic sensors providing measurements of the axial flow velocities and, in the case of point measurements, operate according to the pulse Doppler method based on the backscatter principle and comprise a combined sensor acting as transmitter and receiver within a common housing, said combined sensor providing the measurements required for determining the measuring point position (xp, yp) which can be calculated because the direction of the ultrasonic beam is known (FIG. 3). 28. Device according to one of claims 21 to 24, comprising ultrasonic sensors which provide measurements of the axial flow velocities and which, in the case of line-averaged measurements, operate according to the ultrasonic time-of-flight method based on transmission. 29. Device according to one of claims 21 to 28, comprising sensors (SW) that provide measurements of the liquid level (SW) or the pressure head and comprising pressure cells that operate according to piezoresistive, capacitive or inductive methods. 30. Device according to one of claims 21 to 29, wherein the channel is open and the sensors (SW) provide measurements of the liquid level and comprise ultrasonic sensors installed above the maximum possible liquid level or at the profile bottom. 31. Apparatus as in claim 30, comprising said liquid level sensors (SW) installed dry on a measuring bridge and measuring the liquid level from above and velocity sensors (SV) providing measurements of the axial flow velocity and installed on the channel bottom (FIG. 4). 32.Device according to one of claims 21 to 31, wherein such measuring devices are used which combine the measurement of the axial flow velocity (vp) and the measurement of the liquid level (yw) within a common sensor and said common sensor is an ultrasonic sensor which is installed on the channel bottom in such a way that measurements of the axial flow velocities are provided and the pulse Doppler method based on the backscatter principle is applied.