GB2539369A

GB2539369A - Energy recovery system

Info

Publication number: GB2539369A
Application number: GB1504191.6A
Authority: GB
Inventors: Overton George
Original assignee: POWER FLOW ENERGY Ltd
Current assignee: POWER FLOW ENERGY Ltd
Priority date: 2015-03-12
Filing date: 2015-03-12
Publication date: 2016-12-21
Anticipated expiration: 2035-03-12
Also published as: GB2539369B; GB201504191D0

Abstract

An energy recovery system 10 is disclosed. The energy recovery system 10 is for use with an electricity supply including a microgeneration supply 100 and a utility supply 104, and aim is to reduce export of energy to the utility supply 104 by diverting excess generated electricity to a resistive load, for example an immersion heater 114. The energy recovery system 10 includes power output modulating means for modulating the proportion of maximum power supplied at the power output. The power output modulating means cycles between a power-on state and a power-off state, the relative timing of the pulse width modulation being adjusted to control the duty cycle and therefore the average power supplied. The energy recovery system 10 further includes a power measurement module, which is adapted to read input from a current sensor 16 when the power output modulating means is in the power-off state and to read input from the current sensor 16 when the power output means is in the power-on state, and to derive the power supplied to the load 114 from the difference of those measurements. The current sensor 16 measures magnitude and direction of current flow to/from the utility supply 104, and no separate current sensor is provided for directly measuring the current supplied to the load 114.

Description

ENERGY RECOVERY SYSTEM

The present invention relates to an energy recovery system for optimising use of "microgenerated" power, for example solar or wind power.

BACKGROUND TO THE INVENTION

Microgeneration based on, for example, photovoltaic panels or wind turbines is becoming increasingly popular and increasingly economic. This is due to falling costs of the required equipment, especially solar panels, and also due to government subsidies for renewable electricity generation.

In the United Kingdom, owners of renewable microgeneration systems can benefit financially in three ways: A "Generation Tariff" is paid for each kWh of energy generated, regardless of whether it is used at the point of generation or exported to the grid. The generation tariff is effectively a government subsidy to encourage renewable generation. The amount paid depends on a number of factors, but as an example the current generation tariff for many small solar installations is 13.88 pence per kWh.

An "Export Tariff" is paid (in addition to the generation tariff), for each kWh which is exported to the grid. The export tariff is currently 4.77 pence per kWh.

Finally, where electricity is used at the point of generation, the owner avoids buying electricity (importing) from the grid. The cost of electricity varies by supplier and by area, but a typical charge in the UK is around 10 -12 pence per kWh. This is referred to as the "Import Tariff'.

Due to the tariff structure, and in particular the fact that the import tariff is around twice the export tariff, it is generally preferable for owners of microgeneration systems to avoid exporting energy, for example in the daytime when photovoltaic panels are generating but the house is unoccupied, and then importing energy later on, for example in the evening when photovoltaic panels are operating at reduced power or not at all, but when energy is required for heating and lighting. The householder can benefit financially by using energy at the times it is being generated, so that very little energy has to be imported at the expensive import tariff.

Other factors also incentivise "time shifting" use of energy. For example, many household electricity meters are not capable of measuring exported electricity. Pending installation of "smart" meters, microgenerators are paid an export tariff based on the assumption that they will export a fixed proportion (for example 50%) of the electricity generated. If the proportion exported is in fact less than that assumed amount, then the microgenerator will benefit. Also, some householders have entered into agreements in which a third party will finance solar panel installation in return for receiving the generation tariff and export tariff.

The householder therefore receives a benefit only in terms of a reduced amount of electricity which they have to import at the import tariff, and it is in their interests to minimise import and therefore use electricity at the times when it is being generated.

In principle, there are many types of load in typical households which consume energy, but do not need to be operated at particular times. A good example is an immersion heater in a hot water tank. A hot water cylinder can usually store enough hot water for at least a day's usage, and if it is well-insulated then the storage losses are minimal. The immersion heater can therefore be switched on whenever it is most economic to do so, typically in the daytime for a microgeneration scheme involving solar panels, or whenever the wind is blowing where wind turbines are used.

A simple timer can meet this requirement to some extent. For example, a householder could set an immersion heater timer to switch on between 11am and 1pm, when solar panels are likely to be generating at peak performance. If the house is unoccupied during the day, then it is highly likely that the energy demand at those times would otherwise be low, and that energy would be exported only to be bought back later at a higher price.

However, a timer cannot respond to changes in the weather, which affect the performance of sun-and wind-based microgeneration plant, and neither can it respond to changes in demand within the household. In addition, the load can only be "on" or "off" in such an arrangement. A typical immersion heater for hot water is rated at around 3kW, and many small solar installations will only generate this amount of power at nearly full capacity, or even not at all.

All of these factors mean that there is a demand for a so-called "Energy Recovery System" which monitors export or import of energy, and switches a load so that energy is used when it is being generated. Typically, the load is an immersion heater in a hot water tank, since hot water is routinely required in most households and a well-insulated hot water cylinder is an efficient way of storing energy. However, other loads, for example space heaters, can also be switched. If the building is reasonably well-insulated, then electric space heating may be a good way of "shifting" electricity consumption to times when electricity is being generated. Energy Recovery Systems are known. These known devices monitor the amount of power being exported or imported by means of a current clamp on the mains supply, and respond accordingly by switching the load to consume excess power which would otherwise be exported.

Various techniques are used to switch on a load at less than maximum power. For example, a 3kW immersion heater could operate at half capacity and draw only around 1.5kW when that is the amount of available surplus power from the microgeneration plant. All of these techniques essentially rely on turning the load on for short periods and off in between, so that the average power draw over time is reduced as compared to fully-on operation. For example, if the heater is repeatedly turned on for 40ms and then off for 40ms, the total power drawn is only half of the rated power consumption of the load when it is constantly turned on. As long as the switching cycle is fast enough that the capacitors in the inverter do not discharge during the "on" cycle and will recharge during the "off" cycle, no power needs to be imported to supply the load.

Examples of such switching techniques include phase angle control, in which the supply is switched on at a particular (adjustable) phase angle between ±0 and ±180 degrees in the AC waveform, and then switched off again at the zero crossing (leading edge control). Alternatively, the supply can be switched on at the zero crossing and then off again at an adjustable phase angle (trailing edge control). This technique is widely used, and works well for example in lighting dimmers. However, using phase angle control with large loads (e.g. immersion heaters) causes difficulties. This is because a radio frequency interference (RFI) is caused at the switch-on (for leading edge control) or switch-off point (for trailing edge control), and with large loads the magnitude of this interference is unacceptably high.

Pulse-width modulation is an alternative. In this technique, the supply to the load is typically switched at a much higher frequency than the frequency of the mains supply. For example, the switching frequency may be around 20,000Hz. Pulse width modulation produces radio frequency interference in the same way as described above. Although this interference is generally in high frequencies and can be filtered out reasonably effectively as compared with the interference caused with phase angle control, it is still not as 'clean' as a synchronous switching technique described below.

Synchronous control, also known as time-proportional control, avoids the above mentioned problem of RF interference. "Burst-fire integral cycle switching" is used in heating control applications. In this type of control, the "on" and "off" switching event always occurs at the instant when the AC supply is at zero volts. There is therefore no step-change in current and hence minimal radio frequency interference. However, the frequency of switching this integral cycle switching technique is often low enough that, when a large load is switched (large enough that the impedance of the utility supply line is a significant factor), a noticeable flicker can be caused in lighting connected to the same supply, since the peak and RMS voltage of the supply will vary depending on whether the switched load is in an "on" or "off" state. This is generally undesirable in a domestic installation, and indeed synchronous heating controls are rare in the domestic context outside of energy recovery systems for use with microgeneration.

Despite the above-mentioned drawbacks, energy recovery systems generally use the burst-fire integral cycle switching technique to modulate the supply to a resistive load. Once a first threshold level of energy export is detected, the load will be switched on at low power.

The energy export will be re-measured at regular intervals (for example every few seconds), and at each interval the power supplied to the load will be incrementally increased, for example in 5% intervals, until the measured energy export is at a second threshold level. In this way, excess power is supplied to the load as and when it is generated, for example to heat water in a hot water cylinder and avoid the need to import expensive energy later in the day.

Energy recovery systems usually record the amount of energy supplied to the resistive load, and may store aggregate values for a period of time for display on a screen, or export to an external device for example by Bluetooth (RTM) or Wi-Fi. This is an important feature for many consumers, since having invested in the energy recovery device, an owner generally wishes to measure the financial savings associated with its use. Most known energy recovery systems rely on the power rating of the attached load being manually input at the time of installation, so that the energy "diverted" from export to the load can be calculated simply by multiplying the power rating of the load by the amount of time the load is in the "on" state. However, this can be inaccurate for various reasons, not least that the power rating of a load can be inaccurate (for example an immersion heater sold as nominally 3kW might in fact be 3.1kW or 2.9kW), and the load might change or be replaced without updating the rating programmed into the energy recovery system. Also, requiring the installer to enter the power rating of the attached load is an unwanted step in the installation process. Energy recovery systems generally have only a few pushbuttons for input, since they are mainly designed for unattended operation and do not have many "interactive" features. Entering the power rating of the load may therefore be quite a fiddly operation in many cases.

Since the load will usually be a heating load, it will often be controlled by a thermostat. If multiple loads are connected, which can be switched on and off independently (for example, by independent thermostats) then the energy recovery system which is reliant on load size input on installation cannot know the size of the currently connected load. Known systems where the load size is input on installation are therefore generally not suitable for connecting multiple loads, or at least do not correctly record the energy diverted when the load changes.

Some known systems use a "semi-automated" set-up procedure, where the load is connected and switched on during the set-up process, and the system is able to measure the size of the load and store the measurement at that point. This simplifies installation to some extent, and does avoid the problem of inaccurate load ratings, but such systems are still unable to respond to changes in the load over time, and cannot work well with multiple independent loads.

As an alternative, it is possible to directly measure the energy supplied to the load by providing a current clamp on the output power lead, running from the energy recovery system to the load. However, this increases the build cost of the device significantly.

It is an object of the present invention to solve the above mentioned problems.

STATEMENT OF INVENTION

According to the present invention, there is provided an energy recovery system for use with an electricity supply including a microgeneration supply and a utility supply, the energy recovery system including: a power input for connecting to the electricity supply; a power output for connecting to a resistive load; a current sensor for measuring magnitude and direction of current flowing to/from the utility supply; power output modulating means for modulating the proportion of maximum power supplied at the power output, the power output modulating means cycling between a power-on state and a power-off state, the relative timing of the on and off periods being adjusted to control the average power supplied; a controller for controlling the power output modulating means to modulate the power supplied to the resistive load depending on the current measured by the current sensor; a power measurement module for measuring the power supplied to the resistive load; and accumulator means for recording the energy supplied to the resistive load, characterised in that the power measurement module is adapted to read input from the current sensor when the power output modulating means is in the power-off state and to read input from the current sensor when the power output modulating means is in the power-on state, and to derive the power supplied to the load from the difference of those measurements.

In the energy recovery system of the invention, there is no requirement to input the size of the load at the time of installation, since the power supplied to the load is measured by the power measurement module. The measurement is accurate even if the load changes, and it is possible to connect multiple loads which may be switched on and off independently and without communication with the energy recovery system (most commonly, by independent thermostats). The energy recovery system will always keep an accurate record of the energy supplied to the resistive load, so that the owner of the system can accurately calculate the savings made, and the pay-off for their investment in the system.

A current sensor for measuring current flow to the load is not required, and preferably such a current sensor is not provided. Since only a single current sensor is needed, the cost of the system is reduced as compared to systems which use two current sensors for dynamic measurement of power supplied to the load.

Preferably, the power output modulating means switches the power output at zero-voltage points in the AC waveform. This is known as "synchronous" or "zero-voltage" switching. This technique ensures that radio frequency interference (RFI) is not caused by the power modulating means.

Most preferably, half-cycle switching is used. In half-cycle switching, on or off switching can take place on either the rising or falling zero-crossing point in the AC waveform. In this application, half-cycle switching is preferable to the more commonly used integral cycle switching (in which switching only occurs on a rising zero-crossing), because power output can be modulated accurately at a high enough frequency to avoid unwanted import or export, and without causing noticeable flicker in lighting connected to the same electricity supply. With half-cycle switching, the capacitors in a typical PV inverter will be able to supply the full load for the short "on" periods, and will be able to recharge in the "off" periods. The load "seen" by the solar panels is therefore (for example) 1.5kW, where a 3kW heater is operated at 50% duty cycle. Bearing in mind that a 3kW heater is a large load on a domestic supply (which may have an impedance of up to around 0.4 ohms), the mains supply may react noticeably (i.e. the voltage may drop) when the load is switched on. The voltage drop may be enough to cause lights to noticeably dim, and with repeated switching this can result in a flickering effect. Half-cycle switching however reduces noticeable flicker since the switching is at a higher frequency as compared with integral cycle switching.

Another advantage of half-cycle control is that energy is drawn from the capacitor banks of the PV inverter in shorter bursts. This reduces the amount of heat dissipated in the inverter and in the energy recovery system itself, which would otherwise waste valuable energy.

Thyristers and triacs may be used for controlling the switching in the power output modulating means, and ensuring that switching takes place only at zero crossing points as described above. These devices by their nature switch off at zero current, and hence RFI is minimised.

The current sensor may be in the form of a split-core current transformer (SCT), which can be positioned around the live wire of the utility supply, where it enters the consumer unit. Preferably a burden resistor is provided across the secondary coil, and the voltage across the burden resistor may be measured. The voltage across the burden resistor is proportional to the current in the primary coil (which is the utility supply), and the current flow in the utility supply can therefore be derived from the voltage measurement. The current sensor may further include a pair of half-wave rectifiers, one which passes positive currents and one which passes negative currents. The output from the negative-passing rectifier may then be inverted. The output from each rectifier (in the negative case the inverted output) can be converted to a desired voltage range for measurement.

For example, a typical voltage range which may be sampled for measurement by a digital circuit is 0 to 5 volts. By splitting the output from the current sensor into a positive and negative side, twice the effective measurement resolution is available, since 0 to peak current can be scaled to 0 to 5 volts, rather than scaling negative peak to positive peak to 0 to 5 volts, as is common in known systems. It is found that a rectifier based on an op-amp voltage follower is suitable in this application, although diode rectifiers or other designs may also provide acceptable performance.

Clearly, in any AC utility supply, current will flow in alternating directions in the live utility supply line. However, "direction of current flow" is understood to be relative to the sign of the instantaneous supply voltage. When energy is being imported, the current will be broadly in phase with the supply voltage. As the amount of energy being imported falls, the current will fall. Under no load conditions the current is zero. If energy starts to be exported, then the current will be negative (i.e. in the opposite direction). In other words, the current is the opposite sign of the voltage, that is the current is 180 degrees out of phase with the voltage. The mains voltage can be measured at the power input, and compared to the current measured at the current sensor, to determine the "direction of current flow" to or from the utility supply. Note that the absolute magnitude of the mains voltage does not need to be measured for this purpose -only the sign of the voltage is needed. However, to keep a record of the total power 'savings' (i.e. the power diverted from export to the load), an absolute magnitude is required. This absolute magnitude may be measured, for example at the power input, or may be assumed to be a constant value, which is adequate in many applications.

It will be understood that the current in fact may lag (or, less commonly, lead) the voltage by a few degrees, depending on the power factor of the load. However, this effect is insignificant and it will be clear whether the current is approximately in-phase (importing energy) or approximately 180 degrees out of phase (exporting energy).

Preferably, the controller maintains a time slice period, which is an integer number of cycles. For example, in a 50Hz system the time slice period may be 200ms, or ten full cycles. Changes in power output by the power output modulation means may be synchronized to the start of a time slice -that is, during any given time slice, the power output is modulated to a particular proportion of maximum output, and that proportion will not change during the time slice. During a particular time slice, the controller may measure the magnitude and direction of the current at the utility supply, and make a decision as to how the power will be modulated in the next time slice. For example, if power is being exported at the current modulation level, then the power output modulation may be altered in the next time slice to supply more power to the load. Preferably, increases in supply to the load are incremental. For example, if 5% of full power is being supplied to the load in one time slice and the exported power is measured above a certain threshold, then in the next time slice 10% of power may be supplied, even if the measured export would allow 50% to be supplied without importing. If the level of export remains above the threshold with 10% being supplied, then in the next time slice 15% may be supplied and so on. However, decreases in supply can preferably take place in larger increments. For example, if during one time slice power import from the utility supply is detected, then the supply may be cut off altogether in the next time slice. In turn, this may result in power export and then an incremental increase in the power supplied. This technique of incremental increase and block decrease reduces to an absolute minimum the amount of power which is imported to supply the load.

During every time slice, the magnitude and direction of current to/from the utility supply is measured as described above, both during cycles when the load is switched on and cycles when the load is switched off. As an example, where the power is modulated at 50% the load may be switched on every other half cycle. Different modulation levels have more complex switching patterns which are described in greater detail below, but in all cases, since the modulation is controlled by the controller, the controller can correlate the modulation scheme with the measurements from the current sensor, to determine which measurements are taken when the load is on and which are taken when the load is off. The current flowing in/out of the utility supply can be compared, and the difference between the current flow measured during off periods and that measured during on periods is used to directly infer the current supplied the load. Since the voltage is easily measured, and is generally fairly constant anyway, power supplied can be calculated by P=IV, and the total power supplied to the load can be accumulated and preferably displayed on a screen. In some embodiments, the RMS magnitude of the mains voltage may be measured so that the power can be derived from that measurement. In other embodiments, the RMS voltage may be input at installation time or even when the device is manufactured. Because the variation of mains supply voltage is generally small, this may provide for suitably accurate measurements in many installations.

The power supplied to the load cannot be measured by taking differences in the way described when the power is modulated at 100%, since in this case there are no "off" periods for comparison. However, because the power supplied to the load is preferably increased incrementally, the system can make use of a measured value which was saved at the last increment before 100% power (for example, 95%). In addition, the controller may be configured to measure any change in the absolute magnitude of the current in the utility supply, and if that change exceeds a certain threshold then the power modulation means can be returned to the greatest increment lower than 100% (for example 95%) to update the saved setting. It is found that a suitable threshold in many cases is 100W. This enables the system to take into account changes in the load, which might be caused at any time for example if the load is in fact made up of several heaters each controlled by their own independent thermostat. The system of the invention is able to keep an accurate record of the energy diverted from export to the loads, in spite of changes to the load.

DESCRIPTION OF THE DRAWINGS

For a better understanding of the present invention, and to show more clearly how it may be carried into effect, a preferred embodiment will now be described for the purposes of example only, with reference to the accompanying drawings in which: Figure 1 shows a schematic diagram of an energy recovery system according to the invention, connected to an electricity supply and to a load including electric heaters; Figure 2 is a table showing the on and off duty cycles during a single time slice, which are applied to the load by a power output modulating means which forms part of the energy recovery system in Figure 1; Figure 3 is a timeline showing the sequence of operations that take place in a controller which forms part of the energy recovery system in Figure 1.

DESCRIPTION OF THE PREFERRED EMBODIMENT(S)

Referring firstly to Figure 1, an energy recovery system according to the invention is shown in the schematic at 10. The energy recovery system includes a power input 12 and a power output 14.

The power input 12 is shown connected to an electricity supply. The electricity supply includes a microgeneration supply, which in this embodiment is an array of solar panels 100. The solar panels are connected to the electricity supply via an inverter 102, which converts the DC supply from the solar panels 100 to AC output for distribution. The electricity supply further includes a utility supply 104. The utility supply can supply electricity when the solar panels 100 are not generating, or when the demand exceeds the supply from the solar panels 100. This is referred to as "importing" electricity. On the other hand, when the solar panels 100 are generating in excess of demand, power can flow back into the utility supply 104. This is referred to as "exporting" electricity.

The utility supply 104 and the inverter 102 are connected together at a consumer unit 106.

Note that a function of the inverter 102 is to synchronise the generated AC waveform with the utility supply 104.

Various loads are connected to the consumer unit. As examples, a socket outlet 108 and lamp 110 are shown. It will be appreciated that any number of loads can be connected to the consumer unit 106, essentially in accordance with the normal loads found in a domestic setting. However, a panel heater 112 and an immersion heater 114 have been connected to the power output 14 of the energy recovery system 10, rather than directly to the electricity supply via the consumer unit 106.

The energy recovery system 10 further includes a sensor means 16 for measuring the magnitude and direction of current flowing to / from the utility supply 104. In this embodiment, the sensor means 16 is a current clamp on the utility supply cable. The sensor means may be connected to the rest of the energy recovery system 10 by a cable or by wireless means.

The basic function of the energy recovery system is to monitor the power export/import to/from the utility supply 104, and to supply power to loads 112, 114 in order to reduce power export where possible. At any time, the power generated by the solar panels 100 may change (for example, due to changes in cloud cover), the power demand by the directly connected loads 108, 110 may change (e.g. if someone switches on/off an electric kettle or indeed any other appliance) and the loads 112, 114 may change (e.g. if the hot water tank reaches its set temperature then the immersion heater 114 will be disconnected by a thermostat).

The energy recovery system 10 continuously monitors the power export/import to/from the utility supply 104, and can therefore adjust the power supplied to the loads 112, 114 to take account of these changes.

To adjust the power supplied to the loads 112, 114, the energy recovery system includes power output modulation means, which includes triacs or thyristers for switching the load synchronously. Specifically, half-wave synchronous switching is used to control the amount of power supplied to the loads 112, 114. The power modulation means operates in 200ms time slices, which is equivalent to 20 half cycles at 50Hz. The power output can therefore be modulated in 5% increments so that the loads 112, 114 operate at 0% power, 5% power, 10% power, or any other 5% increment up to 100% of maximum power. The power output is modulated by adjusting the number of half cycles within a time slice for which the power output is switched on. For example, for 5% power, power is supplied for only one half cycle in the 200ms time slice. For 50% power, power is supplied for ten out of twenty of the half cycles in the time slice. The table in Figure 2 shows the switching pattern used in this embodiment. As an example, the table shows that for a 10% power setting, power will be supplied at half cycles 6 and 13, i.e. two out of the twenty half cycles in the time slice. In other words, triacs will be switched on during half cycles 6 and 13, but off during all other half cycles.

The power modulation means with half-wave switching minimises radio frequency interference, does not produce noticeable flicker in lighting, and is also compatible with almost all equipment which may be found in a typical domestic installation. For example, dimmer switches using either leading or lagging phase angle control can form part of the load 108, 110. It is generally envisaged that the loads 112, 114 will be heaters of one sort or another, but any substantially resistive load is suitable.

The energy recovery system 10 includes accumulator means for storing a record of the total amount of energy supplied to the loads 112, 114. This can be used to estimate the total financial savings realised by the use of the unit, bearing in mind the difference between the import and export tariff as described above. For this calculation, the current supplied to the loads 112, 114 needs to be known. However, no current sensor is provided on the power output 14 due to the need to reduce build cost. The current supplied to the loads 112, 114 is therefore inferred by taking measurements of the current in the utility supply 104. The measurements are made both during "on" periods and during "off" periods. For example, if the load is currently at 10% power, then the utility supply current 104 is measured when the load is "on" during half cycles 6 and 13, and when the load is "off" during other time slices. From the difference of these measurements, the current to the loads 112, 114 may be inferred. The voltage can be measured at substantially any point in the system, since any voltage drop between, for example, the utility supply 104 and the power input 12 will be negligible. For convenience, mains voltage is measured at the power input 12 in this embodiment. Alternatively, the voltage can be assumed to be a constant value, for example 240V. In many installations, the utility voltage supply variability is very small, and any change from the assumed value will be negligible.

The energy recovery system includes a controller, which makes various measurements, controls the power modulation means and calculates the power supplied. Figure 3 shows the sequence of operations carried out by the controller in this embodiment. It is noted that the sequence of operations occurs over a 400ms time period and is synchronised with the power output modulation means so that a controller sequence takes place over two full time slices of the power output modulation means.

During operation A -"Data Acq", the import / export current is read from the current sensor 16. Measurements are taken during "off" cycles (as per Figure 2 and the current power output level) and during "on" cycles. Measurements are taken at multiple times during each half cycle to ensure an accurate reading of the RMS current, bearing in mind that the AC waveform may contain distortions/interference. Squared values of the current sensor readings are stored in four accumulators: import acc stores import values when the load is off export acc stores export values when the load is off import acc2 stores import values when the load is on export acc2 stores export values when the load is on During operation B, the values in each accumulator are converted to floating point RMS, taking into account the number of cycles when the load was switched off and the number of cycles when the load was switched on. The net flow with the load off can be calculated from the difference of the RMS values of the import acc and export acc accumulators, and the net flow with the load on can be calculated from the difference of the RMS values of the import acc2 and export acc2 accumulators. The total net flow in a time slice can be calculated from the all four RMS values. The RMS current supplied to the load can also be inferred from the difference between the net utility flow with the load on and the net flow with the load off.

During operation C, other operations are performed which are not described here in detail. For example, the display on an LCD screen may be updated.

During operation D, the power setting is calculated for the next sequence. The power setting for the next sequence is determined by the total net import/export to/from the utility supply as calculated in operation B. If there is a net export of more than, for example 200W, then the power setting for the next sequence is determined to be 5% more than the current power setting. If there is a net export of less than, for example, 25W, or a net import, then the power setting for the next sequence is determined to be 10% less than the current power setting. If there is a net export in the example range of 25W to 200W, then the power setting remains the same. Of course, other thresholds and increments are possible. In particular it is sometimes preferable to impose a maximum number of increments per time slice, for example, one 5% increment increase every ten time slices or one 10% increment decrease every two time slices. In general it is preferable to increase slowly, but decrease power when necessary at a higher rate to minimise unnecessary and expensive import. The target range of 25W to 200W has been found to work well in typical solar installations, bearing in mind typical changes in solar supply and household demand.

During operation E, a measurement of the mains voltage is taken. In some embodiments, this step is not included as the mains voltage is simply assumed to be constant.

Nonetheless, taking regular measurement makes for more accurate calculation of the power supplied to the load. Using the RMS current supplied to the load as determined in operation B and the RMS voltage of the mains as measured in step E, the power supplied to the load may be calculated by P = IV. The energy supplied is given by E = IVt, and is typically displayed on a display screen in kWh units. In some embodiments, a financial saving in pounds and pence (or indeed any other currency) may be calculated based on the number of kWh units diverted from export to the loads 112, 114.

A boost feature is provided, which can be activated to switch on the loads 112, 114, irrespective of whether power needs to be imported to do so. This is useful since sometimes hot water will be required immediately, and the boost feature obviates the need to provide a separate immersion heater for short-notice demands. The boost feature in its most basic form is simply a switch or other input on the energy recovery unit, which when pressed turns on the loads 112, 114 for a period of time, for example one hour. The boost feature is implemented as an interrupt routine in the controller. However, the controller will always wait until the start of the next time slice before any change is applied. There will therefore be a slight delay in activating a boost feature, but such a delay is unlikely to be noticeable to most users.

Claims

CLAIMS1. An energy recovery system for use with an electricity supply including a microgeneration supply and a utility supply, the energy recovery system including: a power input for connecting to the electricity supply; a power output for connecting to a resistive load; a current sensor for measuring magnitude and direction of current flowing to/from the utility supply; power output modulating means for modulating the proportion of maximum power supplied at the power output, the power output modulating means cycling between a power-on state and a power-off state, the relative timing of the on and off periods being adjusted to control the average power supplied; a controller for controlling the power output modulating means to modulate the power supplied to the resistive load depending on the current measured by the current sensor; a power measurement module for measuring the power supplied to the resistive load; and accumulator means for recording the energy supplied to the resistive load, characterised in that the power measurement module is adapted to read input from the current sensor when the power output modulating means is in the power-off state and to read input from the current sensor when the power output means is in the power-on state, and to derive the power supplied to the load from the difference of those measurements.
2. An energy recovery system as claimed in claim 1, in which no current sensor is provided which directly measures the current at the power output.
3. An energy recovery system as claimed in claim 1 or claim 2, in which the power output modulating means switches to/from the power-off state from/to the power-on state when the instantaneous voltage of the electricity supply is substantially zero.
4. An energy recovery system as claimed in claim 3, in which the power output modulating means switches between the power-off state and the power-on state on both rising and falling zero-crossing points in the AC waveform.
5. An energy recovery system as claimed in any of the preceding claims, in which the power output modulating means includes thyristors.
6. An energy recovery system as claimed in any of the preceding claims, in which the power output modulating means includes triacs.
7. An energy recovery system as claimed in any of the preceding claims, in which the current sensor is in the form of a current transformer.
8. An energy recovery system as claimed in claim 7, in which the current sensor further includes a burden resistor.
9. An energy recovery system as claimed in claim 7 or claim 8, in which the current sensor further includes a pair of half-wave rectifiers, one of the half-wave rectifiers passing positive currents and the other passing negative currents.
10. An energy recovery system as claimed in claim 9, in which the half-wave rectifiers include op-amp voltage followers.
11. An energy recovery system as claimed in any preceding claim, further including voltage measuring means.
12. An energy recovery system as claimed in claim 11, in which the voltage measuring means measures the voltage of the electricity supply at the power input.
13. An energy recovery system as claimed in any of the preceding claims, in which the controller maintains a time-slice period, the power modulating means not changing the output level other than at the start of a time-slice period.
14. An energy recovery system as claimed in any of the preceding claims, in which the controller is adapted to monitor the current sensed by the current sensor during periods when the power output modulating means is outputting power at 100% of maximum power, and in which the controller is adapted to reduce the power output whenever the absolute magnitude of the current sensed changes by more than a predetermined threshold value.
15. An energy recovery system substantially as described herein, with reference to and as illustrated in Figures 1 to 3 of the accompanying drawings.