GB2618535A - Multiple Load Energy Recovery Controller - Google Patents

Abstract

The energy recovery controller measure excess export powers to a utility supply 100 from an electricity supply which includes the utility supply 100 and a microgeneration supply, e.g., roof-mounted solar panels 102. The energy recovery controller includes a single CT clamp 12, and outputs for connecting to multiple storage charging loads 108, 112. Storage charging loads may include for example an EV charger 108 and an immersion heater 112. The energy recovery controller 10 can control how much excess export power is diverted to each storage charging load 108, 112. This is said to avoid the poor functionality of systems in which multiple CT clamps, one for each storage charging load, are attached to a single cable from the utility supply (fig 1).

Description

MULTIPLE LOAD ENERGY RECOVERY CONTROLLER

The present invention relates to a controller for controlling multiple loads to use excess energy generated by on-site microgeneration.

BACKGROUND TO THE INVENTION

On-site microgeneration is becoming increasingly popular. This is mainly in the form of solar photovoltaic panels which can be mounted, for example, on the roof of a residential building. However on-site microgeneration also includes for example small wind turbines, hydroelectric generation from a flowing stream, and even small combined-heat-and-power (CHP) boilers powered by natural gas.

On-site microgeneration is any means of generating electricity which is connected on the consumer side of the electricity meter.

VVhere on-site microgeneration is installed, it is preferable to use the electricity at the times when it is being generated. For example, solar PV generates the most electricity in the middle of the day, when the sun is high in the sky. Wind turbines generate electricity when the wind is blowing, hydroelectric generators may generate more electricity after heavy rainfall, and CH P boilers generate electricity when there is heat demand from a space heating system. Most on-site microgeneration generates electricity to some extent intermittently and unpredictably.

At times of excess generation, electricity can be "exported" to the supply network, i.e. sold to an electricity company to be used by other buildings. At times of excess load, electricity can be "imported" from the supply network, i.e. bought from the electricity company to be used on-site. However, the price for which electricity can be sold is generally significantly less than the cost of buying the same amount of electricity at a different time. Therefore it is economically advantageous for the owner of the on-site generation to use the electricity at the times it is generated, rather than selling the excess and then buying back at a higher price at a different time.

It is also environmentally advantageous to use the electricity at the time it is generated, since times of excess generation will generally be roughly synchronised across all owners of similar on-site generation across the country. Obviously, nobody's solar panels generate electricity at night, and solar panels across the country peak during the middle of the day. Consuming electricity rather than exporting it will therefore mean using more electricity at times during the day when the overall mix of electricity in the system has a greater proportion of renewables. At times of high renewable generation there is not necessarily enough transmission capacity to make use of all the available renewable power where it is needed. The proportion of renewables actually being usefully used will therefore be increased if electricity is used in the place when it is generated, at the time it is generated. The corollary of that is that the amount of polluting and non-renewable generation required to make up the balance of the demand will be reduced.

There are many loads which can make use of electricity but do not necessarily require a supply all of the time. For example, electric vehicle (EV) chargers can be turned on and off to store electricity in the electric vehicle battery when the electricity is being generated, and stop charging when doing so would require importing electricity. It is common for electric vehicles to be plugged in for long periods, for example during the working day or overnight, and so it does not necessarily matter if the charging is turned off at times as long as a sufficient amount of overall "on" time is achieved before the vehicle needs to be used.

Likewise, an immersion heater for heating hot water in a well-insulated cylinder can be turned on and off throughout the day, depending on the availability of free solar electricity, as long as the total "on" time achieved is sufficient to heat enough water to meet demand within the household. Even space heating, if the building is sufficiently well-insulated, can usefully warm the building while excess electricity is available to provide comfort a few hours later. Well-insulated fridges and freezers can also be turned off for periods of time as long as the "off" period is short enough to keep the temperature within acceptable bounds.

Batteries for storing excess electricity to be used later by any load are also increasingly being installed. These batteries can be charged when there is excess electricity from on-site microgeneration, and discharged later to avoid the need to import expensive electricity.

Such loads which store energy in any useful way for later use are referred to generally as "storage charging loads". Among other things, any kind of electric heater (whether a space heater or a water heater or both) may be a storage charging load. In particular heat pumps are increasingly popular as energy-efficient space and/or water heaters.

Devices known as "energy recovery systems", "immersion controllers", or "optimisers" detect when excess electricity is being generated, and automatically "divert" that excess energy to a useful source, for example an immersion heater, space heater, EV charger, fridge, freezer, space heater. An example of such a device is described in GB2539369. An energy recovery system uses a current sensor to detect excess generation, and modulates a power supply to a load accordingly. Sometimes, the energy recovery system is integrated into the load device. For example, EV chargers are available which integrate the current sensor and switching logic so that the EV charger can be configured to charge the vehicle when excess electricity is available from on-site microgeneration.

Most energy recovery systems use a current transformer (CT clamp) in order to measure the amount of energy flow through the mains supply cable of the building (i.e. the amount of power being exported to the supply network or the amount of power being imported from the supply network). During times when excess power is flowing back onto the supply network from onsite generation (export), the energy recovery system then uses this information to turn on or increase the power supplied to a storage charging load in order to divert and capture the excess energy available. In isolation with a single storage charging load, this works well.

The problem lies when multiple storage charging loads from various manufacturers are connected, each using their own CT clamp and proprietary measurement system, to measure and divert available export energy. For example, a household may have an EV charger with an integrated CT clamp and controller, an immersion heater controller with its own CT clamp as well, and a storage battery which again has its own CT clamp. VVhere multiple devices are connected the result can be that the devices "fight" each other for the excess energy. It can be unpredictable which device will "win" and illogical and undesirable results could occur, for example the electric car is charged even though nobody plans to drive it for days, but the hot water tank is cold when everyone in the house wants to take a bath in the evening.

Central control of these multiple load systems does not currently exist. There is no way to set or determine which device gets priority over another.

It is an object of the present invention to solve this problem. 30 STATEMENT OF INVENTION According to the present invention, there is provided a multiple-load energy recovery controller for use with a plurality of storage charging loads connected to an electricity supply, the electricity supply including an on-site microgeneration supply and a utility supply, and each of the storage charging loads having an input for connection to an export sensor in the form of a CT clamp for measuring power flow in a cable, and each of the storage charging loads being adapted to charge by drawing power from the electricity supply according to the power flow measured at its associated input, the energy recovery controller comprising: a main CT clamp for measuring current flowing to/from the utility supply; a plurality of outputs, each output being suitable for connection of the input associated with a different one of the storage charging loads; a signal generation module for generating a signal on each of the outputs according to the current measured by the main CT clamp, and according to signals from a control module.

The control module may control the signal presented at each of the outputs so that each connected storage charging load "sees" an apparent current export which is a proportion of the total current export. For example, if the main CT clamp measures an export of lkW then the controller may present signals equivalent to, for example, 40% of this (400VV) at a first output which is connected to an EV charger, 40% (400'A) at a second output which is connected to an immersion heater controller, and 20% (200VV) at a third output which is connected to a storage battery.

By using the energy recovery controller, multiple storage charging loads, potentially from different manufacturers and with no inherent compatibility with each other or ability to communicate with each other, may be connected to the same supply and may use excess electricity from microgeneration in a controllable way.

Preferably, the signal generation module includes a potential divider having a plurality of taps between an input from the main CT clamp and a signal ground. For example, a potential divider comprising five resistors of the same value in series has six taps at 100%, 80%, 60%, 40%, 20% and 0% of the signal input. Hence a fraction of the signal input is presented at each tap from the potential divider.

A buffer, for example a voltage follower, may be provided between the main CT clamp and the potential divider so that the potential divider does not load the input, in effect the input to the potential divider behaving roughly as an ideal voltage source with the magnitude of the voltage being proportional to the export / import power, and the phase indicating the direction (in the sense that in export conditions the voltage waveform will be about 180 degree out of phase compared with the voltage waveform in import conditions).

A multiplexer may be provided between the taps of the potential divider and each of the plurality of outputs. The multiplexer is used to select which of the taps from the potential divider is connected to each output. The multiplexer is controlled by the control module. A signal generation module as described, using potential dividers and multiplexers, essentially generates a signal by following the signal of the main CT clamp with the voltage buffer / voltage follower, and splitting it up with a potential divider. It is possible to generate the signal in other ways, for example by sampling and digitising the signal from the main CT clamp and then generating outputs in software. This alternative has the potential advantage of more flexibility, i.e. not being limited to a number of discrete steps based on the number of resistors in the potential divider.

However, the waveform being measured in a typically installation can be very noisy and this requires a high sample rate. A dedicated, and potentially quite expensive, processor may be needed to perform the task effectively. Using analogue electronics as described results in a very accurate reproduction of the signal, which is "copied" rather than "reconstructed", with a low cost and simple implementation.

Further buffers, for example voltage followers, may be provided between the taps of the potential dividers and the multiplexers, or alternatively between the outputs of the multiplexers and the outputs of the energy recovery controller. This prevents the potential divider being loaded by the CT clamps associated with storage charging loads, which would affect the voltages at the taps.

The control module may include an input of the total measured export power (i.e. from the main CT clamp via the buffer). The control module may have inputs from sensors and other sources, for example room temperature sensors, a water temperature sensor in a hot water cylinder, a battery level of a storage battery or EV battery. User settings, configurable for example via a web interface, will determine how the sensor inputs are used to control the connected energy storage loads. For example, user settings may tell the controller to prioritise heating hot water with an immersion heater up to a power of 1000W, until the temperature of the water in the cylinder is at least 60°. Priorities of different loads may depend on inputs, for example charging an EV could be high priority if its current charge level is less than 40%, a lower priority if its current charge level is 40% -60%, and a lower priority still if its current charge level is over 60%. The priority could be increased by user input, for example if the user needs to make a long journey at the end of the day.

The energy recovery controller may further include means for measuring the power drawn by each storage charging load. Such means may include for example a CT clamp on each cable connecting a respective storage charging load to the electricity supply.

The control module may control the signal generation module according to the power drawn by the storage charging loads. In particular, the control module may control the signal generation module to reduce the apparent export power presented to the storage charging load according to the actual power drawn by the storage charging load. In an example case there may be an export power of 1kW. The control module sees this and, according to its configuration, presents 50% of it, i.e. 500W to an immersion heater controller, one of the storage charging loads connected to an output of the energy recovery controller. The immersion heater controller, seeing an apparent export power of 500W, switches on and modulates to draw a power of 500W. With this extra load, the export power to the utility supply is now half what it was, i.e. 500W and the immersion heater is drawing all the power it was supposed to draw. However, if the CT clamp input of the immersion heater controller is still connected to the same tap from the potential divider then the immersion heater controller will "see" a remaining export of 250W. If that happens then the immersion heater controller will probably increase its power consumption since it will be designed to use any available export power above a small minimum. To avoid this, when the power drawn by the immersion heater controller is 500W (or within a range, for example 500W ± 10%), the control module will control the signal generation module so that it appears to the immersion heater controller that there is no export power left to use. Thus the immersion heater controller will continue to draw about 500W, leaving the other 500W of export power to be used according to configured rules (for example, to charge an EV). The means for measuring the power drawn by each storage charging load therefore allow the control module to present to each load the change in available export power which the load would "expect" if it were the only such load in the system measuring the net export power to the utility supply.

BRIEF DESCRIPTION OF THE DRAWINGS

For a better understanding of the present invention, and to show more clearly how it may be carried into effect, reference will now be made by way of example only to the accompanying drawings, in which: Figure 1 is a schematic of a prior art electrical system including a microgeneration supply, a utility supply, and a plurality of storage charging loads; Figure 2 is a schematic of an electrical system having the same storage charging loads as in Figure 1, connected to an energy recovery system according to the present invention; Figure 3 is a schematic showing in particular the signal generation module of the energy recovery system of Figure 2.

DESCRIPTION OF PREFERRED EMBODIMENTS

Referring firstly to Figure 1 an electricity system is shown schematically. The electricity system includes a utility supply 100 and a microgeneration supply 102, in this case solar photovoltaic panels. An electricity meter 104 measures electricity imported from the utility supply 100 (and may in some cases also measure electricity exported to the utility supply 100). A consumer unit! distribution board 106 connects the utility supply 100 to the microgeneration supply 102, and also connects loads such as lighting circuits, socket circuits, cookers, etc. Some of these loads may be "storage charging loads" such as electric vehicle (EV) chargers and immersion heaters, as discussed in further detail below.

A "storage charging load" is in principle any electrical load which stores energy in a useful way. For example an EV charger can store electricity in the electric vehicle's battery, for use later when the EV is driven. An immersion heater stores energy as heat in a water cylinder. Even a space heater may be considered a storage charging load in a suitably well-insulated building, since the space can retain the heat and remain at a comfortable temperature perhaps for a few hours after the heater is switched off.

In Figure 1 an EV charger 108, heat pump 110, immersion heater 112, and AC-connected battery 114 are all shown as storage charging loads. Each is connected to the AC distribution board and can draw power from the electricity system when it is switched on. These devices also generally have the characteristic that the power draw can be modulated. For example an immersion heater may be rated at 3kW, but by modulating the supply electronically it can be set to draw any amount of power from zero up to that amount.

Of these, the AC-connected battery 114 can of course not only be charged but discharged, and therefore forms in effect both a storage charging load (when charging) and a microgeneration supply (when discharging). In addition, in Figure 1 the solar PV system includes a DC-connected battery, and so the solar PV system with DC battery can also be controlled to, in effect, generate different amounts of power at different times, to some extent independently of the sunlight! weather conditions.

Each of the storage charging loads and the solar PV with DC battery is provided with its own CT clamp 102a, 108a, 110a, 112a, 114a. Each current clamp is disposed to surround the main cable between the meter 104 and the consumer unit 106. A current clamp when clamped around the main cable forms a current transformer, with the main cable forming one turn on the primary, and the secondary coil of the current transformer being used for measurement. A burden resistor is placed in parallel with the secondary coil (the burden resistor typically forms an integral part of the current clamp) and the current in the main cable is measured indirectly by measuring the voltage across the burden resistor. Each storage charging load 108, 110, 112, 114 and the solar PV 102 is in principle trying to measure how much current is flowing to or from the utility supply 100, and in which direction it is flowing (whether it is importing or exporting energy).

Each of the storage charging loads 108, 110, 112, 114 may be made by a different manufacturer, and it is likely that there is no possibility for them to communicate with each other. Such devices are typically designed with the assumption that they will be the only storage charging load in an electricity system. If there is just one storage charging load connected, for example only immersion heater 112, then the current clamp 112a associated with the immersion heater 112 will be able to measure when there is energy export and control immersion heater 112 to be turned on to use up the excess energy. Thus the excess generated electricity is used to usefully heat the water in the water cylinder rather than being exported to the utility supply 100 at small, zero or even negative value. More importantly, the water does not have to be heated later with more expensive imported electricity.

However, with all of the current claims 108a, 110a, 112a, 114a, all of the connected storage charging loads 108, 110, 112, 114 will detect excess export energy at the same time. Which storage charging loads turn on to store electricity is unpredictable because there is essentially a race condition, where what happens depends on the response time of each device. The result may be not the most useful to the householder, for example the electric vehicle may charge up even though it does not need to be driven for days, and the water cylinder may not be heated even though hot water is required that day. As a result, in this example, expensive imported electricity would end up being used to heat the water.

Figure 2 shows a similar system, but using the energy recovery controller of the invention. The utility supply 100, meter 104, consumer unit 106, microgenerafion supply 102 and storage charging loads 108, 110, 112, 114 are all connected together in exactly the same way. Other loads, for example ordinary sockets 122 and lighting circuits 120 are connected to the consumer unit 106 in a conventional way, as shown in Figure 3.

However, current clamps are not connected to the same main cable to the utility supply 100, but instead the inputs 102a', 108a', 110a', 112a', 114a' designed for use with the current clamps (102a, 108a, 110a, 112a, 114a) are connected to outputs of the energy recovery controller 10. The energy recovery controller 10 includes a main current clamp 12, a single current clamp which measures the power being exported or imported to/from the utility supply 100. The energy recovery controller 10 has a web-based user interface and can be configured to set rules to dictate charging priorities depending on time of day, sensor inputs, manual overrides, and any number of other factors. The energy recovery controller 10 generates signals at its outputs, and the outputs are connected to the current clamps 102a, 108a, 110a, 112a, 114a of the storage charging loads. Hence the energy recovery controller 10 can measure the actual export current to the utility supply, and present a fraction of that signal to each of the storage charging loads depending on rules. If the rules in force, for example, say that the immersion heater 112 and EV charger 108 should receive half of excess electricity each, but that until the hot water cylinder is hot and the EV is charged no other loads should be turned on, then the energy recovery controller 10, on detecting an export power for example of 1kW, will present a signal equivalent to 500W to the current clamp 112a of the immersion heater 112, and present a signal equivalent to 500W to the current clamp 108a of the EV charger 108. Thus the storage charging loads will not "fight" for the excess power with an unpredictable "winner", but will be controlled to share out the excess power in a predictable and controlled way.

Referring now to Figure 3, the workings of the energy recovery controller (10) are shown in more detail. The main CT clamp includes a secondary coil 12 and a burden resistor 16 in parallel with the coil 12. Note that the burden resistor is typically an integral part of the CT clamp. The burden resistor associated with main CT clamp 12 is shown expressly in Figure 3, but for clarity burden resistors of other CT clamps are not shown. However, the CT clamp including the burden resistor can be considered as a device for measuring current in a wire, with the output being a voltage signal proportional to the current being measured.

The output of the main CT clamp 12, 16 is connected to the input of a unity gain buffer / voltage follower 18. This buffers the CT clamp 12, 16 and prevents the sensor being loaded, and therefore affected, by connected devices.

The output of the unity gain buffer 18 is then connected to a voltage divider. The voltage divider in this example includes five resistors 20a, 20b, 20c, 20d, 20e. Each resistor has the same value. A divider of five resistors has six taps, i.e. outputs from the potential divider. The six output taps are indicated at 22a, 22b, 22c, 22d, 22e, 22f. It will be seen at output tap 22a the output will be 100% of the output from the buffer 18. At output tap 22b the output will be 80% of that signal, at output tap 22c it will be 60%, at output tap 22d 40%, at output tap 22e 20%, and at output tap 22f (signal ground) zero.

Multiplexers 24a, 24b are provided. There are as many multiplexers as there are outputs for connecting storage charging loads. In Figure 3, for clarity only two storage charging loads are shown and hence two multiplexers. However, it is envisaged that embodiments may have outputs for connection of about six storage charging loads or even more.

In this embodiment, each of the multiplexers 24a, 24b has six inputs, and each multiplexer has its inputs connected to the six output taps 22a, 22b, 22c, 22d, 22e, 22f of the potential divider. Hence each multiplexer output can be connected to a selectable one of the six output taps.

The multiplexers are controlled via select lines 26a, 26b (3-bit digital signals) from control module 28.

The output of each multiplexer is connected to the input of a unity gain buffer 30a, 30b. The output of each unity gain buffer 30a, 30b is effectively an ideal voltage source which can be connected to a respective input 112a, 108a' of a storage charging load 112, 108.

In Figure 3 an immersion heating controller 112 has an input 112a measuring the output voltage of buffer 30a, which is selectably connected to one of the output taps 22a, 22b, 22c, 22d, 22e, 22f of the potential divider. An EV charger 108 has an input 108a' measuring the output voltage of buffer 30b, which is selectably connected to one of the output taps.

By controlling the select lines 26a, 26b of the multiplexers, the control module is able to present a proportion of the signal corresponding to the export power to each of the storage charging loads 108, 112.

The control module may use a variety of inputs 32 in making control decisions. Inputs include manual commands or configurations, for example a requirement to prioritise charging the EV because it is needed for a long journey. Inputs of this type could be generated automatically in some embodiments by linking for example to diary / scheduling software. Inputs 32 also include sensor inputs, in particular sensors which measure how much energy is already stored in the energy stores associated with each storage charging load. For example, sensors may measure how hot the water is in the water cylinder associated with immersion heater 112, how much charge is in the battery of the EV associated with EV charger 108, the room temperature where one of the storage charging loads is a space heater, etc. In addition, the control module has an input 34 from the main current clamp 12, 16 (buffered by buffer 18) so that the control module can measure how much energy is being exported. The control module may have further current sensing inputs 36, 38 for measuring the amount of energy actually being consumed by each storage charging load 108, 112. This allows the control module to adjust the output presented to each storage charging load 108, 112 so that it responds in the expected way to the load being turned on. For example, if total export energy of lkW is detected then the control module may present a signal equivalent to 500W at the output corresponding with load 108. Load 108 will then switch on, modulating its power draw according to the 500W which appears to be available. However, once load 108 is drawing 500W the control module needs to detect this and present a signal to the load 108 indicating that no more power is available. This can be done by switching modulator 24b to connect buffer 30b to the zero tap output 22f. Once in a steady state where the total exported energy is 500W and load 108 is drawing 500W, modulator 24a can be switched to present a signal indicating available power to load 112.

It will be seen that the loads being controlled can potentially be controlled to draw power more finely than the number of resistors in the potential divider would immediately suggest. For example, in Figure 3 only five resistors are shown with taps corresponding to 0%, 20%, 40%, 60%, 80% and 100% of the signal. Another embodiment can be envisaged with ten resistors to give 10% steps which would allow 500W of a 1kW detected export to be presented as explained above. However, even with only five resistors the fact that the storage charging loads will generally try to take all the available power that is apparent on their input can be exploited to achieve finer control. For example, if 1kW export is detected and the program indicates that load 108 and load 112 should receive half (50%) of the power each, then the energy recovery controller can proceed as follows: First, 400W is presented to load 108 by using the 40% tap of the potential divider. This will cause load 108 to modulate to draw about 400W of power, which is detected by the controller with current clamp 38. The export has now reduced to 600W. The load can now be presented with 120W of apparent power by using the 20% tap of the potential divider. Thus load 108 will "see" about 120W of available export power, and increase its power draw to 520W. The controller will then connect the sensing input 108a' of load 108 to the 0% tap so that the load 108 sees no more available export power.

The load is drawing about 520W which is within 4% of the target 500W, much closer than the 20% error which would be suggested by the steps at 400W or 600W available from the five-resistor potential divider.

Even better control can be obtained in more steps. For example if 200W is first presented and drawn by the load, 800W remains. 20% of that is 160W which can be made available, for a total draw of 360W. This leaves 640W of which 20% is 128W, adding up to a total of 488W which is only 12W or 2.4% away from the target 500W.

Finer still control could also be achieved by adding a small number of extra resistors towards the lower end of the potential divider. For example, if resistor 20e were replaced with two resistors each having half the value, that would create a 10% step in addition to the existing steps at multiples of 20%.

Storage charging loads respond to the apparent available power at different speeds and so the strategy to control a storage charging load to draw the target amount of power may be different for different loads. In some embodiments, the controller may learn the response characteristics of the storage charging loads, and improve its control strategy over time.

The energy recovery controller allows for multiple energy storage loads to be connected in a system, and controlled in a predictable way which takes into account the usefulness of storing energy in various different ways according to sensor inputs (e.g. temperature) and configured priorities. This results in a much better use of available excess power when it is available, and hence will reduce the amount of expensive and generally non-renewable energy which has to be imported from the utility supply.

The embodiments described above are provided by way of example only, and various changes and modifications will be apparent to persons skilled in the art without departing from the scope of the present invention as defined by the appended claims.

Claims (14)

1. CLAIMS1. A multiple-load energy recovery controller for use with a plurality of storage charging loads connected to an electricity supply, the electricity supply including an on-site microgeneration supply and a utility supply, and each of the storage charging loads having an input for connection to an export sensor in the form of a CT clamp for measuring power flow in a cable, and each of the storage charging loads being adapted to charge by drawing power from the electricity supply according to the power flow measured at its associated input, the energy recovery controller comprising: a main CT clamp for measuring current flowing to/from the utility supply; a plurality of outputs, each output being suitable for connection of the input associated with a different one of the storage charging loads; a signal generation module for generating a signal on each of the outputs according to the current measured by the main CT clamp, and according to signals from a control module.
2. 2. An energy recovery controller as claimed in claim 1, in which the signal generation module includes a potential divider having a plurality of taps between an input from the main CT clamp and a signal ground.
3. 3. An energy recovery controller as claimed in claim 2, in which a buffer is provided between the main CT clamp and the potential divider.
4. 4. An energy recovery controller as claimed in claim 3, in which the buffer is a voltage follower.
5. 5. An energy recovery controller as claimed in any of claims 2 to 4, in which a multiplexer is provided between the taps of the potential divider and each of the plurality of outputs.
6. 6. An energy recovery controller as claimed in claim 5, in which buffers are provided between the taps of the potential divider and the multiplexers.
7. 7. An energy recovery controller as claimed in claim 5 or claim 6, in which buffers are provided between the outputs of the multiplexers and the outputs of the energy recovery controller.
8. 8. An energy recovery controller as claimed in claim 7, in which the buffers are voltage followers.
9. 9. An energy recovery controller as claimed in any of the preceding claims, in which the control module receives an input from the main CT clamp.
10. 10. An energy recovery controller as claimed in any of the preceding claims, in which the control module receives an input from a sensor indicating the level of charge in an energy store associated with at least one connected storage charging load.
11. 11. An energy recovery controller as claimed in claim 10, in which the sensor is a battery level sensor.
12. 12. An energy recovery controller as claimed in claim 10, in which the sensor is a temperature sensor.
13. 13. An energy recovery controller as claimed in any of the preceding claims, in which means are provided for measuring the power drawn by at least one of the connected storage charging loads.
14. 14. An energy recovery controller as claimed in claim 13, in which the means for measuring the power drawn is at least on CT clamp.