NZ615299B2 - Local demand side power management for electric utility networks - Google Patents
Local demand side power management for electric utility networks Download PDFInfo
- Publication number
- NZ615299B2 NZ615299B2 NZ615299A NZ61529913A NZ615299B2 NZ 615299 B2 NZ615299 B2 NZ 615299B2 NZ 615299 A NZ615299 A NZ 615299A NZ 61529913 A NZ61529913 A NZ 61529913A NZ 615299 B2 NZ615299 B2 NZ 615299B2
- Authority
- NZ
- New Zealand
- Prior art keywords
- power
- control signal
- load
- network
- loads
- Prior art date
Links
- 238000005259 measurement Methods 0.000 claims abstract description 28
- 230000005540 biological transmission Effects 0.000 claims abstract description 10
- 238000004519 manufacturing process Methods 0.000 claims abstract description 4
- 238000011068 load Methods 0.000 claims description 187
- 230000001419 dependent Effects 0.000 claims description 26
- 230000001264 neutralization Effects 0.000 claims description 20
- 238000001514 detection method Methods 0.000 claims description 2
- 230000001808 coupling Effects 0.000 description 19
- 238000010168 coupling process Methods 0.000 description 19
- 238000005859 coupling reaction Methods 0.000 description 19
- 238000001914 filtration Methods 0.000 description 19
- 238000004891 communication Methods 0.000 description 15
- 230000001939 inductive effect Effects 0.000 description 15
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 13
- 238000010586 diagram Methods 0.000 description 12
- 238000004088 simulation Methods 0.000 description 12
- 238000002347 injection Methods 0.000 description 9
- 239000007924 injection Substances 0.000 description 9
- 230000002829 reduced Effects 0.000 description 8
- 238000009826 distribution Methods 0.000 description 7
- 230000000051 modifying Effects 0.000 description 7
- 238000005070 sampling Methods 0.000 description 7
- 238000001228 spectrum Methods 0.000 description 6
- 238000010438 heat treatment Methods 0.000 description 5
- 230000001105 regulatory Effects 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 4
- 230000001934 delay Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 230000005611 electricity Effects 0.000 description 4
- 238000000034 method Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000006011 modification reaction Methods 0.000 description 4
- 238000005096 rolling process Methods 0.000 description 4
- 230000036586 afterload Effects 0.000 description 3
- 238000004140 cleaning Methods 0.000 description 3
- 238000009434 installation Methods 0.000 description 3
- 239000000203 mixture Substances 0.000 description 3
- 238000009987 spinning Methods 0.000 description 3
- 238000007792 addition Methods 0.000 description 2
- 230000001276 controlling effect Effects 0.000 description 2
- 230000000875 corresponding Effects 0.000 description 2
- 230000003467 diminishing Effects 0.000 description 2
- 230000004907 flux Effects 0.000 description 2
- 238000002955 isolation Methods 0.000 description 2
- 230000002441 reversible Effects 0.000 description 2
- 230000003595 spectral Effects 0.000 description 2
- 230000001360 synchronised Effects 0.000 description 2
- 230000037250 Clearance Effects 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000002238 attenuated Effects 0.000 description 1
- 239000003990 capacitor Substances 0.000 description 1
- 230000035512 clearance Effects 0.000 description 1
- 230000002860 competitive Effects 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 230000001143 conditioned Effects 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 230000001809 detectable Effects 0.000 description 1
- 238000000502 dialysis Methods 0.000 description 1
- 238000004146 energy storage Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 230000000644 propagated Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
- 238000003892 spreading Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 230000001052 transient Effects 0.000 description 1
Abstract
Disclosed is an apparatus for production of a control signal for a demand side electric power supply management system. The apparatus is comprised of means to accept set point; measurement means to measure power flow into a supply network; and means to convert information from the measurement means into a control signal for transmission over the network. The frequency of the control signal is indicative of the power available to the network. into a control signal for transmission over the network. The frequency of the control signal is indicative of the power available to the network.
Description
LOCAL DEMAND SIDE POWER MANAGEMENT FOR ELECTRIC UTILITY NETWORKS
Field of the Invention
This invention relates to methods, apparatus and systems for demand side power management
in electrical utility networks. Applications of the invention include, but are not limited to, effective
use of renewable energy generation resources and charging of electric vehicles.
Background
Dynamic Demand Control (DDC) is conventionally known as a demand side management
technique where the frequency of the utility supply, i.e. the grid frequency, is allowed to vary
over a small range in response to fluctuations in the power being generated compared with the
power being used at that moment. If the available power is too high the grid frequency is
allowed to increase by a small amount; if the available power is too small the grid frequency is
allowed to reduce. The grid may be viewed as a huge spinning load and these changes in
frequency correspond to changes in the rotational speed of that load and are large energy
fluctuations. If the frequency is too high then non-essential load can be switched on to absorb
some of that energy; if it is too low then non-essential load can be switched off to free up
spinning power for more important applications. Unless the context clearly requires otherwise,
references to “DCC” herein refer to such a system.
In a practical situation a large number of small DDC capable loads, each with its own controller,
are distributed over the network. As the network frequency varies each controller determines
what load is called for and switches that fractional part load on. As shown in Figure 1 if the
frequency is less than 49 Hz the load switched is zero (W), if it is greater than 51 Hz the load is
the full rated load, and between these two extremes the load varies linearly. Thus if there were
1 million of these devices on the network the actual load applied would be a resistive load
variable from 0 to 2 GW. It should be noted that this is an example only and in practice the
range 49-51 Hz would be a lot smaller, and not all the loads have to be the same. The prime
requirement is that the DDC capable loads can be switched in a continuously variable way
between 49 and 51 Hz – or at least switched on and off inside that range.
There are limitations on the type of load that can be made DDC compliant. In general ‘energy’
loads such as water heaters, battery chargers, freezers, refrigerators, and air conditioners are
suitable but care must be taken where such loads include motors, pumps, and fans as rapid
switching of these devices on and off may adversely affect their life. Nonetheless DDC
compliant energy loads make up a significant fraction of the electric load on any grid system and
make DDC an attractive technology to implement.
DDC is implemented in the simplest possible way by allowing the mains frequency to vary in
response to loads. Schematically the whole grid can be replaced with a generator with inertia J
and a load that varies with frequency shown in Figure 2. A prime mover with no other controller
drives the inertia J representing the Grid and DDC compliant loads (not shown) connected to
the generator moderate the net torque driving the inertia J, via the feedback path.
Symbol Definition
Change in input torque
System inertia
System frequency
Constant for conversion between frequency and torque
Filter time constant
Here changes in the input torque driving the system cause the system to change speed
(frequency) according to the system inertia J. Changes in the speed are observed and used to
change the load on the network to control the change in speed. Practically, there must be at
least some filtering on the frequency measurement to remove non-linear effects caused by
armature reactance changes and other disturbances. In fact without this filtering control is
impossible. Thus aDDC controller has a feedback signal of instead of simply . In a
practical application where generation is at 50 or 60 Hz this filtering may be achieved with a
narrow band single pole band pass filter on the AC waveform to give the same transfer function
for the envelope while at the same time filtering any other noise on the signal so that
determining its frequency is simplified.
The transfer function of the system is therefore:
Which has a damping factor of:
The system performance is therefore dependant on the system inertia, filtering constant and
available controllable load. High gains and short time constants giving rapid response and high
accuracy come at the expense of a low damping factor that is not acceptable. Thus the
essence of control here is to always have enough inertia in the total power system. In our
experience using a filter with a Q of 10 corresponding to a bandwidth of 5 Hz with control
exercised over the range 49.5 to 50.5 Hz gives an acceptable response for inertias of greater
than 0.02 kG.m /2-pole kW. Thus for a 100 kW 6-pole machine the required inertia with a Q of
is 0.02 x (6/2) x 100 = 18 kG.m which is a substantial inertia. Since inertia increases as the
4 power of the machine diameter times the length inertias are more readily achieved with
larger machines. However the biggest problem with DDC is that the system frequency varies so
it cannot be seamlessly integrated into a grid network.
Object
It is an object of the invention to provide an improved demand side control apparatus, method,
system or process, or to at least provide the public with a useful choice.
Summary
Accordingly in one aspect the invention broadly provides apparatus for production of a control
signal for a demand side electric power supply management system, comprising:
means to accept a set point;
measurement means to measure power flow into a supply network relative to the set point;
means to convert information from the measurement means into a control signal for
transmission over the network wherein the frequency of the control signal is indicative of the
power available to the network.
In one embodiment the network is supplied by a transformer and the measurement means
measures the power supplied by or at the transformer.
The control signal may comprise a low voltage signal relative to the voltage of the network.
The apparatus for producing the control signals may be capable of sourcing a high current
relative to the current required by individual loads supplied by the network.
In one embodiment the control signal comprises a signal in the range of substantially 1-3 volts
at 50-500A.
Preferably the control signal frequency is substantially in the range of 300-1200 Hz.
The control signal may be provided between a neutral line and an earth connection of the
network. The control signal may also be inductively coupled to the network.
In one embodiment the apparatus derives the control signal by integrating the difference
between the measured power flow and the set point.
The control signal may comprise the frequency of the power supplied over the network.
In another aspect the invention provides a utility power supply network including apparatus as
set forth in any one of the preceding statements.
In another aspect the invention provides a method of providing a control signal for a demand
side electric power supply management system, the method comprising:
measuring power flow into a supply network relative to a set point.
converting information from the measurement means into a control signal for
transmission over the network wherein the frequency of the control signal is indicative of the
power available on the network.
The method may include varying the set point.
In another aspect the invention provides a demand side electric power supply management
system, the system comprising:
apparatus as claimed in the present invention, and;
a load controller comprising:
priority designation means for designating a priority for one or more loads supplied by
the system;
frequency detection means for detecting the frequency of the control signal;
means to control the one or more loads dependent on the control signal and designated
priority assigned to that or each load.
The control signal may be obtained directly from the network supply power to the one or more
loads.
Preferably the supply network comprises a local demand control network.
Preferably the local demand control network comprises an islanded power system.
Preferably the control signal is nominally 800Hz.
In another aspect the invention broadly provides a demand side electric supply management
system comprising an islanded power system having a point of coupling to a supply grid, the
islanded power system supplying a plurality of consumers, each consumer using one or more
electric loads, each of the loads associated with a load controller to control the power
demanded by that load in response to a control signal, a measuring means associated with the
point of coupling to measure the total power transfer between the grid and the islanded system,
and a system controller which monitors the measured power transfer into the islanded system
relative to a set point and provides a control signal to one or more load controllers by coupling a
variable frequency signal to the islanded system power distribution network to prevent power
transfer into the islanded system substantially exceeding the set point.
Preferably the load controller includes a filter means to detect the control signal.
Preferably the control signal is nominally 800Hz
In another aspect the invention broadly provides a demand side electric power supply
management system including a power system comprising group of loads and/or supplies
having a point of coupling to a supply grid, the system supplying a plurality of consumers, each
consumer using one or more electric loads, each of the loads associated with a load controller
to control the power demanded by that load in response to the frequency of the power supply in
the system, a measuring means associated with the point of coupling to measure the total
power transfer between the grid and the system, and a system controller which monitors the
measured power transfer into the system relative to a set point and adjusts the frequency of the
power supply in the system to prevent power transfer into the system substantially exceeding
the set point.
Preferably the frequency of the power supply in the system is adjusted using an electronic
transformer.
Preferably each load controller receives substantially the same control signal and determines
the maximum power which the or each load associated with the load controller is allowed to
draw from the power system based on information contained in the control signal.
The load controller may prioritise its load(s) with respect to another load or other loads whereby
a load of a first priority is controlled to draw power in preference to a load of a second priority for
a given control signal. For example a load of a first priority is controlled to reduce demand after
a load of a second priority in response to a change in the control signal to indicate that demand
needs to be reduced. The priorities assigned to loads may be changed. In one embodiment
priorities may be changed dependent on the function performed by the load.
The power flow into the system may be substantially maintained at the set point.
In one embodiment the set point represents a base power requirement for the system. The
base power requirement may be established by the consumer(s) and/or by the load controller or
a grid system operator. The base power requirement, and thus the set point, may be varied.
This may be dependent upon factors such as the power requirements of the system, the cost
structure for power supplied by the grid, and the overall power demand on the grid i.e. the
power available to the system from the grid.
The system may include one or more generators. In one embodiment, generation within the
system results in less power transferred from the grid, thereby causing the control signal to
indicate that the loads may demand more power. In one embodiment, if all loads are fully
supplied, then excess generation in the islanded system may be transferred to the grid.
In one embodiment the control signal is derived by measuring the total energy supplied to the
system compared with the energy that would have been supplied if the system had operated
continuously at the set point reference.
In another embodiment the control signal is delivered to the load controller by a low latency
communication system, the system controller monitors power transfer to the system relative to a
set point for power transfer from the grid to the islanded system to thereby establish a
differential power transfer, and provides a control signal to the one or more load controllers such
that the differential power transfer substantially averages zero.
The load controller may prioritise its load(s) with respect to another load or other loads whereby
a load of a first priority is controlled to draw power in preference to a load of a second priority for
a given control signal. For example loads of a first priority are controlled to reduce demand after
loads of a second priority in response to a change in the control signal to indicate that demand
needs to be reduced. The priorities assigned to loads may be changed. In one embodiment
priorities may be changed dependent on the function performed by the load.
The power flow into the system may be substantially maintained at the set point.
In one embodiment the set point represents a base power requirement for the system. The
base power requirement may be established by the consumer(s) and/or by the load controller or
a grid system operator. The base power requirement, and thus the set point, may be varied.
This may be dependent upon factors such as the power requirements of the system, the cost
structure for power supplied by the grid, and the overall power demand on the grid i.e. the
power available to the system from the grid.
The system may include one or more generators. In one embodiment, generation within the
system results in less power transferred from the grid, thereby causing the control signal to
indicate that the loads may demand more power. In one embodiment, if all loads are fully
supplied, then excess generation in the system may be transferred to the grid.
In one embodiment the control signal is derived by measuring the total energy supplied to the
system compared with the energy that would have been supplied if the system had operated
continuously at the set point reference.
In another aspect of the invention there is provided a method of demand side electric power
supply management comprising the steps of:
establishing a set point reference for power transfer from a supply grid to a power
system having a plurality of loads or supplies;
monitoring power transfer from the grid to the power system relative to the set point
reference to thereby establish a differential power transfer;
generating one or more control signals to control the loads present in the system such
that the differential power transfer substantially averages zero, and;
providing the one or more control signals over the power supply network of the system.
The method may include prioritising one or more loads with respect to another load or other
loads whereby a load of a first priority is controlled to draw power in preference to a load of a
second priority for a given control signal. For example loads of a first priority are controlled to
reduce demand after loads of a second priority in response to a change in the control signal to
indicate that demand needs to be reduced. The priorities assigned to loads may be changed.
In one embodiment priorities may be changed dependent on the function performed by the load.
The method may include maintaining power flow into the islanded system at a substantially set
point.
In one embodiment the set point represents a base power requirement for the system. The
base power requirement may be established by the consumer(s) and/or by the load controller or
a grid system operator. The base power requirement, and thus the set point, may be varied.
This may be dependent upon factors such as the power requirements of the system, the cost
structure for power supplied by the grid, and the overall power demand on the grid i.e. the
power available to the system from the grid.
The system may include one or more generators. In one embodiment, generation within the
system results in less power transferred from the grid, thereby causing the control signal to
indicate that the loads may demand more power. In one embodiment, if all loads are fully
supplied, then excess generation in the system may be transferred to the grid.
In one embodiment the control signal is derived by measuring the total energy supplied to the
system compared with the energy that would have been supplied if the system had operated
continuously at the set point reference.
In another aspect the invention provides a demand side electric power supply management
system controller having:
means to monitor power transfer from a supply grid to a power system having a plurality
of loads and/or supplies;
means to compare the power transfer from the grid to the islanded power system relative
to a set point reference for power transfer from the grid to the system to thereby establish a
differential power transfer power flow into the system; and
means to generate a control signal for transmission over the power supply network of
the system to control loads present in the system such that the differential power transfer
substantially averages zero.
The load controller may prioritise its load(s) with respect to another load or other loads whereby
a load of a first priority is controlled to draw power in preference to a load of a second priority for
a given control signal. For example loads of a first priority are controlled to reduce demand after
loads of a second priority in response to a change in the control signal to indicate that demand
needs to be reduced. The priorities assigned to loads may be changed. In one embodiment
priorities may be changed dependent on the function performed by the load.
The power flow into the islanded system may be substantially maintained at the set point.
In one embodiment the set point represents a base power requirement for the system. The
base power requirement may be established by the consumer(s) and/or by the load controller or
a grid system operator. The base power requirement, and thus the set point, may be varied.
This may be dependent upon factors such as the power requirements of the system, the cost
structure for power supplied by the grid, and the overall power demand on the grid i.e. the
power available to the system from the grid.
The system may include one or more generators. In one embodiment, generation within the
system results in less power transferred from the grid, thereby causing the control signal to
indicate that the loads may demand more power. In one embodiment, if all loads are fully
supplied, then excess generation in the system may be transferred to the grid.
In one embodiment the control signal is derived by measuring the total energy supplied to the
system compared with the energy that would have been supplied if the system had operated
continuously at the set point reference.
In another aspect the invention provides a method of demand side electric power supply
management comprising the steps of:
assigning a priority to each of a plurality of loads in a power system;
receiving a control signal indicative of the power available to the power system; and
controlling the loads dependent on the control signal and the priority assigned to each
load whereby a load of a first priority is controlled to draw power in preference to a load of a
second priority for a given power availability indication from the control signal.
In one embodiment the power system comprises an islanded power system.
The control signal may be provided using a low latency communication system. The control
signal may comprise the frequency of operation of the power system.
The islanded power system may receive power from a grid supply.
In another aspect the invention provides a load controller for a demand side electric power
supply management system, the controller comprising:
priority designation means for one or more loads;
means for receiving a control signal, the control signal indicative of the power available
to a power system which supplies the one or more loads; and
means to control the one or more loads dependent on the control signal and designated
priority assigned to that load.
In one embodiment the load controller stores a priority designation for each of a plurality of
loads and controls the loads dependent on the control signal and the designated priority
whereby a load of a first priority is controlled to draw power in preference to a load of a second
priority for a given power availability indication from the control signal.
In one embodiment the power system comprises an islanded power system.
The control signal may be provided using a low latency communication system. The control
signal may comprise the frequency of operation of the power system.
The islanded power system may receive power from a grid supply.
In another aspect the invention broadly provides an appliance for use with a demand side
electric power supply management system, the appliance comprising:
a priority designation means;
means for receiving a control signal, the control signal indicative of the power available
to a power system which supplies the appliance; and
means to control the power demand of the appliance dependent on the control signal
and designated priority.
In another aspect the invention broadly provides a demand side electric power supply
management system comprising an islanded power system having a point of coupling to a
supply grid and a variable power supply from a generator connected to the islanded system, the
islanded power system supplying a plurality of consumers, each consumer using at least one
load, each of the loads associated with a load controller to control the power demanded by that
load in response to a control signal which is delivered to the load controller by a low latency
communication system, a system controller which provides a control signal to the one or more
load controllers such that the power from the generator is preferentially supplied to energy
loads.
In another aspect the invention broadly provides an electric vehicle power supply management
system comprising an islanded power system capable of supplying power to a plurality of
electric vehicle loads and having a point of coupling to a supply grid, each of the loads
associated with a load controller to control the power demanded by that load in response to a
control signal which is delivered to the load controller by a low latency communication system,
a system controller which monitors power transfer to the islanded system relative to a set point
for power transfer from the grid to the islanded system to thereby establish a differential power
transfer, and provides a control signal to the one or more load controllers such that the
differential power transfer substantially averages zero.
In one embodiment the electric vehicle loads are inductively coupled to the islanded power
system.
In one embodiment the islanded system is arranged to to provide power inductively to the
electric vehicle loads when the electric vehicles are on a vehicle carrying surface such as a
garage floor, carpark or roadway.
In another aspect the invention provides a method of demand side electric power supply
management comprising the steps of:
assigning a priority to each of a plurality of loads in a power system;
receiving a control signal indicative of the power available to the power system;
monitoring a characteristic of at least one of the loads; and
reassigning the priority for one or more of the loads dependent on the control signal and
the monitored characteristic.
The monitored characteristic may include one or more of: the power presently demanded by the
load; the state of charge of the load; whether the load has been switched off or on by a user.
In one embodiment the power system comprises an islanded power system.
The control signal may be provided using a low latency communication system. The control
signal may comprise the frequency of operation of the power system.
The islanded power system may receive power from a grid supply.
In another aspect the invention provides a demand side electric power supply management
system comprising an islanded power system having at least one point of coupling to a supply
grid, the islanded power system supplying a plurality of electric loads, each said load associated
with a load controller to control the maximum power demanded by that load, the system further
comprising measuring means associated with the or each point of coupling to measure the total
power transfer between the grid and the islanded system, wherein each load controller
determines the maximum power which the or each load associated with the load controller is
allowed to draw from the islanded power system based on a comparison of the measured power
transfer into the system with a set point.
In another aspect the invention broadly provides a load controller for a demand side electric
power supply management system, the controller comprising:
- a priority designation means for one or more loads or supplies;
- filter means for detecting a control signal; and
- means to control the one or more loads or supplies dependent on the control signal
and designated priority assigned to that load or supply.
In one embodiment the loads comprise a selected group. In one embodiment the group may
comprise an islanded system.
In one embodiment the control signal is indicative of the power available to the group/Island.
In one embodiment the control signal is dependent on a sum, or difference, or ratio, or other
relation between the rate of power supply will drain by the group/Island and a desired or set
point rate.
In one embodiment the control signal is a high ampere, low voltage electrical tone signal which
can be filtered using economical filters out a 50 Hz AC signal. The voltage may be in the order
of one or 2 V and the current may be in the order of an Amp upwards. Preferably the control
signal is detectable at the neutral of electrical supply wiring. This may be in reference to phase
or positive.
The group of loads may or may not be islanded.
In another aspect the invention broadly provides an electric power supply management system
comprising a power system connected to a supply grid at one or more points to transfer a power
to or from the grid, the power system supplying a plurality of consumer sites with power from the
grid or supplying the grid with power, each consumer site using one or more electric loads, each
of the loads associated with a load controller to control the power demanded by that load, a
measuring means associated with measure the total power transfer between the grid and the
islanded system, and a system controller which monitors the measured power transfer between
the system and the grid system relative to a power or energy constraint and provides a control
output suitable the load controllers to use, the control output dependent upon the constraint and
the total power transfer between the system and the grid. The constraint may be a set point of
energy in a time period or a rate of power.
Preferably control data adjusts the frequency of the power supply in the islanded system to
prevent power transfer into the islanded system substantially exceeding the set point.
In another aspect the invention broadly provides apparatus for the production of a control signal
for an electric power supply management system,
the apparatus comprising:
means to accept a desired set-point,
means to measure the power flow into an islanded network,
means to convert the measurement or information relating to the measurement into a control
signal.
In one embodiment the set-point is provided by a grid operator.
In one embodiment the set-point may be updated and controlled at set intervals or continuously
by the grid operator or islanded system.
In one embodiment the difference between the power flow into the islanded network and the
set-point is measured and integrated to create the measurement.
In one embodiment the power measurement is converted into a frequency control signal.
In another aspect the invention broadly provides apparatus for communicating a control signal in
an electric power supply management system, the apparatus comprising:
means to accept an input signal,
means to create a low voltage high current signal,
means to apply the to a neutral line.
In one embodiment the signal is 2-3V at 100A and uses a frequency between 600-800Hz.
In one embodiment an inverter is used to inject or couple the control signal onto the neutral line.
In one embodiment the means for communicating is used to control an LDC system.
In another aspect the invention broadly provides for a load control device, for provision between
a power supply and a load, the load control device comprising:
means to respond to a control signal,
means to change the amount of power provided to a load based on the control signal.
In one embodiment the load control device comprises a dongle.
In one embodiment the load control device is reprogrammable.
In one embodiment the load control device can be connected to an external device for
reprogramming or monitoring.
In one embodiment the load control device includes manual controls to change the priority of the
load switching.
In one embodiment the response to the control signal is dependant on the priority of the device.
In one embodiment the load control device is part of an appliance.
In one embodiment the load control device is part of an LDC system.
In another aspect the invention broadly provides for an appliance with an included load control
device.
In one embodiment a feature of the load control device is dependent on the appliance.
In one embodiment the load control device is reprogrammable through the controls of the
appliance.
In one embodiment the load control device is visible on the appliance.
In another aspect the control signal of the electrical power supply management system is
available to a monitor means, the monitor means being adapted to:
display and record the control signal,
track the historical values of the control signal.
In one embodiment the monitor means is available to a grid operator and allows tracking of the
power used by the islanded system.
In one embodiment the monitor means is available to one or more users of the islanded power
system.
References to loads in the foregoing statements may also include sources or supplies i.e.
generators and/or supplies (such as batteries) of electricity, so that the system can be used to
control a supply which supplies the grid.
Although this document refers by way oe example to an islanded system, use of that term is
intended to generally include a group of electrical loads and/or supplies, as well as the specific
instance of an islanded system.
Further aspects of the invention, which should be considered in all its novel aspects, will
become apparent from the following description.
Drawing Description
One or more embodiments of the invention will be described below with reference to the
accompanying drawings, in which:
Figure 1: is a graph of Power versus frequency for a known 1 kW DDC Controller;
Figure 2: is a Block diagram, and transfer function of a known DDC controller;
Figure 3: is a generalised schematic diagram of a LDC system according to the invention;
Figure 4: is a Block diagram, and transfer function of a known DDC controller;
Figure 5: is a diagrammatic illustration of an inductively powered electric vehicle;
Figure 6: is an illustration of a roadway for inductively powered vehicles;
Figure 7: is a series of graphs showing measured generator frequency and charging
system output power with a DDC controlled IPT battery charging system and a
random torque input;
Figure 8: is a graph showing an example plot of wind supply, grid supply and total
consumption over a one hour period with an LDC controller;
Figure 9: shows two probability distributions, one for power consumed from a grid and the
other for power consumed from wind;
Figure 10: is a graph showing single household demand over a one hour period;
Figure 11: is a graph showing the system response to a 20kW step in wind power;
Figure 12: is a diagram of an islanded power system illustrating generation and use of a
variable frequency control signal;
Figure 13: is a diagram of frequency against voltage for generation of the control signal in
the system shown in Figure 12;
Figure 14: is a diagram showing operation of a load control device;
Figure 15: is a diagram showing information flows in an LDC island according to one
embodiment of the invention;
Figure 16: is a diagram illustrating an example of a hybrid system which integrates DDC and
LDC;
Figure 17: shows plots of wind power and load power over time; and
Figure 18: shows plots of standard deviation against time for Transport Delay and Sample
Time in respect of the simulation relating to Figure 17;
Figure 19: is a block diagram of another embodiment of an LDC contol system;
Figure 20: shows a schematic diagram of another embodiment of an LDC system and
controller;
Figure 21: shows a simulation of an input signal containing both a 50Hz mains and 800Hz
LDC signal;
Figure 22: is a frequency spectrum of the input shown in Figure 21;
Figure 23: shows the output of a high pass filter fed with the input signal of Figure 21;
Figure 24: shows the output of a band pass filter fed with an input signal from the output of
the high pass filter of Figure 23;
Figure 25: shows the input spectrum with noise of .
Figure 26: shows the output spectrum noise of .
Figure 27: is a plot of the raw measurement output.
Figure 28: shows a plot of measurements taken using simple 128 sample averaging;
Figure 29: shows a plot of measurements taken using a weighted rolling buffer;
Figure 30: shows a plot of measurements taken using a combination of 16 averaged data
points and then 16 weighted rolling average points;
Figure 31: shows the mains voltage in a practical LDC system;
Figure 32: shows the high pass filter output of the system of Figure 31;
Figure 33, 34: show the band pass filter outputs for 710Hz and 864Hz respectively from the
output of the high pass filter of Figure 32;
Figure 35: shows the output of the bandpass filter when the signal injection system is turned
off;
Figure 36-39: show the results of use of a schmidt trigger at 733Hz, 800Hz, 868Hz and710Hz
respectively;
Figure 40: shows the output when there is no signal injection;
Figure 41: shows a series of applicances connected to a system such as an islanded
network in accordance with the invention.
Description of One or More Embodiments
A new approach to dynamic demand side control will now be described. In one embodiment
this new approach allows the mains frequency to be constant and also allows for local
Distributed Generation (DG), as shown in Figure 3. This approach may be considered to be a
form of distributed generation demand control, but for convenience this approach is referred to
in this specification as Local Demand Control (LDC). In essence, a load in a power system
comprising a selected group of loads and/or supplies (in one example an islanded system) that
has a connection point to the grid is controlled to prevent the power supplied from the grid to the
islanded system from substantially exceeding a set point. This control concept can also be
used to control supply of power from the system to the grid.
The term “islanded system” is used in this document to refer to a power system or subsystem or
network that may or may not include generation and which has at least one point of coupling to
a utility supply grid. An islanded system may supply power to a plurality of consumers who use
loads (for example domestic appliances), or possibly share a load, connected to the system. In
one example an islanded system may comprise a single household, and in another example
may comprise a city. In another example an islanded system can be defined by a number of
households which are not necessarily located in the same immediate geographical area
collectively agreeing to form an islanded system for the purposes of implementing the invention.
The loads of the islanded system may include any power drawing device, for example including
household appliance, electric vehicle charging device, hot water heater. Loads may also be, or
at certain times act as, sources, such loads include for example renewable energy generators,
inverter outputs, battery banks or energy storage devices. Multiple loads may also be combined
into groups, consisting of a variety of individual loads, so that the operation, control or
monitoring of the loads can be linked. These grouped loads do not, necessarily, share a
controller and may be connected to the islanded grid at one or more points. A group of loads
may then be monitored against a set-point/reference different to that of the main islanded
system.
One example of implementation that will be set forth below is that of an islanded system
comprising a small community such as a farm or a small village. However, as noted above the
system is also applicable on an even smaller scale, such as an individual dwelling. Similarly,
the invention may be implemented on large scale islanded systems such as a city.
In the case of a farm for example, power is available from the grid 1 but it may be at the end of a
long feeder that also drives other isolated farms, so that it is weak and highly variable. This
single, or sometimes three-phase, feeder cannot drive large loads without disrupting the power
supply to neighbours so the idea of charging one or more electric vehicles at perhaps 7 kW is
not practical. In the example here the islanded system includes a generator, so power from the
utility supply grid is available in addition to locally generated power. In this example the network
or islanded system includes a generation in the form of a wind-turbine 2 which drives a single
phase induction generator 3 to produce single phase power. Three phase power can also be
generated. The power available on the farm is then the power from the utility feeder for
example 15 kW, and the power from the wind-turbine which might be only 20 kW in a small
application, and which varies widely as the wind strength varies.
A set point reference can be established for the power available from the utility feeder. Thus the
set point represents a base power requirement for the network or islanded system supplied from
the utility grid. The base power requirement may be established by the consumer(s) and/or by a
load controller or a grid system operator, as discussed further below. The base power
requirement, and thus the set point, may be varied. This may be dependent upon factors such
as the power requirements of the islanded system, the cost structure for power supplied by the
grid, and the overall power demand on the grid i.e. the power available to the islanded system
from the grid. In one example, the system demand on the grid can be monitored and then the
set point can be adjusted based on the demand trend when the next utility billing period begins.
Thus if the utility bills in half hour periods, then the set point may be adjusted to coincide with
the commencement of the next half hour period. The system may signal the intended change in
set point to the grid system operator in advance of the change. In a second example the set
point may react to the above factors and/or the conditions of the grid, including frequency,
voltage or other electrical signal. Alternatively the set point may be transmitted independently of
the grid, using wired or wireless communications including GPRS, Internet, or Wifi. This type of
communications may be of particular use when the islanded system is not geographically
connected. In this case it may be necessary to calculate the set-point at a set of points, or
transformers, use this to calculate the control signal and then send the control signal to the load
controllers, such as dongles (described further below), of the islanded system.
The available power is the sum of the set point power from the grid and the generation within
the islanded system at any instant. A control signal which is indicative of the available power
may be added in common to all the phase voltages of the local system from a system controller
. The power transferred into the islanded system may be monitored compared with the set
point to establish a differential power transfer, and this differential power transfer may be
controlled by appropriate load control to average zero. The monitoring or measuring means
may measure power at the, or each, connection to the grid, or elsewhere if desirable. The
monitor or measuring means may relate the total power into, or out of the islanded system
against the set point and generate an output signal. The output signal may constitute the
control signal to be communicated to the load controllers; alternatively further processing of the
output signal may occur before the control signal is created. The signal from the monitor or
measuring means and/or the control signal may also be made available to independent devices
to enable monitoring by other devices, users or providers. In one example the control signal
may be used by a user to monitor the power in the islanded system or part of thereof. This may
include storing the control signal and/or producing further data based on the control signal. In a
second example the grid controller may wish to monitor the control signal, or power flows
into/out of the system, either at a particular instance or over a period of time. In one
embodiment there may be a monitor for the user which allows the control signal to be monitored
and system characteristics, such as priority, of supply for various loads to be changed.
Similarly, one embodiment provides the grid controller with a monitor of the control signal and
enables the system characteristics such as base power requirement, to be changed. In one
embodiment the control signal is derived by determining the energy transferred to the islanded
system from the grid over and above the energy that would have been taken had the system
operated continuously at the set point. This difference may be represented by a voltage, and
used to generate a control signal as discussed further below.
Further detail is shown in Figures 12 to 14. The system controller 5 may produce a voltage from
which a control signal is generated using a voltage to frequency converter 21. In this example
the control signal is a simple tone e.g. 1 Volt varying from 200 Hz to 1kHz, as shown in Figure
13, corresponding to a “guaranteed” lower power limit to 20 kW for example. Each appliance 37
or household 9 for which the system is implemented is associated with a load controller, such as
controller 30, which receives the common control signal described above from the network
power transmission lines directly for example and controls the appliance or appliances 37
accordingly. The control of the appliance may include providing a defined response (on/off) for
the load based on a feature of the control signal (including frequency of a simple tone or other
electrical property). Each load being controlled may have a different response to the control
signal, thus allowing a priority of loads to be implemented. Referring to Figure 14, one example
of a load controller implementation is shown in which the load controller 30 has filters 31 to 34
(and possibly more) that correspond to different load switching or operation priorities. A dongle
36 is provided connected between the controlled load, such as appliance 37, and the supply (in
Figure 14 represented by power outlet socket 35). Dongle 36 includes a switch to enable on/off
or variable control and is responsive to an instruction from load controller 30 to increase or
decrease the demand of the appliance 37, or other load, to which it is connected. The dongles
may store information, or may be designed so that they react to the control signal only. It is
possible to have a dongle in which the priorities may be changed, including through a physical
device on the dongle or through some wired or wireless programming method. In one
embodiment the dongles select an appropriate reaction based upon the frequency of the control
signal received. Communication between the dongle 37 and the load controller may occur over
a wireless network in one embodiment. As an alternative, the functionality of the controller may
instead be included in dongle 36 so that separate controller hardware is unnecessary. The
controller or dongle 36 may be provided as part of the appliance, so that separate hardware is
unnecessary. Referring to figure 41 each appliance 41 has an included dongle 42 which
receives the control signal 43 described above. An appliance 41 with an included dongle 42
may have modifications which make better it better suited to the local demand control (LDC)
system. In one embodiment there may be multiple control signals used, with different loads, or
groups of loads receiving and acting upon the different control signals.
Thus, any relevant appliance or electrical device will then turn on/off (or have its demand
controlled variably if that is possible or appropriate for the given load type or function the load
performs) as the control signal frequency varies in response to more or less power becoming
available as the wind speed varies, and as other loads turn on or off. The control signal may
also be a digital signal propagated by wire or wirelessly over the community supplied by the
network. All the LDC compliant devices get the signal at substantially the same time and turn
on/off appropriately. Therefore, the control signal is most effective transmitted by a low latency
system. In Figure 3 the control signal is shown as being generated at the local transformer of
the utility grid. This is a convenient practical location for such a controller as it can measure the
power supplied from the grid at this point. However it will be appreciated that the load controller
could be located at another physical location, and may even be located remote from the
islanded system. The control signal may be provided by means other than that described
above. For example, a wireless communication system or network could be used. Thus the
communication of the control signal may occur by varying the system frequency, by radio
signals, by WiFi or Zigbee, or by Internet for example. For example, in a situation where the
islanded system is not geographically connected a radio frequency signal or internet link, may
provide the most effective communication means. In some systems it may be preferable to have
multiple communication methods or control signals. The control signal may also be
communicated to devices other than the controller, including other components of the system,
or external monitors.
As mentioned above, each load may be designated a priority. The order of priority is whatever
that community, or individual consumers in that community, want. Setting priorities should be
considered carefully as devices at the high power end are likely to be turned on and off
relatively often and some devices, for example refrigerators, are not rated for rapid switching.
Devices which consume relatively low amounts of power can be put at the low power end of the
priority list. Those for which frequent on/off switching is undesirable can include an operation
schedule which prevents a switching action for a certain time period. For example a schedule
for a certain load (such as a refrigerator) may include a schedule which requires that whenever
that load is switched on/off it must stay on/off for at least 10 minutes or until it turns itself off.
This may also depend on other factors such as time of day or load condition for example. The
priority for each appliance can be stored in each load controller in such a manner that it can be
varied either by a user or varied intelligently by the load controller depending on parameters
such as the function performed by the load. Alternatively the prioritisation of the loads may be
implemented by adjusting the response of the controllers so that they operate to give the
intended priority. In one example a higher priority load may require a larger frequency change
than a low priority load. This method may not require any storage of priority information on the
controller.
The consumer and/or the “community” in the network or islanded power system can decide on
load prioritisation. For example in Figure 3 each household is shown with a water heater load 6
and an electric vehicle (EV) charging load 7. If a consumer has chosen the EV load 7 to
ordinarily have a lower priority than the water heater load 6, then load 7 will ordinarily be
switched off in preference to load 6 as the control signal indicates that the available power
supply is diminishing. However, the load controller may change the priority if it determines (or
receives feedback indicating) that the EV charge is very low for example, or the water
temperature is sufficient (even if it is not optimal), or dependent on the time of day (for example
cutting water heating in the middle of the night in preference to vehicle charging and
recommencing water heating at an appropriate time). Both loads 6 and 7 are of a type that can
be controlled to be continuously variable, and the load controller may perform that function. It
will be seen that the system of prioritisation described herein is applicable to DDC in general
and is not necessarily limited to use in an islanded power system as the control signal that is
used may be the frequency of the power system. It will be seen that “energy” loads, which are
tolerant to power supply variability, such as water heating and EV charging can be prioritised so
that the variable generation from the generator 3 is effectively used to supply those loads. Thus
the invention can make good use of variable generation such as that from renewable sources
including wind, solar and tidal generation sources, for example.
In the form described above with the wind-turbine driving an induction generator both the
voltage and the frequency are set by the grid. The power taken from the grid can be reduced to
zero and power can even be exported back to the grid if the power is not actually being used i.e.
if all loads are being supplied as required. Indeed in larger applications the System Operator
(SO) for the islanded system can ask for the grid power to be reduced, if possible, or a higher
‘time of day’ pricing schedule might be incurred. If there is a surplus that is not wanted by the
SO it can be used for water heating or dumped. In one embodiment a set point reference can
be established for the power delivered to the islanded system from the grid (the feeder in the
farm example described above) and the controller can provide a control signal to the
controllable loads so that the power delivered to the islanded system does not exceed, or at
least does not substantially exceed) the set point. Furthermore, dependent on the nature of the
loads supplied by the islanded system and the generating capacity in the islanded system, the
islanded system may be managed so that the power delivered from the grid is substantially
maintained at the set point, at least for certain time periods. In this way the demand placed on
the grid is more predictable, with less unexpected change in demand, so spinning reserve can
be lessened or at least be more economically managed by the grid operator.
In some systems, for example those with relatively low power usage and high generation
capacity, the islanded system may operate to feed a substantially constant amount of power into
the grid. The return of power to the grid may include occasions when generation is greater than
the maximum load in the system, at times of high demand in the grid, at times of low demand in
the islanded network or at times set by the LDC controller. In one example the set point could
be adjusted to indicate that energy could be returned to the grid. In a second example, at times
of high demand, the grid controller could request a change in the set point so that power is
returned to the grid, with non priority loads turned off. In these systems the grid controller may
be able to adjust the power drawn from multiple LDC systems to smooth out the load on the
grid.
A significant feature is that fluctuations in the wind speed causing variations in the power being
generated are essentially removed by the LDC controller so that if power is programmed to be
sent back to the Grid then it will be high quality constant voltage grid frequency single phase or
3 phase which has a high value. However if power is sent back to the grid because there is
insufficient load to absorb all of the power available it will be lower quality and consequently of
lower value. In the event of a power cut this system cannot generate as the induction
generators will have insufficient VAR excitation; this is by far the lowest cost implementation and
also the safest as the local generation cannot enliven a line that the power company has turned
off for whatever purpose. Where power continuity is essential, for example for a dialysis
machine, UPS could be used.
A controller for this LDC system is shown diagrammatically in Figure 4. It is similar to the
controller shown in Figure 2 except that the feedback path is now completely in the induction
generator. In these circumstances the output ∆ω is now a change in the slip frequency of the
machine causing a torque feedback of T where ώ is the rated slip frequency of the
induction machine and T is the rated torque. This gives a transfer function
1+ s
This transfer function corresponds to a first order system with a short time constant so that the
expected system response is fast with no overshoot.
A comparison between a conventional DDC controller and the LDC controller is shown in Table
1. The significant differences are that some embodiments of the LDC system need an extra
communications feed to the LDC compliant devices, but it can run in a mixed power mode
where power is taken from the grid and the wind turbine at the same time. The conventional
DDC system is essentially a stand-alone system best implemented with a synchronous
generator whereas LDC operates as an island in a grid system but with its own internal
controller and is best implemented with an induction generator. Conventional DDC is
responsible for its own frequency and voltage control whereas LDC takes its voltage and
frequency from the grid but power can go in either direction and changing the direction of power
flow is simple and seamless.
Attribute DDC LDC
Run Stand-alone Yes No
System frequency Local control Grid
Voltage regulation Local control Grid
Mixed Power Mode No Yes
Frequency range 50 ± 0.2-0.5 Hz Grid 50 ± 0.2 Hz
nd st
Response 2 Order 1 Order
Damping factor Inertia critical Inertia not critical
Generator Synchronous preferred Asynchronous preferred
Switch to Grid Power Complex system Seamless
VAR controller needed No No
Phases 1 or 3 1 or 3
Response time <1 second <1 second
Cost High Lower
Table 1: A comparison between controllers
The most significant difference between the machines and their controllers is possibly the
inertial requirements. Wind machines are relatively low inertia and the LDC system can operate
with low inertia. Conventional DDC systems need approximately 0.02 kg.m of inertia for each 2
pole kW. Thus a 12 pole 1 kW machine needs 0.72 kg.m, and a 12 pole 100 kW machine
therefore needs 72 kg.m . These inertias may be quite difficult to achieve but without them the
damping of a conventional DDC controller may be poor. The LDC controller is helpful in this
respect.
As mentioned with reference to Figure 3, the invention also has application to Electric Vehicles
(EV’s), both for charging and roadway power requirements. Examples of EV inductive charging
and inductive roadway use are described in our published pending applications WO008/140333
and WO2011/016736. Although these publications predominantly refer to inductive coupling of
vehicles to a power system, it will be appreciated that the present invention may find application
to either inductive or non-inductive coupling mechanisms.
Referring to Figure 5, an EV which is charged inductively is shown. A stationary power supply
energises a track or pad 11 in or on a floor or roadway. The vehicle 12 has a pick-up coil 13
and the electric energy transferred to the pick-up is conditioned and provided as DC power for
use with charging and/or operating the EV.
Referring now to Figure 6, when EV’s are in motion along the road 20 they can be powered
inductively from an ‘endless’ string of pads 11 buried in the roadway. These pads are powered
by power supplies 10 spaced perhaps 200m apart and driving 100m of roadway in each
direction. As a vehicle moves along this road 20 the pad(s) 11 underneath it are energised
synchronously with its motion providing a power wave that keeps the vehicle fully charged.
Each pad produces an arched flux across the roadway that switches from pad to pad as the
vehicle moves. The vehicle is powered at 10-20 kW depending on whether one pad or two is
providing linking flux and this power is sufficient to power the vehicle and keep the battery fully
charged. Each 100 m section may or may not have a vehicle on it – if there is no vehicle then
this section switches off. Conversely each section may have 5 cars at 20 kW each with 20 m
spacing between the vehicles. If there are more vehicles then the section is overloaded and a
DDC system is used to reduce the power to each vehicle so that the system does not collapse.
The power supplies 10 provide an IPT frequency of 20 kHz; this 20 kHz is varied between 19.9
and 20.1 kHz to indicate the loaded condition of the section – at 20.1 kHz the vehicles take full
power, at 19.9 kHz they take reduced power in a classical DDC situation. These sections of
roadway could be driven from a mains supply or from local wind or other ‘green’ sources. Thus
these systems may comprise islanded power systems to which the invention is applicable.
Overloaded sections trigger a signal ‘congestion – increase spaces between vehicles’ to the
driver. The introduction of LDC on an islanded system composed of a series of inductive
charging pads may allow for local control of the power drawn from the grid and removes the
need for frequency variation of the grid system. In one example an islanded system consisting
of a set of inductive charging pads, and possibly including energy generation, could monitor
traffic levels, power usage and power cost to balance the needs of the system.
Similar applications – though a lot simpler- will exist in car park buildings that offer parking and
charging. Here one power supply 10 can drive many pads 11 and charge many vehicles at the
same time to give a simpler arrangement than one power supply and pad per parking space as
in a garage or parking place at home.
Measurements and Simulations
1. Simulation of a battery charger with DDC and IPT coupling
A classical DDC controller has been tested under laboratory conditions and by computer
simulation. In the laboratory a controlled AC drive in a torque controlled mode generated a
string of random torques changing each second. The AC drive (variable speed induction motor)
was connected to a 3 phase alternator generating at 50Hz. Two of the phases were on resistive
loads, and the third phase was passed to a DDC controller set up to charge an electric vehicle
battery at 300 V DC. The measured and computer simulated outputs are shown in Figure 7.
The system was controlled by the DDC controller at 1000 rpm with a 4 pole induction motor and
a 6-pole alternator. A huge advantage of this experimental set-up is that the same random
sequence can be used for all of the tests.
The first graph 7 (a) shows the random torque signal used. The second graph 7 (b) shows the
generator frequency (equivalent to shaft speed) with and without DDC control, and the third
graph shows the current into the battery (with DDC control). Since from graph 7 (b) the speed
with DDC is essentially constant the power input is a scaled version of the first graph and the
power output, with a constant voltage battery, is a scaled copy of the battery current. Thus
ideally graphs 7 (a) and 7 (c) should be the same – the correlation between them is
exceptionally good showing the accuracy of the DDC controller. The 4 graph 7 (b) shows a
simulation on Simulink TM for the expected battery current from the circuit. It is a close fit to the
measured data with the same average current and slightly less variation showing that the inertia
figures for the experiment and the simulation are not quite identical.
2. Simulation of the power used in a small community
An LDC system can be used in many circumstances wherever there is a community of common
interest. Perhaps the simplest is a 400/230 V distribution transformer where all the consumers
on the transformer form the LDC system. Here there is no wind power but the transformer load
may be monitored and the connected houses switch LDC compliant loads so that the total load
of all the houses is managed. In this way the load presented by this transformer to the 11 kV
feeder is almost constant. The transformer operates at a higher load factor and problems of
residential infilling are greatly reduced. Also the electronics can monitor the supply frequency
and if it is too low it can drop all non-essential loads, and if it is too high it can switch on all
possible loads.
Here in a slightly more complex situation the power demand for a small community comprising
twenty houses containing LDC compliant loads and EV IPT charging pads has been performed
where mains power to a nominal maximum amount of 20 kW (1 kW/house) is included. Wind
power is added as a random sequence, changing every ten seconds, with an average value of
70 kW. The load taken by each household averages 3.5kW, but can peak at up to 7kW. This
system therefore includes 77% wind which is very high.
Central to the system is the LDC controller which measures power flow to the grid and
compares this with a known limit or set point reference. This set point may be set manually or
provided to the LDC controller from the grid through wired or wireless communication or through
some electrical characteristic of the grid power. A simple integral controller may then be used to
determine the difference between the energy supplied to the islanded system compared to the
energy that would be transferred if power were being supplied at the set point and uses this to
produce a differential power signal which is provided to the system as a power priority signal
that varies from 0 to 10 in real time. The most important device is priority 1, whilst the least is
priority 10. Consequently, devices with priorities below the signal will stay on whilst those above
will be switched off. Different devices and/or dongles may react differently to control signals,
with the difference possibly dependent on the type of load or source being controlled. The
control is thus implemented so that the differential power, i.e. the difference between the power
supplied from the grid and the set point power reference is substantially zero on average. In one
embodiment it may be desirable to measure the energy, or time averaged energy.
Each house consists of a number of LDC controlled loads. These are listed in
Table
Table 2.
Load Average Power Peak Power Priority
EV Charger 2kW 4kW 4-10
HWC 500W 2kW 4-9
Refrigerator 60W 250W 1-6
Base Loads (x4) 250W 250W 1,2,3,4
Table 2: Simulated loads in each household
All loads except the base loads are assumed to vary linearly over their given priority range,
consuming minimum power at a lower priority signal. The four 250W base loads are simply
switched off if the signal goes below their given priority. A small random offset is given to each
of these so that not all houses’ base loads of equal priority switch at exactly the same time.
An example of the simulation output is shown in Figure 8. It can be seen that the wind varies
significantly but the load on the system is kept in step with this varying wind. The power drawn
from the grid is regulated to 20kW. The probability density functions for the power taken from
the grid and the power generated from the wind are shown in Figure. 9. The left plot shows the
power supplied from the grid and gives an idea of regulating efficiency. The right plot gives an
idea of the range of the power output from the wind turbine. Note that the grid power is almost
constant at 20 kW with deviations caused by loads switching on and off. The wind power is a
roughly Gaussian distribution with a wide standard deviation – the ideal result would perhaps be
a Weibull distribution p(x) where x is the wind speed, modified to x to represent the power
output demand in approximately one second.
Figure 10 shows the power usage over 1 hour for a single house. Here the power taken is quite
volatile but when combined with all the other houses the % variation can be improved
considerably. It can be seen that the fridge and hot water cylinder modulate their switching
times to coarsely adjust demand, while the EV charger fills in the gaps. In this way a large load
with continuously variable control is seen to be important to the controller strategy.
The response of the system to a step in wind is shown in Figure 11. The system adds 20kW of
demand in about three seconds in a predictable first order response with no overshoot. It can be
observed that this response is made of steps in load and more continuously variable load as a
function of time. The smaller loads are simply switched on and off depending on the availability
of power while the larger EV and water heating loads are continuously variable and take power
depending on the amount of power available making the overall response more linear.
Figure 15 shows another example of the layout and information flows in a fully LDC island which
includes generation, distribution and a number of houses all with LDC controllers. Here each
LDC controller outputs a signal based on both the signal from the parent node and the power
throughput measured locally, that is, the “set point” of the system controller of the islanded
system can be varied, possibly continuously, based on information from the grid which indicates
the total load on the grid. While deploying LDC network wide would be a significant task, the
system works just as well in isolation. A hybrid of DDC and LDC would also be very easy to
implement and is shown in Figure 16.
In Figure 16 generators so supply loads 52 and 53, and generator 54 supplies loads 53. The
system frequency is used as the signal to LDC controllers built into transformers (not shown).
The LDC controller then takes this into account when calculating the control signal for
subsequent devices running off that transformer. In this way, the LDC system would help
balance both overall supply and demand with DDC and manage local constraints with LDC.
In general usage of DDC and LDC can be categorised into three main usage scenarios based
mainly on size as shown in Table 3. As DDC requires allowing the frequency to vary, it is most
useful in islanded grids. These could be large systems such as the North Island of New Zealand
or small isolated systems such as remote villages. Alternatively, LDC is suitable in mid-size
systems where the frequency may not be allowed to vary or may not represent the generation
constraints of the grid. A community with local wind generation is a good example of this.
TABLE 3
DDC USAGE SCENARIOS
Type Example Constraint Possible Operating
Signal Goal
Grid wide North Total Grid Balance
(DDC) Island generation frequency supply
demand
Localised Single Transformer, Local Keep
(LDC) street, Grid tone or within
Farm interconnect similar ratings
community
Isolated Island Local Grid Fully
(DDC) village generation frequency utilise
renewable
source
Depending on what transmission method is used to transmit the control signal, transport delays
and sampling of the LDC control signal may be unavoidable. This could be introduced by
analogue filtering or using digital communication.
In order to quantify the effects of latency and sampling of the LDC control signal, a modified
version of the simulation mentioned above was run. The simulation was run for 15 minutes with
a square-wave modulated wind turbine output. The turbine output changes between 20kW and
40kW four times during the simulation as shown in Figure 17. Two separate tests were done. In
the first, an analogue delay was introduced that varied between 0s and 1s.
In the second test a sample and hold was added to the priority signal, with the sampling period
also varying between 0s and 1s. In each, the system performance was measured with each
change in delay or sampling rate.
As the LDC controller is trying to regulate the grid interconnect power to a specific level, the
variation is a good measure of how well the LDC system is performing. In order to get the best
performance from the LDC controller, the integrator gain was modified with each change in
delay or sample time in order to avoid overshoots or oscillations caused by the transitions. The
results of both tests are shown in Figure 18.
It can be seen that the ability of the LDC system to regulate power consumption is almost
linearly related to any transport delay or sampling in the system. For this reason, the LDC signal
needs to be transmitted as fast as possible in order to get maximum performance. A delay of
between 0.1 and 0.2 seconds is a realistic goal and this still yields good performance. It can
also be seen that sampling rate has less effect on system performance than transport delay. A
sampling period of <100ms (>10Hz) is sufficient for a good performing system. These figures
put limitations on the size of a network that might be served by a LDC controller in real-time.
DDC is appropriate for levels up to grid scale but LDC may be best restricted to small islands
within that DDC grid.
The basic circuitry required for the LDC controller functionality can be described with reference
to a system that in one embodiment includes a wattmeter and Modulator, Dongles, and a House
Controller. These are discussed in more detail below.
1. Wattmeter and Modulator.
Quantity: 1 per system
Location: ideally (but not necessarily) near the point of common coupling to the grid.
Power requirements: self powered
Inputs: 3-phase 3 wire mains supply
Output: Single turn coupling to the neutral wire on the output 3 phase 4 wire system before the
neutral wire is earthed.
Description:
Conceptually this device measures the power taken from the 3-phase 3-wire mains supply and
gives an isolated output. For an experimental version the input is 3 phase, 400 V, 50 Hz,
current 3-4 A. Output scaled 0-3 kW equals 0 – 3 volts.
Included with the device should be a set point input of for example 0-3 Volts = 0 – 3 kW and an
integrator that can take the integral of the voltage difference between the wattmeter and the set
point with an output scaled to be 0-10 V. This 0-10 V signal is to be used to control a V to F
converter working over a range between approximately 300-1200 for example the range 0 V =
600 Hz, 10 V = 1,000 Hz which signal is then used to produce a 1 V signal on the neutral wire of
the 3 phase system. The 1 V signal will be injected on to the neutral wire using a small inverter
and a 100:1 transformer. Ideally the waveform should be a sine-wave but a square wave could
be acceptable.
In a first Laboratory scale prototype all of these functions except the injection of the 1 V signal
(the Modulator) on to the neutral are included in a prototype microprocessor controlled
instrument. This device measures the power in each of the 3 phases (rated 230 v 10 A per
phase), adds the three outputs, compares with a set point, and outputs a square wave with the
correct characteristics for modulating on to the neutral wire. This particular Laboratory scale
system is not suitable for scaling up to a larger 250 kW system.
2. Dongles
Quantity: one per Appliance
As described earlier in this document, Dongles are devices that sit in the power line between the
switchboard on the house and the appliances in the house. Ideally they would be built into the
appliance (i.e. the load), shown in figure 41 in which dongles 42 are provided in appliances 42
and are connected to the electricity supply line 43. In a second embodiment the dongles 42
could be connected directly between an electricity outlet and the or each appliance. A Dongle
connected to an appliance may include appliance specific features. The dongle consists of a
means of detecting the control signal and a means to respond to the control signal for the range
of a parameter. The response will vary, including simple on/off switching to continuously varying
loads. Some dongles may allow reprogramming of their response to the signal. This
reprogramming could consist of a physical selection switch or mechanism or could be controlled
by wired or wireless communication with another device, such as a computer. The Dongle
makes the appliance LDC compliant so that it can operate in the manner required. There are in
principle four types of dongles:
Type A: Simple on/off type. The Dongle isolates the control signal on the neutral wire and
switches on at a fixed frequency and off at a lower frequency. For example the Dongle may
switch on at 720 Hz and off at 660 Hz. Below 660 Hz the Dongle is always off, above 720 Hz it
is always on, and between these two frequencies it is bistable and its state depends on the past
history in the extant application.
Type B: On/off with minimum switching periods. This type of Dongle is suitable for motorized
devices like a fridge or freezer where the number and/or frequency of switching events must be
controlled. Here the device acts like a Type A Dongle but when it is switched on it must stay on
for some minimum period – eg 10 minutes, and when it is switched off it must stay off for a
minimum period – eg 20 minutes.
Type C: This Dongle is fully proportionally controlled. If the control frequency is 750 Hz or
below the Dongle is off, if it is 850 Hz or above the Dongle is on, and between these two
extremes the maximum output allowed varies linearly proportional to the frequency. The
Appliance must be rated for this type of application. A good application is heat pumps or EV
battery chargers.
Type D: This Dongle is similar to Type C but it is not continuously variable but has say eight
separate states. The control frequency 750-850 Hz is divided into eight regions and these
correspond to the operating states. In the lowest region the device is on at 1/8 of full power, in
the next region it is on at 2/8 of full power, and so on. To achieve this requires a compliant
resistive load switching integral cycles in a random sequence to give the correct power output.
3. Dongle Applications
There are two alternatives to the way that Dongles may be used in a House. For example.
3.1 Alternative 1: Non-intelligent Dongles
With this alternative every appliance has its own Dongle which decodes all its own information.
As outlined above the availability of power is encoded on to the neutral wire by a 1-2 V signal
that varies from 750 Hz (no power available for priority loads) to 850 Hz (ample power available)
on top of the mains voltage. The Dongles filter out this signal and use it to switch devices on
and off, or vary them continuously by switching on mains zero crossings, according to the type
of Dongle used – Type A to D. Here all the appliances/controllable loads are in a strict priority
sequence or order and are switched on and off when activated by the control signal. For
example essential loads are active at all times (if not switched off) and do not have a Dongle,
high priority loads might be set to be active for a control signal in the range 750-850 Hz, and low
priority loads might be active if the signal frequency is above perhaps 820-850 Hz. These trip
points will vary with each Dongle but will be set at the time of installation. The trip points will
have some hysteresis - for example a Dongle may switch on at 820 Hz and switch off at 780 Hz
and both of these points are set at the time of installation. Dongles type C and D are also active
all the time taking power proportionally to the control frequency.
Power: Self powered
Signals: 1-2 V 750-850 Hz,
Software programming: small
Measurement capability: none
Programmability: very limited
3.2 Alternative 2: Intelligent Dongles
This alternative has electronic circuitry – a house controller (HC) - that is preferably, but not
necessarily, located in the meter box. It has the capability to decode the modulated signal on
the neutral wire and know what devices are on/off and it can communicate with all the Dongles.
It can also measure the power flow into the house (essentially Amps) but the flow of power to
the Dongles and the appliances is unchanged. Communications to the appliances by the HC
are for example by WLAN at 2.4 GHz or other, and, as before each appliance has its own
Dongle but now each Dongle has its own WLAN transceiver. The HC is able to reprogram the
Dongles on-line so that the priority order of every appliance is continually changing and only the
default setting is set at the time of installation. Each appliance will be able to report on/off
information and load current back to the house Driver. The Dongles will be able to operate as
all four Types as above - in on/off modes with or without delays, or in proportional control modes
as instructed by the HC. The type selection can be done in real time. As before small devices
will be controlled using on/off switching on zero crossings to reduce RFI, while larger ones – hot
water heater, heat pump, electric clothes drier, and electric vehicle charger will operate in a
continuously variable way to give continuously variable control as described above for Type C
and D Dongles. The Dongles will continually update themselves in response to the extant
circumstances so that the power available is always used in an optimal fashion – for example if
a high priority device is physically switched off the power slot that it was taking – say 660-720
Hz will be dynamically re-allocated i.e. the priority for that load has effectively been reassigned.
The intelligent dongles can act interactively with the appliances and the HC over the WLAN
network. For example they may sense a characteristic such as a power requirement of the load
being supplied, so with an EV battery charging load the HC can be aware of the state of charge
and act so that the battery is fully charged by some specific time. Similarly if a drier is being
used the ‘dryness’ of the clothes may be managed so that they are dry when required. Options
like this will incur a higher price for the electricity but add to the versatility of the total system.
Power: self powered from 230 V 50 Hz
Signal: WLAN 2.4 GHz bi-directional, 1-2 V 750-850Hz on the neutral wire.
Software programming: significant to achieve full potential.
Measurement capability: comprehensive
Fault reporting: comprehensive.
The invention may be implemented to allow a large number of households to be incorporated
into an islanded system and be able to prioritise loads without any impediment to individual
households setting their own priorities. A straightforward controller is used to determine when
those loads can be switched on and when they must be switched off. There are clear
advantages in having the largest loads – EV and hot water – with continuously variable outputs
so that they are essentially available at all times to fill in the gaps between the switching on and
off of other loads. Thus, the invention allows EV’s to be charged as a LDC compliant load and
this extends to the operation of those EV’s in an electrified roadway situation. In a wind-
powered system a community can get great benefit by having a wind turbine with a very large
penetration. Excess power can still be exported to the grid but the total load on the grid can be
managed within narrow limits in most circumstances. This same load management also
extends to interest groups with isolated transformers in a city.
In another embodiment of the invention the system controller does not continuously transmit a
signal, but instead the load controllers poll the system controller (or the measuring means
directly) for updated information. In this case the information received by one load controller
may differ from that received by another, for example if there has been a change to the power
draw on the grid between one load controller requesting information and the next one doing so,
or if the system controller adds a unique identifier to the data sent to a particular load controller.
Such systems may be less desirable than those described above due to the potential to
introduce additional latency into the system.
In some embodiments the islanded system may have more than one point of coupling to the
grid, each point of coupling associated with a means for measuring the power drawn from the
grid through the coupling. The control of the load controllers in the islanded system may be
based on an aggregate or average of the power measurement readings. In a variant of this
embodiment, the different points of coupling may be associated with separate islanded systems
whose occupants have agreed to co-operate such that their combined power usage is
compared to a set point.
Another example of a simple LDC system with power feedback and controllable load is shown in
block diagram form in Figure 19. The symbols used are described in the table below:
Symbol Definition
Input power
Power imbalance
Integration time constant that converts kW to kWh
Energy imbalance signal distributed to loads
Conversion constant between energy imbalance and power
Filter time constant
Since here the system power is being regulated, the output of the system is . The input is the
power disturbance which is the difference between the set point and the power that the local
system is consuming - .
This results in a transfer function of:
Here the damping ratio is also:
Again this means the performance of the system is dependent on the integral time constant,
filtering constant and available controllable load. As mentioned earlier in this document, an
important difference between this response and that for DDC is that the system inertia is not
involved. In fact the inertia of the grid makes the whole network stable without having to add
extra. As far as the controller is concerned the damping factor and hence the stability is
dependent on the controller gain and the integrator time constant and these are easily adjusted.
This transfer function is the expected result as any instantaneous change in will be directly
seen at , hence the “1” in the transfer function. This will then be followed by a second order
response as given by the second term in the transfer function.
It can be observed from the system diagrams of both DDC and LDC that they have the exact
same structure. Since DDC is regulating and LDC is regulating , the outputs are different
but the underlying transfer functions are still the same. The huge advantage of LDC is that in
addition to the filtering constants and being configurable, the integral constant can also
be changed without affecting the mechanical machines. This means the speed of response and
the stability are not dependant on any physical properties of the system as it is in DDC.
However, as mentioned above, the disadvantage of LDC is that the control signal L has to be
distributed over the network accurately, without corruption, and with low latency. With DDC this
is not a problem as the system frequency is everywhere the same but here it is more difficult.
In a practical grid environment, a demand response system should be able to react in less than
a second to any appropriate signals or disturbances. This requirement is often specified by the
system operator. The only physical constraint in an LDC system is delays in generating,
distributing, filtering and responding to the LDC signal. The time constant of the filter is designed
to dominate any other delays such that the operation of the system is predictable and stable. As
is demonstrated later, a filtering time constant of is realistic.
The integral constant can then be designed to give an ideal response for a given amount of
controllable load as follows:
For example, if there is 1kW of controllable load ( , the filtering time constant is
and a damping ratio of is desired, an integral constant of will result.
This gives a settling time to within 5% of approximately 0.6s, which is well within the 1s usually
prescribed by system operators.
The operation of another embodiment of an LDC system can be explained with reference to
Figure 20. In that Figure 3 phase power from a grid is measured by a wattmeter and then drives
a local grid that can have transformers on it with multiple housing loads, and generators –
shown here as a wind-turbine. The household loads are on a 4-wire system but generators are
on a 3 wire or 4 wire connection as appropriate. The input 3-phase power is measured and
compared with a grid reference set point. The result of this comparison is integrated and
converted to a frequency control signal that is inductively coupled into the connection between
the transformer star point and the earthed neutral. All the houses are fed from one or more
phases and the phase-neutral voltage that they receive has the frequency signal with it. Inside
the house there are dongles between wall plugs and appliances and these dongles are
sensitive to the low power frequency signal and switch on or off or linearly control the appliance
load as may be appropriate. The load may be controlled with a triac or other bi-directional
switch as required. The household loads may also be prioritised (as described elsewhere in this
document) such that as the availability of power increases and decreases the loads switch on
and off according to their assigned priorities. The priorities may be fixed or variable and even
dynamically variable and may be reassigned as the user requires as often as required – without
limit. These options are shown schematically in Figure 20 (d) with fu and fl designating upper
and lower frequencies for a given priority. The frequencies are detected by detector 60 to
enable latch 61 and gate driver 62 to trigger triac 63 and thus turn the load on or off
(Figure20C). The controller of Figure 20 (c) can be provided in dongle 64 between the
appliance 65 (i.e. the load) and the power point 66. In one example (shown in Figure 20(a)) the
control system is added to an existing transformer, this may require the addition of a
communications system to receive the set-point information and the output of a signal, or the
LDC control signal, to be communicated to the control signal generator. In a second example
the control system could be built in to a transformer. In a third example the control system,
transformer and control signal generator may be combined in a single device.
The control signal, possibly a V to f signal created by the control system, must be small but
capable of spreading through the local network. In one embodiment an inverter is used to
produce a 2-3 V signal capable of 100A or more e.g. a range of 50-500A so that the signal does
not get lost in the network. This signal is inductively couple to the neutral line between the star
transformer and the phase-neutral voltage or ground. This means of placing the control signal
on the neutral wire enables fast communications and reduces the possibility of a break in
transmission. The control signal can control a range of different loads, including digital loads,
linearly variable loads or any other type of load as required. The control signal must be
recoverable in all the dongles on the network in real time so that the LDC control action can be
implemented accurately, without delay to keep the network stable. To do this requires low cost
easily constructed filters that can fit into appliances while taking little space and little power.
These dongles require a special filtering capability as described below.
An analogue communication system has been designed in order to simply distribute the LDC
signal around a microgrid. This design requires the signal be unidirectional, of medium
resolution (< 8 bits) and have very low latency. A system whereby an 800Hz tone is injected at
the star point of the local distribution transformer and picked up and filtered at each load has
been created. This tone is varied by in order to represent the maximum and minimum
LDC signal value. If the tone is at 750 Hz or below all dongle loads are switched off, if it is at
850 Hz or above all user loads may be switched on, and between these two extremes loads
can be switched in a priority sequence.
As shown in Figure 20(a), power supplied to a network may be measured by wattmeter 71. The
difference between the measurement and the set point or reference is integrated by integrator
72 and a voltage to frequency converter 73 for example can be used to produce a control signal
having a frequency dependent on the power available. To inject an 800Hz tone at the star point
of the transformer, an inverter and a transformer together generally referenced 74 are used in
this embodiment. The inverter consists of a 3-phase rectifier, DC Bus, H-Bridge and a 100:1
transformer for isolation and some output filtering. The signal is small to the point where it has
no effect on electrical loads.
One side of the injection transformer secondary is connected to neutral / ground (or earth) of the
network being supplied and the other to the star point of the local transformer. In this way the
800Hz tone can be picked up at any outlet within the system. The frequency of 800Hz is in
th th
between the 15 and 17 harmonic of the mains, is far enough away from 50Hz to be filtered
and yet is low enough to still propagate well through standard wiring. It can be seen that with
the Delta-Star transformer used the tone is a common mode and cannot propagate to the delta
side of the transformer. Thus all local islanded systems or networks connected to the same grid
are independently controlled and there is no leakage from one network to another.
This filter design requires that each controllable load has circuitry for filtering the 800Hz signal
added to the 50Hz mains network supply. The inverter drives a 100:1 transformer and runs off
the same voltage source as the distribution transformer. Given that the inverter input is rectified,
there will initially be a 43dB ( difference between the mains (50Hz) and LDC signal
(800Hz). To reliably pick up this LDC signal, the filter needs to have a relative gain of
significantly greater than 43dB in order to be reliable.
There are numerous filtering designs that could be used to provide this level of performance,
passive networks, active filters and digital filters were all considered. In this example a design
uses a combination of a passive filtering network and a digital filter inside a PSOC. An RC high-
pass filter (HPF)is used to step down and bias the input signal about 2.5V in order to be
accepted by the PSOC. This has the added benefit of attenuating the mains component
significantly more than the LDC component.
The input is first stepped down using a 1:10 resistor divider to a voltage level of <30V in order to
be suitable for standard capacitors. An RC network is then used step down again to a 5V P-P
signal.
The RC HPF uses and has a transfer function of:
An attenuation of -4.0dB at 800Hz and 20.3dB at 50Hz is achieved which gives a 16.3dB
relative gain at 800Hz.
This signal is then suitable for processing with a bandpass filter built from functional blocks
inherent in a PSoC microcontroller. The filter is designed with a centre frequency of nearly
800Hz and a bandwidth of 100Hz. An exact frequency may be difficult to achieve depending on
the PSoC frequency of operation and the division cycles that are available in the processor.
The PSoC has the option of both a two-pole and a four-pole filter. A four-pole filter is achieved
by chaining two two-pole filter stages together.
The transfer function of a two-pole filter is as follows:
For and , a relative gain of 17.8dB is achieved, bringing the total differential
gain between the 800 Hz signal and the mains voltage to 34.1dB. If a second two-pole filter is
used there is another 17.8dB which gives a total of 51.9dB.This shows that a four pole filter is
required in order to reliably differentiate between the two signals. This will give a total of 8.9dB
signal to noise ratio given an initial ratio of -43dB.
As noted above, the band pass filter is realised inside a PSOC microcontroller, which places
constraints on which values can be chosen. Using the PSOC Designer software, it was found
that the following numbers were possible for a nominal desired 800Hz centre frequency and
100Hz bandwidth:
Filter centre frequency: Hz
First pole:
Second pole:
The transfer functions mentioned previously were realised in MATLAB/Simulink in order to
further verify the design. The input signal containing both the 50Hz mains and 800Hz LDC
signal is shown Figure 21. It can be seen that the 800Hz signal is barely noticeable on the
outline of the mains waveform, with small peaks and troughs just visible on close inspection.
A frequency spectrum of this input is shown in Figure 22. The main signal components are of
course the 50Hz mains and the 800Hz LDC signal. It can be seen that there is around -42dB of
relative gain between the mains and LDC.
The high pass filter output shown in Figure 23 brings this relative gain to around -22dB. This is
larger than, but in line with, what was calculated previously. The band pass filter then lifts this
800 Hz signal to +60dB, as shown in Figure 24. Again this is larger than that calculated but not
too dissimilar. These simulations show the filter performance should be at least equal to, if not
better than that calculated manually.
While demonstrating correct performance in a perfect environment is one thing, determining that
the system will work in a non-ideal environment is also important. To this end, wide band noise
was added to the simulation. To measure the exact output frequency, componentry similar to
that which could be implemented in a microcontroller was used. The final band pass filter output
is put through a Schmidt trigger to create a digital signal which can then be timed and filtered.
Multiple software filters were tested in order to find the most suitable method.
With zero noise as shown above, the frequency measured is a perfect 750Hz. Measurements
were taken with noise at andat . Figure 25 shows the input spectrum with
noise of . Figure 26 shows the output spectrum noise of .
Clearly the 750Hz signal is significantly higher in magnitude than the system noise, and should
still be measureable. When the raw measurement output is plotted, the output is somewhat
stochastic as shown in Figure 27. This signal has a mean of 750.078Hz and a standard
deviation of 5.2848Hz.
Using simple 128 sample averaging, a more stable result is generated and is shown in Figure
28. Here the mean is 750.003 and standard deviation 0.1008Hz. 128 sample points were used
as this generates a new data point every 0.16 seconds. This is close to 0.1 and therefore on the
order of the desired filtering delay.
Using a weighted rolling buffer of the same length gives a mean of 749.997Hz and a standard
deviation of 0.0797Hz. This has a lower overall delay and narrower spread than the standard
averaging method and the result is shown in Figure 29.
Using a combination of 16 averaged data points and then 16 weighted rolling average points
gives a better result, achieving a mean of 749.999Hz and a standard deviation of 0.0385Hz.
This technique would not be computationally intensive to implement in a microcontroller.
These graphs show that with of noise in the system, it is still possible to achieve a very
accurate measure of the LDC frequency, with standard deviations of no more than 0.04Hz. It
can also be seen that even with unrealistically high amounts of noise, a relatively high level of
measurement accuracy can still be obtained. With of noise in the system, the
standard deviation of the final method is 0.6142, which is still usable even with this
unrealistically high amount of noise.
A practical system has been tested within a laboratory scale micro grid. The signal injection is
setup as described previously. A 300:4 turn injection transformer was used, with the injection
inverter running off the same voltage as the rest of the system.
The upper and lower frequencies used here were 710Hz and 864Hz. These are right on the
outside of the filters bandwidth, so are used to show the worst case scenario. The 710Hz has
the worst performance as it is not only on the very outside of filter band but is closer to the 50Hz
and consequently further attenuated by the high pass filter. The mains voltage of the system is
shown in Figure 31.
Here there is 42.5dBV of the 50Hz component and 10.625dBV of the 864Hz component, giving
a -31.875dB difference between the two. The high pass filter output is shown in Figure 32.
There is 3.125dBV of 50Hz and -19.375dBV of 868Hz making the new difference -22.5dB
between the two.
The band pass filter outputs for 710Hz and 864Hz are shown in Figures 33 and 34. At 864Hz,
the signal magnitude is 1.875dBV and at 710Hz, the output is -3.125dBV. There is now no
perceivable 50Hz, but there is at least 20dB of clearance to the nearest spectral component.
Since the 50Hz has been eliminated, other spectral noise components must be investigated.
The band pass filter itself has above unity gain within its band, and is therefore capable of
amplifying noise in the system. When the signal injection system is turned off, there is -
.625dBV of 850Hz as shown in Figure 35. With -3.125dBV of actual signal at 710Hz, this
gives 12.5dB as the minimum signal to noise ratio. At 868Hz it is 17.5dB which is significantly
better.
A Schmidt trigger is then used to square up the signal for measurement. The hysteresis band is
designed such that a signal just within the desired band is picked up and the rest ignored.
At 733Hz there is 3.5V of signal and a clean square wave as shown in Figure 36.
At 800Hz there is 4V of signal and a clean square wave as shown in Figure 37.
At 868Hz there is 3.5V of signal and a clean square wave as shown in Figure 38.
At 710Hz there is 2V of signal and gaps in the square wave as shown in Figure 39.
With no signal injection there is 1V of signal and no square wave output as shown in Figure 40.
These results show that within the filtering band, the LDC signal can be picked up reliably and
used for local demand control even in the presence of noise. The two filtering stages provided a
total of over 70dB of differential gain between 50Hz and the 710Hz to 848Hz band. This is
consistent with the results from both the analytical solution and simulation.
The LDC system using a low frequency tone as part of the control loop is a simple direct system
enabling good control as it has very low latency. An alternative method for the future uses an
electronic transformer in place of the conventional 50/60 Hz transformer and eliminates the
need for an 800 Hz or other frequency tone. Here the conventional street transformer forming
the hub of the LDC micro-grid system is replaced with the electronic transformer. In this
electronic transformer the input power typically at 11 kV is rectified to a high DC voltage which is
then switched electronically with a power electronic inverter producing high frequency power at
a very high voltage and a high frequency of perhaps 20 kHz. This power is then transformed
down in a high-frequency transformer to reduce the voltage and a 3-phase (or single phase)
output voltage is synthesized at 50 Hz using another inverter. The system may use a direct AC
to AC conversion or rectify to DC and invert to AC after rectification. The output voltage and all
the converters in the process are reversible so that power may be sent in either direction. But
the output frequency is no longer restricted to be 50 Hz and by controlling this frequency to vary
according to load an alternative control signal for the islanded network may be produced. Thus,
as described, if the frequency is 49.5 Hz all controllable power is switched off and if the
frequency is 50.5 Hz the entire controllable load is switched on, and there is a linear variation
between these two extremes. In one example the islanded system may be connected to the
grid by one or more electronic transformers and the local frequency, generated at the
transformer, may be used as, or as part of the control signal. In this example the tone on the
neutral line, or other communications systems, may not be needed.
At present this method would be more expensive than the method described above for LDC
using a low frequency tone but the high frequency transformer is already relatively lower cost,
smaller, lighter and more efficient than a conventional transformer, and as semiconductor
prices continue to fall the inverter costs will reduce and this method will be cost competitive. At
the terminals the two systems appear identical except that the LDC one has an impressed 800
Hz tone on the utility voltage, and the electronic transformer has its own local frequency. As the
frequency changes any motors on the system will change speed but as the system is fully
reversible transient energy flows will be available from the grid system to enable those speed
changes and stability will not be an issue.
Unless the context clearly requires otherwise, throughout the specification, the words
“comprise”, “comprising”, and the like, are to be construed in an inclusive sense as opposed to
an exclusive or exhaustive sense, that is to say, in the sense of “including, but not limited to”.
It should be noted that various changes and modifications to the presently preferred
embodiments described herein will be apparent to those skilled in the art. Such changes and
modifications may be made without departing from the spirit and scope of the invention and
without diminishing its attendant advantages. It is therefore intended that such changes and
modifications be included within the present invention.
Claims (16)
1. Apparatus for production of a control signal for a demand side electric power supply management system, comprising: 5 means to accept a set point; measurement means to measure power flow into a supply network relative to the set point; means to convert information from the measurement means into a control signal for transmission over the network wherein the frequency of the control signal is indicative of 10 the power available to the network.
2. Apparatus as claimed in claim 1 wherein the network is supplied by a transformer and the measurement means measures the power supplied by or at the transformer. 15
3. Apparatus as claimed in claim 1 or claim 2 wherein the control signal comprises a low voltage signal relative to the voltage of the network.
4. Apparatus as claimed in any one of the preceding claims wherein the apparatus for producing the control signals is capable of sourcing a high current relative to the current 20 required by individual loads supplied by the network.
5. Apparatus as claimed in claim 3 or claim 4 wherein the control signal comprises a signal in the range of substantially 1-3 volts at 50-500A. 25
6. Apparatus as claimed in any one of the preceding claims wherein the control signal frequency is substantially in the range of 300-1200 Hz.
7. Apparatus as claimed in any of the preceding claims wherein the control signal is provided between a neutral line and an earth connection of the network.
8. Apparatus as claimed in any one of the preceding claims wherein the control signal is inductively coupled to the network.
9. Apparatus as claimed in any one of the preceding claims wherein the apparatus derives 35 the control signal by integrating the difference between the measured power flow and the set point.
10. Apparatus as claimed in claim 1 wherein the control signal comprises the frequency of the power supplied over the network. 5
11. Apparatus as claimed in any one of the preceding claims wherein the frequency of the control signal is proportional to the power available to the network.
12. A utility power supply network including apparatus as claimed in any one of the preceding claims.
13. A method of providing a control signal for a demand side electric power supply management system, the method comprising: measuring power flow into a supply network relative to a set point. converting information from the measurement means into a control signal for 15 transmission over the network wherein the frequency of the control signal is indicative of the power available on the network.
14. A method as claimed in claim 13 including varying the set point. 20
15. A demand side electric power supply management system, the system comprising: apparatus as claimed in any one of claims 1 to 10, and; a load controller comprising: priority designation means for designating a priority for one or more loads supplied by the system; 25 frequency detection means for detecting the frequency of the control signal; means to control the one or more loads dependent on the control signal and designated priority assigned to that or each load.
16. A system as claimed in claim 15 wherein the control signal is obtained directly from the 30 network supplying power to the one or more loads.
Publications (1)
Publication Number | Publication Date |
---|---|
NZ615299B2 true NZ615299B2 (en) | 2015-07-28 |
Family
ID=
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10270249B2 (en) | Local demand side power management for electric utility networks | |
US10454271B2 (en) | Local demand side power management for electric utility networks | |
EP2577832B1 (en) | Reactive power management | |
EP2675034B1 (en) | Power flow measurement and management | |
WO2012128626A2 (en) | System for charging the battery of at least one electric vehicle, charger and method | |
JP6701201B2 (en) | Real-time adaptive distributed intermittent power coupled to grid | |
US11982544B2 (en) | Distributed control of energy storage device charging and grid stability | |
JP7000448B2 (en) | A system for distributing locally generated energy to multiple load units | |
US9989949B2 (en) | Grid tied, real time adaptive, distributed intermittent power | |
Lee et al. | Improved grid dynamics using a localized demand control system | |
NZ615299B2 (en) | Local demand side power management for electric utility networks | |
NZ615299A (en) | Local demand side power management for electric utility networks | |
D'hulst et al. | LV distribution network voltage control mechanism: Experimental tests and validation | |
Talpur et al. | Application of demand response and smart battery electric vehicles charging for capacity utilization of the distribution transformer | |
Lee et al. | Communications requirements and performance of a distributed demand control system | |
Aldejohann et al. | Current modulated messages generated by PFC circuits for allocating appliances with management systems in smart grid applications | |
AU2015203539B2 (en) | Power flow measurement and management | |
Gu et al. | Research on Differentiated Load Response Method Considering Operation Costs of Load Clusters and its Real Time Verification Scheme | |
Tushar et al. | Volt-VAR optimization by using electric vehicle, renewable energy and residential load-shifting | |
Joshi et al. | A Review on Up Gradation of Conventional Grid to Smart Grid for Optimal Residential Power Scheduling |