Conductivity
measurement of fish and instrumentation
of swim tunnel.
Conductivity
measurement of fish and instrumentation
of swim tunnel.
Master of
Science Physics –Industrial Instrumentation
proposal for
Endre Grimsbø
Print friendly version in
pdf-format
Supervisors:
Associate Professor Bjørn Tore Hjertaker and Professor Ole Brix
Introduction
This master project consists
of two parts
(i) instrumentation of a swim
tunnel and
(ii) measurements of
conductivity in living fish.
The instrumentation of the
swim tunnel is implemented not only with the intention of measuring
conductivity in fish, but also to make efficient use of the tunnel as a
respirometer.
The experimental results of
the conductivity measurements will be used in the development of a Fish Flow
Meter by Christian Michelsen Research AS (CMR).
The model for conductivity measurements developed in this project is also
expected to be used in upcoming projects on conductivity measurements of living
fish.
The basic principle of CMR’s
Fish Flow Meter is to use conductivity measurements (s) to calculate the volume of fish passing through a pipeline.
Conductivity measurements is a well-developed measurement principle within
multiphase flow in the oil and gas industry.
Multiphase flow is a flow consisting of different components, e.g. oil,
gas and water.
Conductivity measurements
The use of conductivity for
measurement of fish is a transfer of technology from the oil and gas industry.
In order to better understand the background of the technology an introduction
to the measurement methods used in the oil and gas industry is useful as a starting
point.
Figure 1. Cross section of
pipeline filled with two component flow consisting of oil and saline water
In figure 1 the two phases,
crude oil and saline water, are separated and the measurement principle shown
is based on measurement of electrical capacitance to determine the electrical
permittivity ε of the two
component flow.
In order to denote multiphase
hydrocarbon flow consisting of oil, gas and water, it is common to use the
following abbreviations:
α - The oil volume fraction of
the total flow
β - The water volume fraction
of the total flow
γ - The gas volume fraction of
the total flow
Assuming the total volume of
measurement volume is unity and that the flow only consists of oil, water and
gas α+β+γ=1.
Oil and gas have low
conductivity compare to water, especially saline water which has relatively
high conductivity. As a result of this the previous mentioned permittivity
measurement based on electrical capacitance could only performed if the oil and
gas fractions of the flow prevents the saline water from short-circuiting the
electrodes. Therefore, it is important to know the distribution and behavior of
each component in the measurement volume. An issue when monitoring multiphase
flow in a pipeline is to determine which of the different phases is continuous.
It is obvious that it is only
possible to use the capacitance principle for measuring in oil or gas
continuous phase. In water continuous multiphase flow the electrodes will be
short-circuited due to the ions in non-distilled water. Due to the conductivity
of the water component, measurements based on conductivity instead of
electrical capacitance has to be used in a water continuous multiphase flow.
Figure 2. Flow of water with gas
bubbles in a pipeline
A flow of water with gas
bubbles in a pipeline is also called a two phase water continuous flow.
In order to calculate the
mixture permittivity and conductivity of two flowing phases, expressions has
been developed by e.g. Maxwell and Bruggeman. These formulae assume that the
two different phases are homogeneously mixed. Since water has a relatively high
conductivity, a commonly used expression is the formulae developed by Ramu and
Rao.
The basis for Ramu and Rao’s
equations are that σ1 << σ2 and ε1 << ε2, which is true for a
two component flow contains of water and oil/gas. When this parameter is
fulfilled, and the continuous phase consists of component one, then the
equations are as follow:
In the equation σm and εm is the total the conductivity
and permittivity of the mixture and b is the water volume fraction
often called water cut.
When the continuous phase
consists of component two the following equation has to be used:
The tag used at σ’m and ε’m indicates
that this is the total the conductivity and permittivity of the mixture where
continuous phase consists of the second component and b is none continuous.
Connection between hydrocarbon
flow and conductivity measurement in fish
The starting point was that we
wanted to measure the volume of fish in a pipeline.
Figure 3. Cross sectional area of a pipeline with a
fish
It is now possible to look at
the pipeline in figure 3 as a water continuous two component flow, and the fish
is like a gas or oil bubble in a water continuous flow. By measuring the cross
sectional areas it is possible to describe the total conductivity of the two
components as follows:
σtot = σwater + σfish
It is now obvious that a flow
of fish and water is much the same as a multiphase flow of water and oil.
Conductivity measurements of
living fish has not been done previously, only measurements of dead fish. A
significant part of the master project will therefore be to find a suitable
model for conductivity measurement of a living organism that not necessarily
wants to be measured.
Figure 4. Principle of the conductivity measurement
The principle of the
conductivity measurement that is expected to be used is named a “four-terminal
conductivity meter”. The measurement principle is to send an electrical current
through the water and measure the voltage drop over the voltage
electrodes. From the voltage drop it is
possible to calculate the conductivity. The measurement taken does not
represent the cross sectional area but the entire volume between the voltage
electrodes. Two measurements have to be
taken, one with fish in the swim tunnel and one without fish. Based on the
difference between the measurements it is possible to calculate the
conductivity of the fish itself.
It is also necessary to
calculate the volume of the fish between the electrodes. Therefore it is
required to find a suitable method to calculate the volume of living fish. In
previous experiments this volume measurements, related to the conductivity
calculation, have only been done with dead fish. Again, the challenge probably
will be to measure a fish which does not necessarily wants to be measured.
Probably some kind of anesthetic has to be used in the experiment just so that
the fish relaxes when the measurements are done.
The conductivity in the water
is strongly dependent on two factors, temperature and salinity. It is therefore
important that the temperature is stable under the entire experiment, and an
efficient temperature stabilization is needed. Temperature stabilization will
be done by the flow tunnel control unit and a cooling/heating system. The
cooling/heating system consists of a pump which pumps water from the flow
tunnel through to a cooling/heating loop. The cooling loop has to be placed in
a cooler or a heater dependent of the wanted effect.
The conductivity in seawater
is normally 5 S/m and will vary with salinity. Hopefully the salinity of the seawater
delivered from the supply system in the laboratory is stable. The salt water
delivered by the supply system is taken from the fjord outside the laboratory,
and hopefully there are no or insignificant variations in salinity.
Figure 5. Conductivity in Siemens
pr. meter versus salt in mg pr. liter water
Figure 5 is taken from a
previous master thesis written by Anders Nødseth. The figure shows the
relations between salinity pr mg/l in water and conductivity. Before the
experiment starts it is necessary to verify that variations in salinity are
insignificant and therefore can be neglected.
Figure 6. Conductivity in Siemens
pr. meter versus temperature
Figure 6 is also taken from a
previous master degree thesis written by Nils Bjarte Dyrøy. The figure shows
the variations in conductivity caused by temperature variations.
In order to avoid fluctuations
in the conductivity it is important to have a stable temperature. Variations in
temperature will give an unwanted uncertainty in the conductivity measurements
and has to be avoided.
The control unit used in this
experiment must include a control algorithm for stabilizing the temperature in
the water. It is expected that an algorithm based on a simple proportional
controller (P-controller) is adequate for this experiment. If the temperature
stabilization is insufficient with the P-controller more advanced controllers
can be used like PI or PID-controller. The control algorithm will be
implemented in a Compact FieldPoint (cFP) unit used for controlling the flow
tunnel and will be discussed later in this document. The programming tool used
for configuration of the unit is called LabVIEW and is a graphical programming
language from National Instruments.
Swim tunnel as
respiriometer
To get a better understanding
of what a respiriomer is it is useful to take a look at gas exchange in fish in
general.
Fish live in a dense and very
viscous environment, which for a terrestrial animal would appear to be almost
depleted of oxygen. At best, one volume of water contains a little less than 4%
the amount of oxygen present in the same volume of air. In air the solubility
of a gas would roughly be 1000 mL/101.3 kPa = 9.87 mL Gas.L-1.kPa-1
(at 15°C). In practice we treat the
respiratory gases as ideal gases, which mean that O2 and CO2
would dissolve more or less equally in air. In water, however, the solubility
of gases depends on:
·
The nature of the gas
·
The pressure of the gas in the gas phase
·
The temperature
·
The presence of other solutes
Thus, while the solubility of
O2 in water is only about 1/30 the solubility in air, the solubility
of CO2 is more or less the same in air and water depending on
temperature (similar at about 15°C). For calculating gas fluxes
we use a concept which gives the following three fundamental gas equations:
is the volume of gas taken up (O2) or
produced/released (CO2). 
is the flow of medium over the gills. 
is the
capacitance coefficient (solubility) of the gas in the medium. P is the partial pressure of gas (i: before gills; e: after gills). From these expressions, we may calculate:
RQ is
the respiratory quotient (R for
exchange ratio is used when examining isolated parts of the fish). Rearranging
the formulas and ignoring Pi
we get:
If the fish extracts all
the oxygen in the water (Pi @ 21 kPa, Pe
= 0 kPa), the CO2 partial pressure can, at a maximum, be
about 0.7 kPa in the exhalent water. This means the maximum partial pressure in
the arterial (post branchial) blood will not exceed this value, which is almost
a factor 10 lower than for air breathing animals.
This low oxygen availability
has undoubtedly contributed to the evolutionary development of:
·
Large gill surface areas for extremely efficient gas exchange.
·
Developments of air breathing organs and the necessary circulatory
arrangements.
·
Acid-base regulation depending more on ion than ventilatory exchanges.
To measure the gas exchange in
fish it is common to use a respiriometer. There are several different types of
respiriometers based on different principles or systems.
Figure 7 Closed
system respiriometer.
Let’s have a look at the type
called closed system, shown in figure 7. The construction may vary depending on
the material used. There may either a continuous registration of oxygen partial
pressure by an internal sensor or more primitively water may be sampled (and
replaced) for O2 external analysis. It is important that the sensor
is kept at the same temperature as the fish. If the fish is sluggish or we
measure the consumption of fish eggs O2 gradients around the fish or
eggs may build up, and we need to stir the water either by a magnetic stirrer
or a small pump (the later can be combined with an external sensor). We also
need to optimize the size of the chamber to prevent long experimental runs with
building up of organic waist products from the fish, which may influence the
metabolic rate (i.e. carbon dioxide). For measuring the metabolism gas tight
glass syringes have often been used. The closed principle is also used in
activity studies in the so called tunnel respirometer.
Due to the measuring principle
in the closed system we will never be able to measure the oxygen consumption at
a constant oxygen partial pressure. In order to obtain a proper precision of
measurement we need a decrease in oxygen partial pressure of minimum 0.7 kPa.
However, this technique is very useful to obtain the critical oxygen tension,
Pc, at which the fish is no longer able to regulate its oxygen consumption in
relation to the decrease in ambient oxygen partial pressure (see figure next
page).
Figure 8. The
different types of metabolism in relation to environmental oxygen partial
pressure.
The method is very simple and
the precision of measurement is very high. At the same time we can reduce the
response time to a minimum. In addition to the problems regarding accumulation
of organic waist products we also have to take into account the metabolic state
of the fish, since the measurements starts immediately after transferring the
fish to the respirometer. We will thus never obtain the so called standard
metabolism but rather a more active metabolism which we call the routine
metabolism.
Figure 9. The open respirometer
The open – or flow through
system. This system has been preferred by many because it is easy to control
the water quality. This feature also allows for acclimation of the fish prior
to measurements allowing it to calm down. As for the closed system the
respirometer consist of a gas tight Perspex housing. There is a whole in both
ends allowing a constant flow of water through the respirometer. The flow is
either controlled by a leveling system or by a pump. The only draw back of the
system is a ‘time lack’ of response caused by a combination of flow, chamber
size, metabolism and oxygen gradients in the system. As a result we don’t get a
100% registration of all changes in metabolism at any time. On the other hand
we are able to do long term runs without accumulation products and with full
control of physical parameters in the water.
For both systems it’s
important that we use a homogeneous material that has been acclimated to the
proper condition without feeding 24 hr prior to experiment. Further more any
disturbances of the fish should be avoided during experiment.
Figure 10 Stop-flow
respiriometer.
Finally the principle used in
this experiment called the stop-flow system, shown in figure 10. This system is
a combination of the open flow system and the closed system. At the figure
there is a closed circulatory system with a propeller which stirrers the water.
It is a electric motor that rotates the propeller. The electric motor get it’s
power supply from a adjustable frequency AC drive that is controlled by the
swim tunnel’s main controller unit.
The closed circulatory system,
or inner camber, consists of flow tunnel and a camber where the fish swim.
(position 2 and 3 at the figure)
An optode for oxygen
measurement and temperature sensor is mounted in flow tunnel and connected to
an instrument who communicate whit the main controller unit. Position 1 at the
figure is the outer chamber or reservoir.
An open system is connected to the reservoir, where water quality and
physical parameters in the water are controlled. Water oxygen partial pressure
is controlled by a feedback control using magnetic valves connected to air
compressor/oxygen tank and a nitrogen tank. When the flush pump is running,
water from reservoir is pumped in to inner camber. Water from inner camber is
now flushed out the camber and replaced with oxygen rich water. Figure 10 also
shows the temperature regulation pump that pumps water through the
cooler/heater depended on the temperature wanted. All the actuators, i.e.
pumps, electric motor and magnetic valves are controlled by the main
controlling unit.
Figure 11. Oxygen partial pressure
when running the respiriomerer.
The oxygen level increase when
the flush pump starts and flushing the system with water from the reservoir.
(Ascending curve) The flush pump stops i.e. switching to a closed system and
wait to brake down oxygen gradients. Then a descending straight line showing
the decrease in oxygen partial pressure as a result of the oxygen consumption
of the fish measured. Let’s now take a look at the swim tunnel before
discussing the main control system.
Figure 12 the swim tunnel seen
from above, with all the equipment
including the electrodes for
conductivity measurement.
The swim tunnels main control
device named compact FieldPoint (cFP) and is manufactured by National
Instrument (NI). The cFP is described by the producer as a programmable automation
controller (PAC). A PAC is a combination of a personal computer (PC) and a
programmable logic controller (PLC). The common used Norwegian name of a PLC is
“programmerbar logisk styring” (PLS).
The cFP look like and acts
pretty much like a module based PLC with controller and I/O modules mounted on
a back plane.
Figure
mounted on a backplane. (from NI’s web site)
The I/O modules to control the
swim tunnel are a relay module, an analog output module and a possibly a
digital output module. The relay module controls the flush pump, cooling pump,
oxygen and nitrogen valve connected to the swim tunnel. Connections of devices
to the relay module implement a “fail to safe” condition, i.e. the flush pump
is running and the oxygen valve is open. The analog output module is intended
to control the adjustable frequency AC drive and thereby the flow in swim
tunnels inner chamber. Measurements of oxygen and temperature are done by an
optode and a sensor, previous mentioned, these are connected to instrument
named Microx TX3. The Microx TX3 instrument is connected and communicates with
cFP through the RS232 communication protocol.
The instrument for measuring
conductivity in fish, the 3522-50 LCR HiTESTER, will be connected to the
electrodes placed in the swim tunnel. The conductivity instrument also
communicates with the cFP through the RS232 communication protocol.
The controller and its
interface is the significant difference between a regular PLC and a cFP. In
fact the controller unit on a cFP is a small PC with a built-in web server and
support for file transfer protocol (ftp). The cFP communicates over a regular
Ethernet and connects to the network at UiB with its own IP-address. The
network properties make the cFP, a suitable device in a so called distributed
control system (DCS), i.e. a control system with several control devices
communicating over a network.
A special real-time operating
system is running the controller and not a common PC operating system like
Windows or Linux. Because of the characteristic of the cFPs operating system
the device is also called a real-time controller (RTC).
The programming of cFP has to
be done with NI’s development tool named LabVIEW and a dedicated real-time (RT)
software module. A control program with
necessary functionality to measure conductivity and run the swim tunnel as a
respirometer has to be developed. The developed program must be compiled as an
executable file, i.e. a so called exe-file, so it can be used by others without
any knowledge of LabVIEW.
It is important to note that
the LabVIEW runtime engine needs to be installed on users PC before it is
possible to run the exe-file. To make a simplified picture of the situation the
runtime engine is a software layer between the operating system and the
exe-file. The concept with LabVIEW runtime engine is pretty much the same as
JAVAs virtual machine or Microsoft .NET-framework.
Figure 14. Arrangement of the
swim tunnels control system
When the exe-file is executed
it gives the system a human machine interface (HMI) at the laptop/PC, that
gives the operator access to the system.
It is also needed to compile
and store a version of the exe-file in the RTC’s memory as an embedded
application. When the PAC is connected to a power supply and switched on, the
embedded application has to be executed under the RTC’s boot up sequence.
The PAC will, because of the
embedded application, have the ability to be a stand-alone device if the
connection to other DCS devices fails. As a stand-alone device the PAC has only
some LED’s and opportunity for some switches as the only HMI and the user will
have little or almost no access to the system.
As an option it is possible to
use one switch and a diode to make a calibration procedure for the Microx TX3
that is connected to the cFP.
Figure 15. The cFP unit connected
in the network,
including other network devices needed for the
experiment
The web cameras are needed in
order to observe the fish during the experiment. During the conductivity
experiments it is important to know exactly where the fish is located compared
to the placement of the electrodes. It is also an option to make an automatic
recognition to determine the behavior of the fish.
The measurements taken will be
stored on a server in the network, named “data” in the figure.
If possible it is interesting
if the experience with the swim tunnel and it’s equipment can result in
suggestions for improvement of future swim tunnels or experiments. It is also
of interest if this master project may lead to new ideas and technology.
Print friendly version in
pdf-format
MSc. student Endre Grimsbø
Address:
Allégt. 55,
N-5007
E-mail : endre.grimsbo@student.uib.no