FIELD OF THE INVENTION This invention relates to heat pump systems, in embodiments for combined heating and cooling, which are powered by grid mains electricity. More particularly the invention relates to improved management of the use of grid mains electricity by such systems. BACKGROUND TO THE INVENTION

We have previously described, in our co-pending UK patent application number 151 1907.6 filed 8 July 2015, combined heating and cooling systems with improved efficiency, in embodiments using carbon dioxide as a working fluid. These systems included a thermal storage tank, in particular a stratified thermal storage tank, which facilitated control of the system based upon stored energy rather than on temperature per se.

We now consider techniques for improving management of the electrical power supply to such systems.

SUMMARY OF THE INVENTION

According to a first aspect of the invention there is therefore provided a grid-mains powered heat pump system with controllable grid-mains power consumption, the system comprising: a refrigerant circuit including a compressor having a compressor motor, the refrigerant circuit having a raised temperature portion following said compressor and a reduced temperature portion prior to said compressor; a heating circuit thermally coupled to said raised temperature portion of said refrigerant circuit, said heating circuit including a thermal storage module, wherein said thermal storage tank comprises instrumentation to measure thermal energy stored in said thermal storage module; an internal power supply bus coupled to said compressor motor via a motor driver, coupled to a grid mains input via a mains front end, and coupled to a rechargeable battery via a bi-directional battery interface; and a system power controller, the system power controller having one or more inputs to receive: a stored electrical energy signal representing a level of stored charge in said rechargeable battery; a stored thermal energy signal representing a level of thermal energy stored in said thermal storage module; and a demand control signal indicating a desired reduction in grid mains power consumption; and having at least one control output to control power provided from said rechargeable battery onto said bus; wherein said system power controller is configured to limit grid mains power consumption in response to said demand control signal, wherein said operation of said limiting is dependent upon a combination of said stored electrical energy signal and said stored thermal energy signal.

In broad terms the inventor has recognised that in a heat pump system of the type described, where there is stored heat (and/or stored coolth), this can be exploited to manage the mains electricity requirements of the system. More particularly the mains power drawn can be controlled, for example, in response to a demand control signal which may be an external control signal, for example provided by the mains electricity supply/distribution system. An electricity distribution system may provide such a signal in order to temporarily reduce demand from the grid by reducing the power consumption of end-users, and this in turn can be used to keep the amount of electricity demand within national/local power grid limitations.

Embodiments of the system thus provide both electrical and thermal storage, and balance electrical and heat flows into/out of this storage in order to manage the overall power consumption of the system. The skilled person will appreciate that such "balancing" may be performed in a number of ways, although some particularly preferred approaches are described below. In broad terms, when a desire for demand control is indicated by the demand control signal then, if the electrical storage is lower than a threshold, the system will ration the heat pumping capability. The degree of rationing is dependent upon the degree of thermal storage in the system, in particular more rationing being applied if the thermal storage is below a threshold.

In embodiments the thermal storage module comprises a thermal storage tank, more particularly a stratified thermal storage tank (which may comprise multiple individual tanks). In principle, however, other thermal storage media may be employed. In some preferred implementations the techniques are applied to a combined heating and cooling system of the type we have previously described in GB151 1907.6 (ibid). Thus in embodiments the working fluid (refrigerant) circuit may also include a gas cooler and an evaporator. The heating circuit may be thermally coupled to the working fluid circuit via the gas cooler and, in embodiments, a cooling circuit is also provided, thermally coupled to the working fluid circuit, in particular via the evaporator. Although preferred embodiments of the system we have previously described employ carbon dioxide as the working fluid (refrigerant) the techniques we describe here are not limited to a particular refrigerant and may, for example, also be employed with propane. Embodiments of the technique may be applied both to heat pumps that recover their coolth output, for example for use in food conservation, air conditioning and the like, and also in heat pumps that reject their coolth output, for example in to the ground, a water course, air, and the like. In practice the compressor may be implemented as a group of refrigerant compressors having an aggregate output.

The system power controller will typically comprise a processor coupled to memory storing processor control code, and to working memory and having one or more sensors/signal inputs, and one or more control outputs. Inputs to and outputs from the controller may be wired and/or wireless and may be multiplexed over a bus or network. In embodiments the level of charge stored in the rechargeable battery may be sensed by the battery interface and/or using separate sensing circuitry. The level of thermal energy stored in the thermal storage module may be sensed by one or more sensors. The demand control signal may be received from an external source, for example over a network. In embodiments the system power controller may be implemented using a programmable logic controller. (PLC). Typically the heat pump system also includes a thermal system controller, which may be implemented in a corresponding manner to the system power controller. Optionally both these controllers may be implemented on the same or shared hardware, and in embodiments are in communication with one another.

In embodiments the heat pump system includes a power supply architecture to manage power within the system. In embodiments this architecture comprises a power supply this coupling together a mains power supply input, a compressor motor power output, a heating circuit pump output to a heating circuit pump, and a cooling circuit pump output to a cooling circuit pump (where the cooling circuit is present). The bus is also connected to a secondary power source. In the above described aspect of the invention the secondary power source comprises a rechargeable battery, coupled to the power supply bus via a bi-directional power converter. Additionally or alternatively, however, the secondary power source may comprise a source of renewable energy, such as a source of electrical power from one or more solar panels. In embodiments the compressor motor(s) and pump(s) are driven by variable frequency drive (VFD) motor controllers. These enable the speed/power of the drive to be adjusted in response to a control signal. As the skilled person will appreciate, in a typical system there may be other components coupled to the power supply bus, but typically the compressor motor(s) has/have the largest power consumption. Preferred embodiments of the system employ a dc power bus. Embodiments of the system use one or more high voltage batteries, but lower voltages may also be used, for example for safety reasons. In some preferred embodiments of the system the power supply bus comprises a dc bus, preferably operating at a relatively high voltage, for example greater than 100 volts, 150 volts, 300 volts, or 500 volts. In embodiments this dc supply may be obtained by rectifying the grid mains input (which may be single phase or three phase). In preferred embodiments the dc bus is coupled to the grid mains via an active front end (AFE) or passive front end, which may also provide isolation between the dc bus and grid mains. In embodiments the interface between the grid mains and power supply bus provides control over the power flow from the mains onto the bus, for example by controlling the current through the interface. Optionally the interface between the bus and grid mains may be a bi-directional interface, to allow energy to be provided back to the grid.

In embodiments the bi-directional battery interface between the rechargeable battery and the bus may control power flow into/out of the rechargeable battery, for example by controlling a current through the interface. In other embodiments the power into/out of the battery (charging/discharging) may be controlled indirectly, for example by controlling a voltage of the dc bus (this may be performed by the grid mains interface). Optionally the voltage (dc or ac) of the power supply bus may be monitored to provide such control. The skilled person will appreciate that there are many different configurations which may be employed for the power supply bus - for example in another approach a 3-wire dc bus (+, - and 0) may be employed, where 0 volts may correspond to grid mains neutral. The particularly configuration selected for an application will depend upon trade-offs such as system cost, component availability and the like. Optionally in embodiments the power supply bus may be used for supplying power to one or more devices or systems external to the heat pump, such as lighting, in particular dc lighting such as LED (light emitting diode) lighting.

In broad terms, embodiments of the system may be employed to reduce mains power consumption, or to balance mains power consumption by the heat pump system over a period of time such as a day or week. Thus additionally or alternatively to controlling (limiting) power consumption in response to a demand control signal, more generally the system, or demand control signal, may be employed to control a profile of power consumption from the grid mains over time, by employing a combination of the electrical and thermal storage. This "balancing" of power consumption may thus additionally or alternatively be employed to reduce power consumption for a period or periods when power is less "green" and/or more expensive, at the expense of increasing power consumption at some other times. Here "green" is used to refer to a degree to which the power generated is provided from a renewable or low-carbon source. Electricity may be categorised as green or not green with reference to a threshold; in practice whether the grid mains is (sufficiently) green may be defined by a signal, for example from an electricity generator or distributor (network operator) and/or may be determined by time, for example according to a known variation over a period of a day, week, month, year, or the like.

In embodiments the system power controller is configured to limit the power drawn by the compressor motor(s) when the "capacity" remaining in the rechargeable battery is less than a threshold level. Here the "capacity" is preferably expressed in terms of the duration for which the rechargeable battery is able to run the compressor motor(s), this duration being represented by the stored electrical energy signal (which may be combined with predetermined information regarding the motor power requirements). In embodiments the effect of limiting the power drawn by the compressor motor(s) is to limit the flow rate through the refrigerant compressor (or the flow rate of an aggregate output of a group of refrigerant compressors). In embodiments the power drawn by the compressor(s) may be limited by limiting a maximum speed of operation of a compressor motor, for example to a particular number of revolutions per minute (rpm). In embodiments such a power limit signal overrides a thermal control system/signal, in particular overriding a signal to replenish the stored heat in the thermal storage module (and/or stored coolth, where present). This may, for example, limit a maximum temperature achievable by the heating circuit; optionally the heating circuit may also be controlled, for example by means of one or more controllable valves, to limit or turn off non-essential heaters driven by the circuit.

Typically in embodiments of the heat pump system the one or more compressor motors dominate the electrical power consumption but, as described above, additional motors such as pumps and the like may also be fed from the internal power bus. Thus, optionally in embodiments of the system, the operation of one or more of these other devices, for example of one or more pumps of the system, may also be limited in a corresponding manner to the compressor. Similarly the two stage limiting process described below may also be applied to these other devices. Embodiments of the heat pump system include a thermal system controller (which may be implemented using the same physical hardware as the system power controller) to control thermal operation of the heat pump system. We describe later details of some preferred embodiments of a control scheme for a system incorporating a heat store and/or coolth store. However efficient thermal operation is preferably achieved during a normal thermal control mode of operation by controlling the compressor, in response to the stored thermal energy, to maintain the level of stored thermal energy within a first, stored heat target range. In a system as described above which includes electrical storage the system power controller may control or interact with the thermal control mode to invoke a limited thermal control mode of operation of a system, to limit power drawn by the compressor. In this way efficient thermal control may be achieved whilst maintaining stable running at low power and, in embodiments, maintaining a more constant overall electrical power consumption (fewer/reduced peaks/troughs). As previously described, embodiments of the heat pump system may also include a cooling circuit, thermally coupled to the reduced temperature portion of the refrigerant circuit. One or more sensors may be employed to sense stored coolth in the system. Then the thermal system controller may, in its normal thermal control mode of operation, further control the compressor motor in response to the sensed stored coolth to maintain the stored coolth within a second, stored coolth target range. In embodiments this stored coolth control may also be relaxed when the system is operating in its limited thermal control mode of operation. Some preferred embodiments of the system provide two power limiting modes, a first, major power limiting mode and a second, minor power limiting mode with less power limiting than the first mode, for example a less significant limit on the compressor power or flow (an increased cap or threshold). Then one or other of these modes may be selected dependent upon a combination of both the stored thermal energy signal and the stored electrical energy signal. More particularly the system power controller may react to the demand control signal to select the moderate power limiting mode (minor limit) where the ability to run the system from the rechargeable battery is below a threshold ("capacity") but where the stored thermal energy is greater than the threshold level. The system power controller may select the more severe, major power limiting mode where the capacity to run the system on the stored electrical power is similarly limited but where the level of stored thermal energy is less than the threshold. In embodiments the threshold level of stored electrical energy is determined by the capacity of the stored electrical energy to run the system for a threshold duration of time. This time duration may be the time interval to the next update of the demand control signal - that is, in embodiments, the stored electrical energy is evaluated to establish whether this stored energy is sufficient to run the system until the next demand control signal update (which may then indicate that there is no longer any need for the system to run in a limited grid power mode). Additionally or alternatively the demand control signal may define a duration of a control event. Optionally the demand control signal may also define a target or upper limit for power to be drawn from the grid, in which case this may be used to limit the grid power drawn by the system as previously described, in particular by limiting grid power by relying on the combination of stored thermal and stored electrical energy and, if necessary, also limiting the power drawn by the compressor motor(s).

In embodiments where it is determined that the stored electrical energy is below the threshold the system power controller may (progressively) reduce power delivered from the battery to the internal power bus, in embodiments until no power is being delivered from the battery. In broad terms, if the battery is low, power drawn from the battery is ramped down; this may result in power from the grid mains increasing, even during a period when the demand control signal indicates a desired reduction in grid mains power consumption. Thus, in order to at least partially compensate for this, the operation of the compressor is limited, in particular at a value dependent upon the level of thermal energy stored in the thermal storage module (tank). The skilled person will appreciate that, as previously mentioned, determination of the stored electrical energy may be direct (for example, by a measurement performed on the battery), or indirect (for example inferred from the behaviour of the system/bus).

In some preferred embodiments the system power controller is configured to limit the power drawn by said compressor motor(s) conditional upon the stored thermal energy signal indicating less than a threshold level of stored thermal energy in the thermal storage module. In broad terms, in embodiments battery power draw monitoring/control may only be invoked where it is determined that the thermal store is low/insufficient.

Where, during a demand control event, it is determined that the stored electrical energy is sufficient to run the system for greater than a threshold duration (for example until the next update of the demand control signal), the thermal control operates in its normal mode of operation, that is without a cap on the power drawn by the compressor. In this case preferably the power delivered (more precisely, deliverable) by the rechargeable battery on to the internal bus may be increased towards a maximum, to thereby minimise power drawn from the grid mains. The skilled person will appreciate that in such a situation the power drawn from the internal power bus, and hence from the battery, will be dependent upon the power drawn by the compressor variable frequency drive. Typically, however, there will be some power drawn from the grid mains even when the battery output is at a maximum (and thus the power drawn from the battery is effectively controlled by controlling the power permitted to flow through the battery interface on to the bus).

In preferred embodiments the system power controller controls the bi-directional battery interface to charge the rechargeable battery at times when a demand control event is not indicated by the demand control signal. More particularly, however, in some preferred embodiments the system power controller controls charging of the battery, more particularly a charging rate of the battery, dependent upon one or both of the "greenness" of the grid mains supply and the cost of electricity from the gird mains. In this context greenness refers to the level of carbon usage in generating the grid mains power supplying the system. Whether or not the energy supply is "green" may be determined by an external signal, for example from an electricity generator or distributor (network operator), and may indicate a low carbon supply in a binary fashion or may indicate a degree of "greenness" of the electricity supply. Potentially, depending upon how the supply is generated, an indication of date and/or time of day may be used to infer a level of "greenness" of the supply - for example where a network operator provides a significant proportion of power from solar panels. Thus, in embodiments, the system includes an input to receive a carbon level signal. The system power controller may then restrict the charging rate in response to this signal indicating greater than a threshold level of carbon usage in generating the grid mains power and the demand signal indicating no desired reduction in grid mains power consumption, and may derestrict the charging rate in response to this signal indicating less than a threshold level of carbon usage in generating the mains power and the demand signal indicating no desired reduction in grid mains power consumption. Additionally or alternatively the system may be configured to increase power drawn from the grid in response to an external signal indicating a surplus grid mains power availability (for example when there is excess renewables supply). In embodiments this is achieved by controlling the battery interface to store charge in the rechargeable battery.

As previously described, the demand control signal may be from an external source such as an electricity supplier/distributor. Thus the system may include an electricity supplier/distributor interface for one or more such grid-mains electricity suppliers/distributors; this may for example comprise an Internet or WAN (Wide Area Network) interface. As the skilled person will appreciate, the particular form of the interface will depend upon the signal/data format provided by an electricity supplier/distributor.

Additionally or alternatively the demand control signal may be from a user interface for a "local" operator (where here "local" is contrasted with, for example, a supervisor or coordinator of multiple such systems). Such a user interface may comprise, for example, a local terminal or a mobile terminal such as a smartphone or the like, or a web interface. In this way a local operator may control or override the system operation. Embodiments of the system also provide a "fleet control" mode of operation. With such an arrangement a plurality of heat pump systems is provided and together with a fleet control server, coupled to each of the systems, to coordinate the operation of the systems. In embodiments the fleet control server may be configured to provide a respective demand control signal for each of the plurality of grid-mains powered heat pump systems, for example in response to a fleet control signal. The fleet control signal may be, for example, a demand control signal from the grid, more particularly from an electricity supplier/distributor interface. Additionally or alternatively the fleet control server may have a user (supervisor) interface (local or remote) to enable a system supervisor to control a fleet of the heat pump systems together. In this way an overall limit may be applied to the grid mains power drawn by a group of heat pump systems, enabling a significant level of control, more particularly limitation, on grid power demand. In embodiments the server may allow a user to define which systems belong to a group of heat pump systems which are controlled together. The user interface may enable one or more such groups of systems to be defined.

Although we have described above some preferred embodiments of the system which include both a refrigerant circuit and a heating circuit, the invention also contemplates systems as previously described in which the heating circuit is omitted. In such arrangements the function of the heating circuit thermal storage module (in embodiments, a stratified thermal storage tank) may be substituted by thermal storage provided by a refrigerated chamber (and excess "hot energy", for example at a temperature above ambient, may be rejected to waste). Thermal energy of this general type may be referred to by those skilled in the art as coolth, and thus in embodiments the stored thermal energy signal may comprise a measure of stored coolth in the refrigerate chamber. The stored coolth may be a function of a difference between two temperatures, a first, higher/reference temperature (which may be the ambient temperature), and a second lower temperature; more particularly it may be a measure of the thermal energy removed (from the chamber) in lowering the temperature of the refrigerated chamber below the higher (ambient) temperature. Thus in embodiments one measure of the stored coolth may comprise this difference in temperatures, or it may simply comprise the lower temperature if an ambient temperature is assumed.

Thus in a related aspect the invention provides a grid-mains powered heat pump system with controllable grid-mains power consumption, the system comprising: a refrigerant circuit including a compressor having a compressor motor, to cool a refrigerated chamber, wherein said refrigerated chamber comprises instrumentation to measure coolth stored in said refrigerated chamber; an internal power supply bus coupled to said compressor motor via a motor driver, coupled to a grid mains input via a mains front end, and coupled to a rechargeable battery via a bi-directional battery interface; and a system power controller, the system power controller having one or more inputs to receive: a stored electrical energy signal representing a level of stored charge in said rechargeable battery; a stored thermal energy signal representing a level of said coolth stored in said refrigerated chamber; and a demand control signal indicating a desired reduction in grid mains power consumption; and having at least one control output to control power provided from said rechargeable battery onto said bus; wherein said system power controller is configured to limit grid mains power consumption in response to said demand control signal, wherein said operation of said limiting is dependent upon a combination of said stored electrical energy signal and said stored thermal energy signal.

In embodiments the power drawn from the grid may be limited to zero by the demand control signal. As previously described the demand control signal may be generated in response to control of a user interface, and/or may be programmed or otherwise defined to occur during one or more time windows - for example to limit the power drawn from the grid during these intervals. Additionally or alternatively however the signal may be generated automatically, for example when there is a temporary absence of grid mains supply. In one example application, an embodiment of this system is installed in a refrigerated electric vehicle, grid power drain being reduced to zero when the vehicle is mobile. The heat pump system rechargeable battery may be shared between the refrigeration system and the vehicle drive system - for example a battery used to drive the vehicle may also be used for refrigeration. Optionally, therefore, the demand control signal may also be used to control the battery power drain by the system whilst the vehicle is mobile.

In embodiments the system power controller limits power drawn by the compressor motor from said internal power supply bus when the stored electrical energy signal indicates less than a threshold stored electrical energy, more particularly less than a threshold (predicted) duration the stored electrical energy will last whilst the system is running partially or completely off-grid. The power drawn by the compressor motor may be limited by limiting the degree to which the compressor (motor) is able to ramp up, in particular when the stored thermal energy signal indicates that there is less than a threshold level of coolth stored in the refrigerated chamber - for example because one or more temperatures in the refrigerated chamber have risen above a set point which would normally trigger compressor ramp-up.

The skilled person will appreciate that the previously described features of aspects and embodiments of the invention may also be employed in the presently described aspect of the invention, although the relevant description is not repeated for conciseness.

In a further related aspect the invention provides a method of controlling a grid-mains powered heat pump system to control grid-mains power consumption, the system comprising: a refrigerant circuit including a compressor having a compressor motor, the refrigerant circuit having a raised temperature portion following said compressor and a reduced temperature portion prior to said compressor; a heating circuit thermally coupled to said raised temperature portion of said refrigerant circuit, said heating circuit including a thermal storage module, wherein said thermal storage tank comprises instrumentation to measure thermal energy stored in said thermal storage module; and an internal power supply bus coupled to said compressor motor via a motor driver, coupled to a grid mains input via a mains front end, and coupled to a rechargeable battery via a bi-directional battery interface; the method comprising: inputting a stored electrical energy signal representing a level of stored change in said rechargeable battery; inputting a stored thermal energy signal representing a level of thermal energy stored in said thermal storage module; inputting a demand control signal indicating a desired reduction in grid mains power consumption; and controlling power provided from said rechargeable battery onto said bus; and wherein said controlling comprises limiting grid mains power consumption in response to said demand control signal, wherein said limiting is dependent upon a combination of said stored electrical energy and said stored thermal energy.

In embodiments the method may include corresponding aspects to those described above in relation to the heat pump system. Thus, in particular, embodiments of the method include a fleet control mode to control a fleet of the grid-mains powered heat pump systems in a coordinated manner such that each is provided with a corresponding respective demand control signal and the systems work together to limit power drawn from the grid mains by the group of heat pump systems. Such a fleet control mode may be controlled by an external demand control signal, say from the grid, and/or by a fleet supervisor. The invention also provides processor control code to implement a method/system/ power controller as described above, for example on a general purpose computer system or on a digital signal processor. The code is provided on a non-transitory physical data carrier such as a disk or programmed memory. Code and/or data to implement the method/system may comprise source, object or executable code in a conventional programming language (interpreted or compiled), or assembly code, or code for a hardware description language). As the skilled person will appreciate such code and/or data may be distributed between a plurality of coupled components in communication with one another. In a further aspect the invention provides a combined heating and cooling system, the system comprising: a working fluid circuit comprising a compressor including a compressor motor, a gas cooler, and an evaporator; a heating circuit thermally coupled to said working fluid circuit; and a cooling circuit thermally coupled to said working fluid circuit; wherein said heating circuit further comprises a heating circuit pump; wherein said cooling circuit further comprises a cooling circuit pump; and a power supply architecture, the power supply architecture comprising: a mains power supply input; a compressor motor power output to said compressor motor; a heating circuit pump output to said heating pump; a cooling circuit pump output to said cooling pump; and a power supply bus connecting interfaces to said mains power supply input, to said compressor motor power output, to said heating circuit pump output, and to said cooling circuit pump output; wherein at least one of said heating circuit and said cooling circuit includes thermal storage; and further comprising a secondary power source coupled to said power supply bus. In a still further aspect the invention provides a combined heating and cooling system, the system comprising: a working fluid circuit comprising a compressor including a compressor motor, a gas cooler, and an evaporator; a heating circuit thermally coupled to said working fluid circuit; and a cooling circuit thermally coupled to said working fluid circuit; wherein said heating circuit further comprises a heating circuit pump; wherein said cooling circuit further comprises a cooling circuit pump; and a power supply architecture, the power supply architecture comprising: a mains power supply input; a compressor motor power output to said compressor motor; a heating circuit pump output to said heating pump; a cooling circuit pump output to said cooling pump; and a power supply bus connecting interfaces to said mains power supply input, to said compressor motor power output, to said heating circuit pump output, and to said cooling circuit pump output; wherein said power supply bus is a DC bus, the power supply architecture including a controllable AC to DC converter in an interface between said mains power supply input and said DC bus; the power supply architecture further comprising: one or more variable frequency drive (VFD) motor controllers in one or more interfaces coupled between said DC bus and one or more of said compressor motor, said heating pump, and said cooling pump, said one or more VFD motor controllers each having a DC input and a variable frequency AC output; and a power system controller coupled to said controllable AC to DC converter and to said one or more VFD motor controllers to control power drawn from mains power supply input in response to an external power limit control signal.

To aid in understanding the operation of embodiments of the invention we now describe some aspects of combined heating and cooling systems in which the above described techniques may be employed.

Combined heating and cooling systems

We have previously described a combined heating and cooling system, the system comprising: a working fluid circuit comprising a compressor, a gas cooler and an evaporator; a heating circuit, thermally coupled to said working fluid circuit via said gas cooler; and a cooling circuit, thermally coupled to said working fluid circuit via said evaporator; wherein said heating circuit further comprises a thermal storage tank, in particular a stratified thermal storage tank, controllably coupled to said heating circuit to controllably store heat for said heating circuit.

Embodiments of the above described system are adapted for control to facilitate efficient operation of the system. In particular by providing a controllable thermal store it is easier both to operate the system in an efficient regime when simultaneously heating and cooling and also to achieve stable control of the combined heating and cooling system.

In preferred embodiments the thermal storage tank is configured such that the tank is stratified, in particular into one or more of layers of fluid separated by one or more thermoclines (although a mechanically stratified tank may alternatively be employed). In one embodiment, as described later, a single tank is employed with a thermocline which moves up and down within the tank according to the amount of heat stored in the tank; optionally multiple such tanks may be provided "in parallel". Additionally or alternatively however the stratified thermal storage tank may comprise a plurality of tank vessels or chambers coupled "in series", for example stacked one above another or side-by side, and coupled together by one or more fluid flow conduits to allow fluid to move between them. Where the vessels or chambers are coupled side-by-side a conduit from the top (or an upper portion) of one vessel/chamber may couple to the bottom (or a lower portion) of the next vessel/chamber. Optionally two (or more) conduits may couple each vessel to each adjacent vessel, optionally one conduit allowing fluid flow in one direction (for example, up) and a second conduit allowing fluid flow in a second direction (for example, down). In a tank comprising a set of vessels the moving thermocline may be approximated by a change in the number of vessels containing hot (warmer) as opposed to cold (cooler) fluid (for example water). Suitable devices are known to those in the art and are also available for purchase.

Use of a stratified thermal storage tank is advantageous as this separates relatively warmer and cooler portions of the heating circuit fluid. This in turn facilitates achieving a low temperature for the input to the gas cooler, this low temperature in the heating circuit facilitating efficient operation of the working fluid circuit. Thus, paradoxically, employing stratified thermal storage not only facilitates obtaining a higher temperature for the portion of the heating circuit used for heating, it also facilities achieving a lower temperature in a different portion of the heating circuit which facilitates efficient operation of the overall system.

Further, and importantly, use of a stratified thermal storage tank allows the thermal storage tank to be used as a gauge in which the degree of stored thermal energy can be determined from a set of temperature sensors at different levels within the tank. This in turn facilitates control of the system based upon stored energy rather than on temperature per se. In the case of a tank comprising a set of vessels/chambers the stored thermal energy gauge may be provided by a count of the number of vessels/chambers containing hot (warmer) as opposed to cold (cooler) fluid. It is not, however, essential to employ stratified thermal storage as in principle the advantages of such an approach may be achieved in other ways, for example by means of multiple smaller thermal storage tanks.

In some preferred embodiments the system also includes a thermal dump system to enable heat to be dumped from said heating circuit, again to facilitate efficient overall system control, counter-intuitively dumping thermal energy facilitating an overall energy saving. In embodiments the heating circuit is configured, for example using valves, to direct or switch flow in the circuit between the thermal storage tank and the thermal dump, for example when the thermal store does not need to be replenished. In preferred embodiments a heating side control system controls storage and dumping of heat for the heating circuit, in particular based on a level of stored energy. In embodiments this is measured by a set of temperature sensors in the stratified thermal storage tank, that is control is effectively based upon thermal energy stored in the heating circuit. In embodiments the heating side control system controls the compressor with the aim of maintaining the stored thermal energy at a substantially steady state, which may be defined by a target range. For example the virtual temperature gauge provided by the stratified thermal storage tank may be controlled to a target percentage/percentage range (of full capacity). Preferred embodiments of the system also include a measure of stored coolth (the term of art "coolth" is explained in more detail later). The stored coolth may be measured by one or more temperature sensors in the cooling circuit. In preferred embodiments the heating side control system is responsive to both stored heat energy and to stored coolth, in particular to control the compressor and/or to dump heat from the heating circuit.

Preferred embodiments of the system also include a controllable coolth dump to allow coolth to be dumped from the cooling circuit. This may comprise, for example, a controllable heat exchanger to exchange heat with an ambient, typically external environment. Such a heat exchanger may also be used for "warmth dumping", that is for free cooling where the ambient temperature is less than a target temperature for the cooling circuit. A cooling side control system is preferably included to control the coolth dump responsive to a sensed temperature of coolant within the cooling circuit, in particular to control a heat input to the evaporator provided by the coolant. In some preferred embodiments the coolant temperature for the cooling side control system is measured in a coolant flow path to an input to the evaporator, more particularly in the mixer header described later.

In some preferred embodiments of the system the cooling circuit has a coolth output (the 'ring-main' in the example described later), and is controllably reconfigurable between parallel and series modes of operation. In the parallel mode the controllable coolth dump is coupled in parallel with the coolth output, and in the series mode these are coupled in series. Optionally a configuration may be provided in which a portion of the output of the coolth dump may be mixed with a portion of the output from the evaporator. In preferred embodiments control is provided, preferably by the cooling side control system, to switch the cooling circuit between the series and parallel modes of operation dependent upon an ambient temperature (of the controllable coolth dump). More particularly the series mode may be selected when the ambient temperature is below a target temperature for the coolth output. Optionally a solar thermal energy capture system may be coupled to the heating side of the circuit to facilitate use of 'free' solar heating when available.

Although preferred embodiments of the systems and control techniques we describe are applied in the context of a combined heating and cooling system, in particular using carbon-dioxide as the working fluid, in principle some benefit is obtainable by applying the techniques we describe separately to one or other side of a system, that is to a system which heats but does not cool and vice versa.

We have also described a method of controlling a combined heating and cooling system, in particular as described above, the method comprising: determining one or both of a stored heat in said heating circuit and a stored coolth in said cooling circuit; and controlling one or both of said compressor and said coupling of said stratified thermal storage tank to said heating circuit responsive to said determination of stored heat/coolth to maintain one or both of said stored heat and said stored coolth in a steady state, more particularly within a respective target range. Again the stratified thermal storage tank may comprise a single tank or a set of stacked vessels (tanks/vessels may be coupled "in series" and/or "in parallel"). In embodiments of the method the heating circuit is controlled to control the stored heat and/or stored coolth, for example to maintain this is a steady state, in embodiments to achieve a target or target range of stored heat and/or stored coolth. In embodiments this is achieved by controlling the compressor and preferably also by controlling dumping of heat from the heating circuit. Whilst it might be thought that it would be most efficient to store all the heat generated, counter-intuitively the overall system efficiency can at times be increased by dumping heat.

We have also described a method of controlling a combined heating and cooling system, in particular as described above, the method comprising: determining a temperature of coolant circulating in said cooling circuit; and controlling dumping of coolth from said cooling circuit responsive to said coolant temperature to control a heat input to said evaporator provided by said coolant.

Again counter-intuitively, the overall system efficiency can at times be increased by dumping coolth from the cooling side of the system, so as effectively to heat the input to the evaporator (from the cooling circuit) and increase the efficiency of the working fluid circuit. In embodiments the coolth may be dumped by controlling the rate at which a heat exchanger operates (in embodiments, to exchange heat with an ambient, generally external environment); and/or by selecting a series or parallel mode of operation for the cooling circuit. In embodiments the method further comprises selecting a series mode of operation when 'free' cooling is available, that is when an ambient environment of the coolth dumping device is less than a target desired temperature for the coolant in the coolth output of the cooling circuit.

In preferred implementations both of the above control methods are implemented in a combined heating and cooling system of the type we describe. Preferably the working fluid comprises carbon-dioxide. In preferred embodiments the working fluid circuit is configured to operate (by default) with a transcritical cycle.

We have also described a carbon dioxide-based combined heating and cooling system, the system comprising: a working fluid circuit comprising a compressor, a gas cooler and an evaporator; a heating circuit, thermally coupled to said working fluid circuit via said gas cooler and having heat output; and a cooling circuit, thermally coupled to said working fluid circuit via said evaporator and having a coolth output; and a first control system to control said working fluid circuit responsive to one or both of heat stored in said heating circuit and coolth stored in said cooling circuit to partly satisfy heat and coolth demands from respective said heat and coolth outputs with stored heat and/or coolth.

Thus in preferred embodiments the system is controlled based (partly) on stored heat and/or coolth, and preferably operates to control storage and/or dumping of heat and/or coolth. In embodiments the control is further facilitated by controlled dumping of either heat or coolth, for example to facilitate achieving an approximately steady state of stored heat and coolth (even with varying heat and coolth demands). In embodiments such a steady state may involve the stored heat and coolth being approximately in the middle of their respective ranges (for example 50% +/- 30%), but alternatively the system may be biased towards greater or lesser storage of heat and/or coolth.

In preferred embodiments a further control system, preferably operating with a shorter cycle time, operates to control the cooling side circuit to control the temperature of the coolant flowing into the evaporator, again preferably by controlling at least dumping of coolth from the cooling side circuit. In embodiments the second control system may operate to control the coolant temperature up towards a target temperature. In broad terms this is advantageous because this raises the pressure of the working fluid (carbon dioxide) in the evaporator, which in turn increases the mass flow of working fluid through the evaporator and compressor, without causing a proportionate rise in the shaft power of the compressor. This helps to raise the "Coefficient of Performance" (COP) of the refrigeration cycle.

Preferably the second control system also controls the coolant temperature down towards a target to achieve a target cooling effect (from a coolth output of the cooling side circuit). As previously described the second control system may also selectively couple the coolth dump (heat exchanger) either in series or in parallel with the coolth output responsive to a sensed ambient temperature. That is a parallel configuration may be selected if the ambient temperature is greater than a target temperature of the coolant. As previously described in preferred embodiments the system controls the heating side of the system in response to a determined stored heat gauge and/or a determined stored coolth gauge. As previously described, preferably the heating circuit includes a controllable heat dump to (indirectly) control the efficiency of operation of the working fluid circuit, in particular to facilitate dumping heat from the working fluid (carbon dioxide) circuit.

BRIEF DESCRIPTION OF THE DRAWINGS

These and other aspects of the invention will now be further described, by way of example only, with reference to eh accompanying figures in which: Figures 1 a and 1 b show a cooling circuit of an integrated heating/cooling system according to a preferred embodiment of the present invention, configured respectively into a parallel mode of operation and a series mode of operation;

Figure 2 shows a heating side of the integrated heating/cooling system of Figure 1 ;

Figure 3 shows a control system for the integrated heating/cooling system of Figures 1 and 2;

Figure 4 shows a pressure-enthalpy curve illustrating operation of part of the integrated heating/cooling system of Figures 1 and 2;

Figure 5 shows a stratified thermal storage tank instrumented with a temperature sensors for use with the integrated heating/cooling system of Figures 1 and 2; Figures 6a and 6b show, respectively, a compressor control procedure a thermal store/dump control procedure for the integrated heating/cooling system of Figures 1 and 2;

Figure 7 shows regions of a lookup table for controlling operation of the integrated heating/cooling system of Figures 1 and 2; Figure 8 illustrates a first example of operation of the integrated heating/cooling system of Figures 1 and 2; Figure 9 illustrates a second example of operation of the integrated heating/cooling system of Figures 1 and 2;

Figures 10a and 10b these show, respectively, an embodiment of a power control architecture for a combined heating and cooling system, and a perspective view of a physical configuration for such a system ;

Figures 1 1 a and 1 1 b show example procedures for controlling a combined heating and cooling system in response to a demand control signal, according to embodiments of the invention;

Figure 12 shows a power control procedure for controlling combined heating and cooling system in the absence of an external demand reduction signal, according to an embodiment of the invention; Figure 13 shows example energy flows for the procedures of Figures 1 1 and 12;

Figures 14a and 14b show, respectively, an embodiment of a heat pump system according to the invention, and an energy management system comprising a fleet of heat pump systems of the type shown in Figure 14a; and

Figures 15a and 15b show, respectively, an example of a refrigeration-only heat pump system installed in a building and an example of a refrigeration-only heat pump system installed in an electric vehicle.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

To aid in understanding the invention we will first describe an example of a combined heating and cooling system (heat pump system) in the context of which embodiments of the invention may operate. Combined heating and cooling systems We thus first describe details of an integrated heating and cooling system (also referred to as a heatpump-chiller) which employs a water to water heat pump and which uses C0 ₂ as the primary refrigerant. We will also describe details of an associated thermal control system that automatically regulates the heating and cooling of the heatpump- chiller.

We will consider an application where both a chilled water circuit and a heated water circuit are provided, for example a chilled water ring main and a low temperature hot water (LTHW) ring main. In the description below we generally refer to chilled water but the skilled person will understand that this may include an antifreeze such as glycol or salt (in which latter case the chilled water may be brine). However the skilled person will also recognise that the techniques we describe are not limited to use with any particular working fluid such as water (in the heating or cooling circuits), or even carbon dioxide. Similarly the skilled person will recognise that although in the description below we refer to a "ring main", this merely refers to any form of circuit (which may include series and/or parallel sub-circuits).

We will describe techniques for employing a common, potentially single heat pump- chiller as the thermal engine. However to implement such a system in practice involves considerations relating to both the apparatus and the control system. For example, to cope with year round varying load profiles it is desirable to implement a control strategy that facilitates the heat pump-chiller adapting to out of balance situations. In broad terms this occurs where the quantum of warmth or coolth required by the LTHW & brine ring-mains respectively does not correlate with the achievable ratio of warmth and coolth that can be cogenerated by the heatpump-chiller. (In the context of a refrigeration system the heat input is sometimes termed the coolth output, measured in Joules or Jules/second).

We describe further below what is meant by "free cooling", but in outline some preferred embodiments of the system employ a heat exchanger, more particularly an air-blast heat exchanger in the chilled water circuit, located in the external (outside) environment. In this way, when the external ambient temperature is below a target for the chilled water circuit, the ambient air can be used for "free" cooling of the chilled water. In embodiments the heating/cooling system uses a vapour-compression refrigerating principle and is fitted with a gas cooler that is controlled to operate in a transcritical range of C0 ₂ pressures. In some preferred embodiments the integrated heating/cooling system includes a system for heat (hot water) storage, in embodiments a thermally layered storage tank. This storage facilitates control of the system, in particular facilitating the joint operation of the heating and cooling portions of the system in a manner which provides some tolerance to varying loads on either side of the system (heating and cooling), such that the operation of the system can be overall more efficient. In some preferred embodiments the integrated heating/cooling system additionally or alternatively includes a mechanism to effectively provide cold (chilled water) storage, for similar reasons. In practice this can be considered to be provided by the circulating chilled water in combination with a measure of a cooling requirement to be provided by the circulating chilled water. The measure may comprise a temperature difference between a target cold temperature and the temperature of the circulating chilled water at a point prior to its use for cooling, that is, for delivering coolth to a process. Thus in embodiments a measure of such a temperature difference may be used as a proxy for the stored coolth in the chilled water (brine) circuit. We will sometimes refer to this later as a "virtual" coolth store.

Referring now to Figures 1 a, 1 b, 2 and 3, these show a preferred embodiment of the integrated heating/cooling system 100. A control system 300 is shown schematically; in embodiments it comprises a general purpose computer system or dedicated controller 310 including stored program code to control the operation of the system. The control system 300 also includes one or more interfaces 302 for controlling the heating and cooling circuits, more particularly valves, optionally pumps, and for controlling a C02 refrigerant compressor of the combined heating/cooling system. It preferably also includes one or more sensor interfaces 304 for sensing temperatures and/or pressures within the system, and optionally for monitoring correct operation of components of the system. The control system may still further comprise a user interface 306 for interacting with, monitoring, configuring, and/or controlling the heating/cooling system; and optionally a wired or wireless network interface 308 for external communications and remote monitoring/control.

Figures 1 a and 1 b show a cooling (chilled brine) circuit 100a of the system, with valves controlled to put the circuit into parallel and series modes of operation respectively. Here, broadly speaking, "parallel" refers to an air blast heat exchanger and chilled ring- main configured so that fluid flows through both systems take place in parallel (ie fluid enters both systems from a common source at the same time), whilst "series" refers to an air blast heat exchanger and chilled ring-main configured so that fluid flows in series, firstly through the air-blast heat exchanger and then after that into a chilled water circuit.

Figure 2 shows a heating (LTHW) portion of the circuit 100b. Figures 1 and 2 together show a complete system - evaporator 102 is common to Figures 1 and 2. The evaporator has a relatively lower pressure side 102a connected to the cooling circuit, more particularly through which the chilled water (brine) circulates, and a relatively higher pressure side 102b connected to the C02 circuit, more particularly through which the C02 liquid circulates. Referring to Figures 1 a and 1 b, in the symbols for the valves open triangles indicate the flow is allowed through that part of the value and solid triangles indicate that flow is blocked. Thus the dashed (blue) lines indicate liquid (water/brine) paths in which there is substantially no flow whilst solid (red/green) lines indicate paths of liquid (water/brine) flow. Thus in the parallel mode of Figure 1 a air-blast heat exchanger 104 is connected in parallel with a chilled ring main 107 thus allowing chilled water flows from the evaporator to be distributed through the chilled water circuit and through the air-blast heat exchanger 104 at the same time whereas in the series mode of Figure 1 a the air-blast heat exchanger 104 is connected in series with chilled ring main 107 so that the air-blast heat exchanger can be used as a means of supplying extra coolth into the chilled water (brine) circuit

The chilled ring main typically comprises one or more cooling devices such as (air) coolers 107a-c which may be used, for example, for comfort cooling (air conditioning), for food conservation (to cool a food chiller cabinet), or the like. Air-blast heat exchanger 104 is preferably located outside a building or in some other environment external to or outside the regions heated/chilled by the system 100. A typical air-blast heat exchanger used in embodiments of the invention may comprise a coiled or meandering piped liquid path in combination with one or more fans to force air over the piped liquid.

The external air-blast heat exchanger may be used to dump (jettison) excess coolth from the chilled water (brine) circuit in what might be termed a "sky cooling mode" of operation. This mode is operable when the chilled water (brine) temperature inside the air-blast heat exchanger is more than a threshold colder, for example 2 or more degrees Celsius colder, than the ambient air. Similarly when the chilled water (brine) temperature inside the air-blast heat exchanger is more than a threshold warmer, for example 2 or more degrees Celsius warmer, than the ambient air then the air-blast heat exchanger may be used to dump (jettison) excess warmth from the chilled water (brine) circuit

Depending upon the geographical location, time of year, and operating environment the range of values of the ambient/outside temperature, perhaps -20°C to +30°C, may overlap with the target temperature range for the chilled water circuit, perhaps -5°C to +15°C (the higher end encompassing air conditioners). These values are merely illustrative but are intended to convey the inventor's recognition that, at times and with a suitable control strategy, ambient air may provide an important part of the overall system control. By way of example, a chilled water circuit operating with a 12°C flow and 7°C return temperature can be maintained solely by an air-blast heat exchanger system without the need for mechanical refrigeration, so long as a sufficiently cold flow of ambient air is passed across the air-blast heat exchanger so that the temperature of the water at the exit of the air-blast heat exchanger is reduced to within the maximum allowable flow temperature of 12°C (in this case).

In the illustrated embodiment the chilled liquid circuit comprises two main pumps, a first pump 106 and a second pump 108. The first pump 106 may be termed a chilled circuit pump because it circulates chilled liquid through the air coolers 107a-c of ring main 107, either directly (parallel mode) or indirectly (series mode). The second pump 108 may be termed an evaporator pump because it draws chilled liquid through the evaporator. In the parallel mode first pump 106 circulates the cooled liquid through cooling devices (air coolers) 107a-c and the second pump 108 circulates the cooled liquid through air- blast heat exchanger 104 to chill the ambient air (dump excess coolth). In the series mode the first pump 106 circulates chilled water (brine) sequentially through the air- blast heat exchanger 104 and next through air coolers 107a-c; this may be a substantially constant circulation of liquid. The second pump 108 may operate intermittently or not at all to circulate liquid in the cooling circuit, in particular to mix liquid in a mixing tank, mixing header 1 10. Mixing header (or "discharge header") 1 10 is, essentially, a junction box so that the circulating flows meet and mix prior to flowing back through evaporator 102 and/or air coolers 107. In a similar manner preferably a suction header 1 12 is included in the path of liquid exiting the evaporator 102, to provide a "junction box" with parallel flow outputs on this side of the evaporator (in operation the "suction header" is substantially entirely full of liquid).

In addition, as previously outlined, the cooling circuit 100a of Figure 1 includes a number of controllable valves, which may be controlled by the control system to configure the circuit for either a series or a parallel mode of system operation. The valves thus operate as mode selection valves; in embodiments the valves are controlled electrically by control system 200 to select fluid flow paths through the valves.

Thus in one embodiment the lower pressure side 102a of evaporator 102 has an output fluid flow path or conduit 130 coupled to (suction) header 1 12, which in turn has first and second output conduits 132, 134. (Here "conduit" is used broadly to mean any fluid flow path). Conduit 132 provides a first input to a first controllable (mode selection) valve 1 14, which has a second input from conduit 150 and an output conduit 136. Valve 1 14 is controllable to direct either the first or second input to output 136.

The second output conduit 134 of evaporator 102 is coupled to an input (suction side) of evaporator pump 108. The output conduit 136 provides a first input to an optional (but preferable) mixing valve 1 16, which has a second input conduit 138, and an output conduit 140. Conduit 140 provides an input to chiller circuit pump 106. The second input 138 to mixing valve 1 16 is from liquid which has passed through the chilled circuit 107, and which has thus been warmed by this passage, thus helping to warm the input to the chilled "ring main" in a condition in which there is excess coolth in the suction header. Valve 1 16 is preferably controllable to selectively mix input 138 with the flow between input 136 and output 140. We describe control of valve 1 16 later, but in embodiments if the chilled water (brine) temperature inside the suction header is greater than a threshold difference lower than a desired target temperature some of the output of the ring main may be mixed in.

An output conduit 142 of chiller circuit pump 106 provides an input to a second controllable (mode selection) valve 120, which has a first output conduit 144 to air-blast heat exchanger 104 and a second output conduit 146 to chilled "ring main" 107. Valve 120 is controllable to direct the input 142 to either the first or second output 144, 146.

The output conduit 146 of valve 120 is coupled to ring main 107, to provide a chilled water input to the air coolers 107a-c of ring main 107. An output conduit 148 from ring main 107 provides a first input to (mixing) header 1 10.

Pump 108 has an output conduit 152 which provides an input to a third controllable (mode selection) valve 1 18. Valve 1 18 has a first output conduit 154 joining conduit 144, to provide an input to (external) air-blast heat exchanger 104, and a second output conduit 156 which provides a second input to (mixing) header 1 10. Valve 1 18 is controllable to direct the input 152 to either the first or second output 154, 156.

An output conduit 158 from air-blast heat exchanger 104 provides an input to a fourth controllable (mode selection) valve 122, which has a first output conduit 160, and a second output conduit 162 which provides a third input to (mixing) header 1 10. The first output conduit 160 is coupled to conduit 146, so that circulating liquid from air blast heat exchanger 104 can provide an input to ring main 107. Valve 122 is controllable to direct the input 158 to either the first or second output 160, 162 (in series and parallel mode respectively).

The (mixing) header 1 10 has a first output conduit 150, which provides an input to valve 1 14 to enable circulation of liquid through the external air blast heat exchanger 104 and ring main 107 in series mode. The (mixing) header 1 10 has a second output conduit 164 which provides an input to the higher temperature side 102a of evaporator 102 and, in embodiments, provides a continuous path through evaporator 102 to output conduit 130. As described further below, evaporator 102 acts as a heat exchanger and the path 164, 130 through evaporator 102 provides one part of the flow through this heat exchanger. The skilled person will appreciate that within the heat exchanger the fluid conduit may be defined in many ways, for example by tubes, plates, baffles and the like.

The skilled person will appreciate that the configuration of the valves and conduits may be varied whilst still providing a cooling circuit which can be switched between parallel and series modes of operation. As previously described in such a system in the series mode liquid circulates sequentially through the external air-blast heat exchanger and chilled "ring main" (optionally but preferably water is also circulated through evaporator in a second circuit). In the parallel mode liquid from the evaporator is circulated in parallel through the external air-blast heat exchanger and chilled "ring main", optionally mixing the output from the ring main with the input to the ring main, so that the ring- main feed-in temperature is not excessively cold.

It will be appreciated that the temperatures of the circulating liquid depend upon the application. Nonetheless it is helpful for understanding embodiments of the invention to provide illustrative example temperatures.

Thus for a system in which the chilled ring main cooling devices 107a-c comprise air coolers for chilling food the input 146 to these may be at around -5°C whilst the output 148 from these may be at around -1 °C (for an air conditioning circuit the input temperature may be higher, for example around +10°C). In this example an air cooler has air at, say, +1 °C blown over a conduit, radiator or the like carrying the chilled liquid (brine), so that the air is cooled to, say, around -2°C whilst the brine is heated from, say, -5°C to -1 °C. In this example in parallel mode the output 130 from the evaporator 102 may have become significantly colder than the targeted chilled ring-main feed-in temperature (it may have reached -8°C for example, at the same time brine being fed into the air-blast heat exchanger 104 would also be at -8°C). Parallel mode is selected when the external air temperature is high enough to allow dumping of coolth, for example -8°C coolth can easily be dumped into ambient air of +9°C, and thus the output 158 from the air-blast heat exchanger could be substantially warmer than the water exiting air coolers 107 a-c. Thus in a food chilling application the mixing header may receive one flow at around -1 °C and another flow at around +5°C (assuming 9°C ambient). By enabling unwanted coolth to be dumped through the air-blast heat exchanger in this way it is possible to impose a heat load on the evaporator 102 during times that require the C02 refrigerating system to be operated primarily for the production of hot water, even though an excessive cooling effect (relative to that required by the chilled ring- main 107) may be emanating from the low pressure side of the C02 refrigerating system.

In series mode the outlet 148 from coolers 107 a-c on the chilled water (brine) ring- main may again be at around -1 °C. However the ambient temperature is sufficiently low in this mode of operation for the air-blast heat exchanger 104 to be able to at least partially cool the output from the chilled ring-main 107. In preferred embodiments of series mode the cooling for the chilled ring-main can at times be provided solely by the air-blast heat exchanger 104.

Referring now to Figure 2, this shows a heating (LTHW) portion of the circuit 100b, in which evaporator 102 is common to the circuits shown in figures 1 a, 1 b and 2. In embodiments the heating circuit side of system 100 includes a vapour (gas) compression refrigeration system. Thus the evaporator 102 of Figure 2 is the same evaporator 102 as illustrated in Figures 1 a and 1 b. We will describe a preferred implementation of the system which employs carbon dioxide as the vapour but in principle other gases may also be employed.

In more detail, conduit 202 carries high pressure dense gas, in particular carbon dioxide, towards an expansion valve 204, which allows the pressure to reduce, with a concomitant reduction in C02 temperature. In embodiments expansion valve 204 may be an automatically adjustable small bore needle valve. Expansion valve 204 converts the gas, which at this point is a dense mist, into a cold, mostly liquid state in conduit 206 leading away from expansion valve 204. By way of example, the temperature of the liquid in conduit 202 may be around 30°C whilst the temperature of the fluid in conduit 206 may be around -10°C. For efficient operation it is preferred that there is some turbulence in the flow within conduit 206. This is facilitated, in particular, by a mixture of "wet" and "dry" fluid, which in turn can be achieved by arranging for the vapour (carbon dioxide) in conduit 206 to be in a region of the pressure enthalpy curve in which the vapour is not fully wetted but is also not subject to a vapour fraction in excess of 0.5

Conduit 206 provides a continuous path through evaporator 102 to output conduit 208. During its path through evaporator 102 the carbon dioxide is boiled, for example at -10°C to the point where 5K or more of superheat is generated by the counter-flowing liquid (brine/glycol) in the conduit 164 to 130. As energy is passed from the brine to the C02 then the brine is reduced in temperature eg. from -1 °C at inlet to 102 to say -5°C at the outlet from evaporator 102.

During its passage through the evaporator the carbon dioxide is also "dried", although there may still be some residual dampness. For this reason conduit 208 is preferably provided with a droplet separator 210 to remove residual liquid droplets from the flow prior to compression. Thus in embodiments an output 212 of (optional) droplet separator 210 is provided to a compressor 214, and this in turn has an output 216 to an optional oil separator 218 (to remove residual compressor oil), and thence to conduit 220. Compressor 214 raises the pressure of the carbon dioxide, and also the temperature of the carbon dioxide, for example to around 85°C.

Conduit 220 provides an input to a heat exchanger 222, which may be referred to as a "gas cooler": During its passage through this element the temperature of the carbon dioxide is reduced prior to its passage through expansion valve 204, using the preceding figures to around 30°C (the temperature in conduit 202). A mechanical filter 226 is preferably provided between an output conduit 224 of heat exchanger 222 and expansion (needle) valve 204.

Heat exchanger 222 is coupled to a water heating circuit. For convenience we will sometimes refer to this circuit as an LTHW (low temperature hot water) circuit, but the skilled person will appreciate that other types of heating circuit may also be implemented. Thus in embodiments heat exchanger 222 has an input conduit 228 and an output conduit 230, together forming part of the LTHW (heating) circuit. By way of example, input conduit may carry water at around 28°C, which cools the counter- flowing carbon dioxide gas flowing through conduits 220, 224, in turn heating the water so that output conduit 230 may, for example, be at around 55°C. The skilled person will appreciate that the degree of heating depends both on the temperature of the counter-flowing gas and also on the mass flows of the gas and water. The output conduit 230 from heat exchanger 222 provides a source of hot water for thermal storage device 232, in preferred embodiments a layered (stratified or thermocline) thermal storage device. Conduit 230 provides an input to an upper, high temperature region of the tank where the water temperature may be, for example, around 55°C. Stratification within tank 232 maintains a temperature differential between upper and lower regions of the tank and thus a lower region of the tank may at the same time be at a much lower temperature, for example around 30°C, without substantial mixing between the stratified layers. An output conduit 234 from a lower region of tank 232 provides a lower temperature outlet and a lower temperature return to conduit 228, an input to heat exchanger/gas cooler 222. In embodiments this return is via a liquid pump 236 (although it will be appreciated that the pump may be located elsewhere); this pump may be termed a gas cooler pump.

In preferred embodiments the LTHW circuit also includes a heat exchanger 240 which is usable to dump excess heat, and which may therefore be termed a hot side dump exchanger. This heat exchanger may be a liquid-to-liquid heat exchanger or a liquid- to-gas heat exchanger; in the latter case it may be situated in any convenient location to dump heat, for example external to a heated environment/building. An input to heat exchanger 240 is provided by a conduit 248, a branch of conduit 230. A return conduit 242 from thermal-dump heat exchanger 240 is coupled a first input of a (dump) changeover valve 244. Valve 244 has a second input from conduit 234, and an output conduit 246 which provides an input to pump 236. Thus valve 244 is controllable to either permit or inhibit a heated water flow through thermal-dump heat exchanger 240. Valve 244 is likewise controllable to either permit or inhibit a heated water flow into (through) the thermal store 232, and thus effectively to switch flow between the thermal store and thermal dump. One or more additional shut-off valves may be operated in coordination with valve 244 so that gas passing through 222 is cooled either via the thermal store 232 or via the hot-side dump exchanger 240.

Optionally a solar water heater 260 may also be coupled to thermal store 232, as shown in simplified form, to enable solar thermal heating input to the system. In this configuration the solar thermal panel(s) would serve (instead of the gas cooler) as a means of generating heat into the thermal store.

A heating circuit 250, in embodiments an LTHW "ring main", is coupled to thermal storage tank 232. More particularly in embodiments a first conduit 252 from an upper, heated region of tank 232 provides an input to heating circuit, which has a second, output conduit 254 providing an input to a lower, cooler region of the tank. A pump 256 pumps water through heating circuit 250, which includes one or more heating devices 256a-c, for example radiators.

Referring again to the carbon dioxide vapour (gas) compression refrigeration circuit, in preferred embodiments this operates in a transcritical mode, that is as it circulates through the refrigeration circuit the carbon dioxide defines a transcritical cycle (enclosing the critical point) on a pressure (p) enthalpy (H) graph for the carbon dioxide. We use the terms vapour and gas interchangeably herein.

Figure 4 shows such a pressure-enthalpy curve with labelled points A to D corresponding to labelled locations A to D in Figure 2. In Figure 4 the critical point is labelled X and lies at the top of the dome defined by the saturated liquid line (to the left) and saturated vapour line (to the right). It can be seen that the closed curve defining the carbon dioxide cycle extends above the critical point encompassing part of the supercritical region.

We now describe an example cycle: Point D labels the output from expansion valve 204, the input to evaporator 102. As the (mostly liquid) carbon dioxide passes through the evaporator at substantially constant pressure it is boiled at -10°C (at the saturated vapour line), and then heated-up further, to -5°C (in the superheated region), to arrive at point A where it has 5°C of superheat. The skilled person will appreciate that the extent of the superheating at point A is variable.

The vapour is then compressed by compressor 214, moving from point A to point B at the output of the compressor and heating the gas to, for example, 85°C. At point B the gas is beyond the critical point X, in a supercritical region. During its passage through gas cooler 222 the vapour is cooled at approximately constant pressure, for example to 30°C, to arrive at point C, the input to the expansion valve 204. Expansion valve 204 reduces the pressure, down to point D, reducing the temperature, for example to around -10°C, liquefying the vapour. It should be noted that whilst transcritical operation is the default mode (when the gas cooling pump connects via the thermal storage tank) it may be the case that the C02 refrigerating cycle operates entirely below the critical point when the hot-side dump exchanger is engaged.

Control schemes

We next describe some preferred examples of control schemes for efficient operation of the above-described system. In broad terms it is desirable to provide thermal storage on the heating and cooling sides of the system. This may be inherent, provided by the heated/cooled liquid circulating within the respective circuits and/or additional thermal storage may be provided, for example using a layered thermal storage tank as previously described. Thus we refer below to stored heat and stored coolth. Efficiency may still further be increased by taking advantage of effectively free cooling which may be provided by a cold external ambient environment (by comparison with a target temperature) at certain times of year.

Consider, for example, a system which is providing cooling for air conditioning units. If, say, the system makes "10 units" of cooled liquid but the air conditioners need only 6, there is excess coolth. Thus it is beneficial in such a situation to dump some excess coolth. Where the external ambient temperature is high enough this can be achieved by using air-blast heat exchanger 104 to "refrigerate the sky", mixing the warmed brine return from air-blast heat exchanger 104 in mixing header 1 10.

The above example illustrates an example of an unbalanced condition which the control system should aim to address. We now describe preferred embodiments of suitable control schemes. In embodiments two control schemes operate, one for the C02 refrigerating system and one for the air-blast heat exchanger 104. Although these systems are essentially distinct it is the case that changes made by one control system may lead to changes needing to be made by the other control system.

Heat storage - temperature sensing In embodiments the stratified thermal storage tank 232 is instrumented with a set of temperature sensors at different levels within the tank, so that the stored energy in the tank can be determined. Referring to Figure 5 there may, for example, be four temperature sensors sensing temperatures T1 to T4 at successively lower levels within the tank. The stored energy level (SEL) may then be defined as 25%, 50%, 75% or 100% according to whether T1 ; T1 and T2; T1 , T2 and T3; or T1 , T2, T3 and T4 are at greater than a threshold temperature (preferably the same but potentially different for each of T1 to T4). Optionally a 0% level may be defined if T1 is not greater than the threshold. An example table for determining the stored (heat) energy level is shown below (Table 1 ). Conceptually this provides a stored heat or energy level "fuel gauge" for the system, as shown in the inset to Figure 5. It will be appreciated that this approach may be adapted for other numbers of temperature sensors.

Table 1

In preferred embodiments the control system 300 measures temperatures T1 to T4 at intervals, for example every five minutes, and determines the stored (heat) energy level, for example in terms of a percentage, as above.

Coolth storage - temperature sensing

In a similar manner, the cooling (chilled brine) circuit 100a is instrumented with at least one temperature sensor to measure a temperature of liquid circulating within the circuit, in embodiments a temperature, T _S H, of the input to chilled ring main 107 (Figures 1 a and 1 b), measured by a sensor located in the suction header 1 12. A difference between this temperature and a target temperature T, _ar g _e , (see below) is used as a measure of the stored coolth. This can be seen by considering a target temperature of -5°C; where T _SH is, say, -8°C the chilled brine ring-main could in this circumstance run for some time without further cooling of the brine because of the coolth stored in it, whereas if T _SH is at a temperature of, say, -4°C then the chilled brine ring-main could not operate for long (if at all) without additional coolth being provided by the evaporator 102. Preferably, but not essentially, when the target temperature is approximately the same as the measured T _SH the stored coolth is deemed to be at a 50% level. In embodiments the target temperature is a target temperature for the circulating liquid (brine) but in principle other temperatures may be used, for example a target temperature of fresh foods being chilled by an air cooler

An example table for determining a stored coolth level is shown below (Table 2), where ΔΤ = TsH-Ttarget- Conceptually this provides a stored coolth gauge for the system. It will be appreciated that the threshold temperature values may be varied, and that this approach may be adapted for finer or coarser gradations of stored coolth. In embodiments a stored coolth level of 0% is not used.

Table 2

In preferred embodiments the control system 300 measures T _SH at intervals, for example at around the same time as T1 -T4, say every five minutes, and determines the stored coolth level, for example in terms of a percentage, as above. This gives information about the level of cooling that is (or is not) being delivered whilst heat is also being (or is not being) delivered by the compressor to the thermal storage tank.

Heating-side/compressor control In preferred embodiments of the system data from the stored energy level (SEL) and stored coolth gauge (SCG) are processed jointly to control the compressor 214, gas cooling pump 236, and changeover valve 244 (to control heat dumping). In preferred embodiments the control system operates with the aim of keeping both the stored heat and stored coolth ("gauges") at around 50% full. In this way there is heating potential available in the heating circuit and cooling potential available in the cooling circuit. This approach also tends to optimise (power) efficiency. Alternatively, however, the system may be run with a bias towards either the heating or the cooling circuit.

As described above, embodiments of the system may operate with a control cycle which operates at intervals of, say, 300 seconds, to monitor the SEL and SCG and in response control the heating circuit side of system, more particularly the compressor, gas cooling pump and dump changeover valve. However in embodiments the control cycle is adaptive, and may change depending upon the SEL and/or SCG.

Preferred embodiments of the control system define a set of (discrete) compressor speeds and then control the compressor by incrementing or decrementing the compressor speed between one level to another (although a similar concept may also be applied to a continuously variable compressor speed). In the description below, increasing the speed is referred to as "loading" the compressor and decreasing the speed as "unloading" the compressor; "stay" denotes leaving the compressor speed unchanged.

In one, preferred approach the control system employs a lookup table to determine whether to load, stay or unload the compressor, as shown in the table below (Table 3). Table 3 also defines whether the dump changeover valve 244 is controlled so that pump 236 pumps to the layered thermal store 232 ("thermal store") or to the hot-side dump heat exchanger 240 ("hot dump").

SCG SEL Compressor Pump 236/valve 244 directs

(%) (%) heat to

25 25 Load Thermal store 25 50 Load Thermal store

25 75 Load Hot dump

25 100 Load Hot dump

50 25 Load Thermal store

50 50 Stay Thermal store

50 75 Unload Thermal store

50 100 Unload Hot dump

75 25 Load Thermal store

75 50 Unload Thermal store

75 75 Unload Thermal store

75 100 Unload Hot dump

100 25 Load Thermal store

100 50 Unload Thermal store

100 75 Unload Thermal store

100 100 Unload Hot dump

Table 3

The underlying logic of Table 3 may also be implemented by the procedures shown in Figures 6a and 6b (although use of a lookup table is potentially more flexible).

Thus referring to Figure 6a, which shows a compressor control procedure, at step S604 the system reads the SEL and SCG gauges and, at S606, determines whether either is at 25%. If so the compressor is loaded (sped up), S608, and the procedure loops; otherwise if both are at 50% the compressor speed is unchanged (S612). At step S614 one or both of SEL and SCG is at 75% or 100% and neither is at 25%; in this situation the compressor is unloaded (the compressor speed is decreased).

Figure 6b shows a control procedure for the thermal store/dump: At step S624 the procedure reads the SEL gauge and if this is less than or equal to 50% (S626), heat is stored (S628), and if the gauge is at 100% (S630) heat is dumped (S632). Otherwise SEL is at 75% and the procedure stores heat if the compressor is being unloaded by the compressor control procedure and dumps heat if the compressor is being loaded by the compressor control procedure. Optionally the above described control logic/procedures may optionally be state- dependent, that is the control applied may depend upon a previously applied control output. In particular if the compressor was unloaded (decreased) in a previous control cycle, and a subsequent unload (decrease) is indicated then it may be unloaded (decreased) faster. A similar approach may be applied when loading (increasing) the compressor. Additionally or alternatively the cycle period may be shortened in such situations.

It will be appreciated that the two control loops of Figures 6a and 6b interact as the SEL gauge of Figure 6a is affected by the thermal store/dump control of Figure 6b, and the control loop of Figure 6b is affected by the compressor speed adjusted by the procedure of Figure 6a. In broad terms, embodiments of the combined procedures aim to keep one or both the SEL and SCG gauges at a steady state. This may, for example, be at around 50% or some other value - for example if providing a constant summer air conditioning load, say during times of low demand for low temperature hot water, then the SEL might read say 75% all day long whilst the SCG oscillates between say 25% and 50%.

Nonetheless, although joint control of the compressor has been described based upon both the SEL and SCG "gauges", in principle (but less preferably) compressor control may be based upon just one of these by simply assigning a false value to the gauge that is not needed. For example the SEL parameter could be assigned a value of 100% to inhibit the compressor from altering its speed for any reason other than for the maintenance of temperatures in the suction header (chilled brine circuit).

Cooling-circuit control

In broad terms the air-blast heat exchanger control operates to facilitate a mixing header temperature that is consistent with energy efficient operation. It also preferably (but not essentially) switches the brine circuitry between series and parallel modes of operation, in such a way that the mixing header (1 10) temperature can be raised (parallel) or lowered (series)

In embodiments the air-blast heat exchanger control measures a temperature of the circulating coolant (for example water/brine/glycol) in the mixing header (1 10), and an ambient (air) temperature at entry into air-blast heat exchanger 104, TMH and Tan* respectively. The circulating coolant temperature is preferably measured at a location in the return path to the evaporator 102, conveniently at the mixing header 1 10. The air-blast heat exchanger control system operates to control a rate of exchange of heat with the ambient air by controlling the air-blast heat exchanger fan speeds, series- parallel switchover valves, and on/off operation of pumps 106, 108. In one, preferred approach the control system employs a lookup table to define a control strategy for controlling these elements.

Table 4, below, shows an example lookup table, where T_glycol refers to TMH (noting that the coolant need not be glycol). As described further below, Table 4 encodes information for mode, pump and air-blast heat exchanger control but in practice multiple separate tables may be employed. Table 4 is also referred to later as the Glycol Ambient Matrix (GAM).

Table 4a

Table 4b

The different regions of Table 4 above may be appreciated by cross-referencing to Figure 7. Table 4a shows regions A, C and E of Figure 7; Table 4b, which should be considered as lying horizontally to the right of Table 4a, shows regions B and D of Figure 7.

Figure 7 shows a version of the table in which regions A, B and C are regions in which the system operates in a parallel mode (Figure 1 a) and in which region D denotes a region of series operation (Figure 1 b). Region E denotes a region in which the system may be configured for either series or parallel operation - for example this region may be selected to be one or the other mode (by default, series), or hysteresis may be applied so that the region denotes parallel operation if approached from a parallel region (typically region C) or series operation if approached from series region D. The numbers in Table 4 denote a percentage rate of fan speed of the air-blast heat exchanger fans. The quantum of heat exchange in air-blast heat exchanger 104 is a function of fan speed and temperature difference (between the ambient air and the brine flowing through the air-blast heat exchanger).

The evaporator pump 108 is preferably ON in all operational regions except for region B (orange "0"s in Table 4), where evaporator pump 108 is preferably OFF.

Table 4 above is an example table for a target coolant temperature, TMH, of -1 °C (suitable for merchandising refrigerators). The target coolant temperature is preferably, but not essentially, measured in the mixing header. In general different target temperatures will employ different lookup tables. For example for an air conditioning circuit TMH may be around +14°C, whilst for (industrial) process cooling TMH may be higher, for example around +20°C. We describe below the principles underlying the table, which enable an adapted table to be constructed for different target coolant temperatures. Nonetheless in embodiments the use of a lookup table is preferred over a procedural approach, again, because it provides flexibility.

Series/parallel mode In Series Mode air-blast heat exchanger 104 is controlled so as to reduce the value of TMH until it is equal to (or close to) its target temperature. Air-blast heat exchanger 104 is able to work in this way so long as the ambient air being passed across the air blast heat exchanger is colder (by a margin, for example of 2 or more degrees Celcius) than the targeted TMH. The heat exchange capacity of air-blast heat exchanger 104 is proportional to the temperature difference between the brine inside the heat exchanger and the temperature and mass flow of the air flowing across the outside of the heat exchanger.

In parallel mode air-blast heat exchanger 104 is controlled so as to increase the value of TMH until it is equal to (or close to) its target temperature. If the ambient air is colder than the targeted TMH then the air-blast heat exchanger fans are operated only to the point where they can no longer increase actual TMH (regions A and C of Figure 7). For example, if the actual TMH is -8°C and if the air ambient air temperature is -3°C then it might be possible to warm the brine inside the mixing header to a temperature of say - 6°C (allowing for a temperature difference between the ambient air and the brine inside the mixing header). There would be no benefit to be gained by attempting to warm the brine any further and so fans would at this point be switched off even if the actual TMH was colder than the target TMH. Coordinated control

As described, there are two control loops; one that operates according to a set of temperature measurements taken in the thermal storage tank and in the suction header (1 12) and which affects the behaviour of the compressor and the heat storage / dumping system and another that acts according to prevailing temperatures in the mixing header (1 10) and which influences the behaviour of the fans fitted to air-blast heat exchanger 104, pumps 106 and 108 and a set of mode selection valves for parallel and series configuration These control loops interact indirectly in that the temperature inside the mixing header (which may be influenced by the control of the air blast heat exchanger) and the temperature inside the suction header (which may be influenced by the C02 compressor) will affect one another. For example it can be the case in series mode that the mixing header is being cooled by the air-blast heat exchanger whilst the suction header is simultaneously being cooled due to the compressor. In this situation it can be considered that the air-blast heat exchanger inhibits the build-up of excessively warm water at the exit of the cooling ring-main 107 whilst the compressor system inhibits the build-up of excessively warm water at the entry into cooling ring-main 107.

Achieving balanced operation in which the two control loops reinforce (rather than work against) one another is facilitated by setting appropriate time intervals for the operation of the two control loops. For example the air-blast heat exchanger ("GAM control") may operate every 90 seconds, whilst the compressor (SEL/SCG control) loop may operate once every 300 seconds. Preferably the air-blast heat exchanger control operates 5 or more times more frequently than the SEL/SCG control; preferably the intervals at which each control scheme operates are adjustable. Balanced operation is further facilitated by assigning appropriate target temperatures to TMH and TSH.

For example, in an application where the temperature set point for air inside an air- conditioned room is 22°C and the volume of air and brine flowing through the coolth emitters in the room is fixed then a practical method of controlling the thermal output of the coolth emitters is to maintain a stable brine temperature at exit from the emitters whilst at the same time allowing the brine temperature at inlet to the emitters to float.

For example the brine exit temperature (TMH) from a chilled brine ring-main could be held at say 14°C whilst the brine inlet temperature (TSH) might in practice be allowed to fluctuate between 9°C and 12°C. Thus in the case of the heatpump-chiller described here it is feasible to task the air blast heat exchanger 104 to maintain 14°C (TMH) at the same time as assigning the SCG a set point of say 10°C (TSH). In this example if the ambient air temperature was 4°C and in the absence of exceptionally high loads being imposed on the coolth emitters then the air blast heat exchanger 104 would by targeting a brine exit temperature of 14°C effectively be delivering a brine temperature at entry to the coolth emitters of 9°C to 12°C. Allowing for dead band effects on temperature set points it can be seen that in this example a SCG set point of say 10°C may well result in the SCG reading a stored coolth value of 50%, which in the absence of a low SEL reading for stored warmth would not necessarily require for the compressor to be put into operation. In other words a free- cooling service via the air blast heat exchanger 104 acting alone would be sufficient to keep the air conditioning system operating within its required operating range.

If in the same example the compressor was later compelled by a low SEL reading to operate (at the same time as which the air blast heat exchanger 104 was operating) then in practice the air blast heat exchanger would need to reduce its cooling output and its fans would be therefore slowed down. Indeed if the compressor was operated at high enough speed for long enough then the air blast heat exchanger 104 fans would be switched off and parallel configuration might be engaged, eventually leading to the fans being switched back on so that the air blast heat exchanger could dump excess coolth emanating from the evaporator as a result of the compressor being in operation.

Broadly speaking we have described how a LTHW system and a chilled brine system may be driven from a common heatpump-chiller that automatically switches its emphasis between LTHW production (in which case the heatpump-chiller focuses on providing load to the thermal storage tank) and heat dumping (in which case heat is deliberately dumped in order to generate extra cooling effect in the brine circuit). We now give some examples of operation of the cooling side control.

Example A: series-mode circuit

Referring to Figure 8, which corresponds to Figure 1 b, this example illustrates the effect of using an air blast heat exchanger 104 for cooling during colder weather. Figure 8a shows temperatures in the system at t = 0 minutes and Figure 8b at t = 2 minutes. Figure 8c corresponds to a version of Figure 7 and illustrates the main features of a lookup table for control of the (fan of) cooler 104. Figure 8a corresponds to point a _! in Figure 8c and Figure 8b to point a ₂ .

The target coolant temperature (in the mixing header) is, in this example, 21 °C, a temperature suitable, for example, for an industrial process. The ambient temperature is significantly lower than this and if the temperature of the mixing header becomes excessive this can quickly be brought under control by the control of the cooler fan(s), even if the compressor is not providing much cooling effect. As previously discussed, in embodiments the scan rate of the GAM control procedure may be 90 seconds as compared to 600 seconds for the heating side (SCG + SEL) control. Thus the GAM control "fine tunes" the step changes implemented by the heating side control.

Example B: parallel-mode circuit

Referring to Figure 9, which corresponds to Figure 1 a, this example illustrates operation of the system with the cooling side in a parallel mode of operation. Thus Figures 9a to 9d show temperatures in the system at, respectively t = 0, 6, 12, 18 minutes, and Figure 9e illustrates the main features of a lookup table for control of the (fan of) cooler 104. In Figure 9a the recirculation port 138 of mixing valve 1 16 is closed; in Figure 9b it is partially open; in Figure 9c it is wide open; and in Figure 9d it is partially open once again. The SEL and SCG "gauges" are also illustrated as insets. The target coolant temperature (in the mixing header) is, in this example, 15°C, and the target suction header temperature is 10.0 °C. In broad terms, the air blast heat exchanger fan(s) and valves are operated in such a way as to influence the temperature of the mixer header 1 10. The SCG uses the temperature of the suction header 1 12 as a means of determining how much or how little coolth is available to the pumps which draw from the suction header. A short explanation of the operation at each step is given below (where b _! to b ₄ refer to the points in Figure 9e):

t = 0 mins (bi) - Figure 9a

T _A MB = 23°C, T _M ix = 22°C, Cooler fan(s) = NIL. The cooling side circuit is set in parallel mode.

The set point for T _SUC . _H DR = 10 C; T _SUC . _H DR (ACTUAL) = 14.5 ^U C. suc-HDR (ACT-TGT) >1 therefore the stored Coolth Gauge (SCH) reads 25%

t = 6 mins (b?) - Figure 9b

Due to the low cooling potential of the suction header (SCG = 25%) the compressor speed has been incremented. After a short time the suction header became colder and a value of T _SUC . _H DR = 9.0 °C was read, which is within the target range for T _SUC . _H DR - and so SCG = 50%. At the same time it was noted that T _M IX HDR = 14.5 °C, which is still in the no fan zone (allowing for a dead band).

t = 12 mins (ba) - Figure 9c

Strong demand for heating has depleted the thermal store. The Stored Energy Level (SEL) in the thermal store reads 25%. The combination of SEL = 25% and SCG = 75% sends a "load signal" to the compressor. This forces a low value of T _M ix HDR = 6.0 °C and the cooler 104 is set to run at "Parallel-Mid Speed Fan" (Figure 9e). t = 18 mins (b ₄ ) - Figure 9d

T _M ix HDR = 14 °C, which is close to ideal, so the cooler fan(s) are slowed down. Since increasing the compressor speed (see above) the thermal store has been replenished and SEL = 50%. At the same time T _SU C-HDR = 7.0°C which equates to SCC = 75%. At SEL = 50% and SCG = 75% the compressor is unloaded.

POWER CONTROL SYSTEMS

Referring now to Figures 10a and 10b these show, respectively, an embodiment of a power control architecture 1000 for the above described combined heating and cooling system, and a perspective view of a physical configuration for such a system.

Thus Figure 10a shows a 3 phase grid mains supply 1002 which, in the UK, provides a nominal phase voltage of around 400 volts RMS. This is provided to a so-called active front end (AFE) 1004 which converts this to a DC supply, typically around 700 volts DC, which may then be reduced to, for example, around 600 volts DC. A passive front end (passive rectification) may also be employed; in the following description although the acronym "AFE" is used this is shorthand for either an active or a passive front end. A typical AFE employs an IGBT (insulated gate bipolar transistor) half bridge for each phase, controllable to control current through the AFE and indirectly to control the output voltage. Preferably, though not essentially, an EMC (electromagnetic compatibility) filter is provided on the AC side of the AFE; optionally an isolation transformer may also be included.

The AFE 1004 provides power to internal DC power bus 1006 but which may be a 2 wire bus as shown or a 3 wire bus. The previously described heating and cooling system includes a number of motors, in the illustrated example a compressor motor 1008 (for compressor 214) a gas cooling pump motor 1010 (or gas cooling pump 236) and an air blast fan motor for dry air cooler 104). Each of these is powered by a respective variable speed (frequency) drive module 1014, 1016, 1018, coupled to bus 1006. Present but not shown in Figure 10a are also various stepper motor valve actuators which may also be powered from bus 1006, albeit indirectly since they typically have a low voltage supply.

The system includes a Battery Energy Storage System (BESS) 1020 coupled to the bus 1006 via a bi-directional power converter interface 1022. In this example bi- directional converter 1022 is a DC/DC converter which allows control of the current flow in either direction; the skilled person will be aware of many suitable topologies for this module. In embodiments the Battery Energy Storage System comprises a pack of rechargeable batteries; any of a range of technologies may be employed, depending upon trade-offs including cost, including for example, lead-acid and lithium ion technologies. In some preferred embodiments the battery pack voltage is less than the voltage of the DC bus, but still relatively high for efficient operation, for example greater than 100 volts, in embodiments around 400 volts.

Optionally the system may include a further DC/DC converter 1024 to enable a feed into bus 1006 from a solar array 1026. Further optionally (although not shown in Figure 10a) the system may be used to export power to grid mains 1002. Conveniently this may be achieved by coupling the output of a variable frequency drive, such as VFD 1014 for the compressor motor, to the grid mains. This may be achieved by providing a changeover contact on each output of AFE 1014 to direct the AC power either to motor 1008 or to grid mains 1002.

The bus-connected elements of Figure 10a are, in embodiments, each controllable although the control lines are omitted from the Figure for ease of representation. In embodiments these control lines are connected either separately or via a further bus to a programmable logic controller (PLC) 1028. PLC 1028 is optionally provided with a user interface 1030 locally and/or remotely via a network, for system control and management. The PLC 1028 may also provide monitoring and control of the thermal system by means of sensors and valve actuators as previously described. In operation PLC 1028 is able to control the current (power) flow through AFE 1004, bi-directional power converter 1022, and the VFDs 1014, 1016, 1018, as well as controlling other functions/behaviours of the system. In embodiments the PLC 1028 may receive a demand control signal from user interface 1030, and/or from an external source such as an electricity generator or distributor, and/or from a fleet control server as described later, for example via a network interface 1031 .

Thus, broadly speaking, the motors may be driven by power from the grid-mains, or by power from the battery pack, or by a combination of the two. In addition the battery pack can be charged from the grid-mains. Optionally a proportion of power for the system may be provided from a solar array. Additionally (though not shown) in Figure 10a, bus 1006 may be employed to transfer power between one unit and another, more particularly to transfer energy in either direction between the battery pack of a first unit and the battery pack of a second unit. This may be achieved by means of a further bidirectional DC/DC interface onto bus 1006 to provide a (DC) link interface for the system.

In broad terms the architecture of Figure 10a is able to provide a number of energy efficiency benefits. Some of these benefits are "local" to the system, others are benefits to the gird mains electricity supplier/network operator since they permit demand management/capping of the unit (although this may indirectly translate to a benefit for the user of the system via a reduction in electricity cost). As described later, co-ordinated management of the battery storage and thermal storage allow improved profile of the electricity demand from the grid-mains and facilitate a user in reducing their energy cost for, for example, HVAC (heating, ventilation and air-conditioning). Embodiments of the system also facilitate use of renewable energy generated either locally or remotely, because of the electrical energy storage provided.

It is helpful to aid understanding of the operation to provide some illustrative numbers for power flows. Thus the compressor motor 1008 may draw around 20kW peak; the pump and other motors may draw substantially less, for example 3-1 OkW peak; and the stepper motors typically require no more than a few hundred watts in total. The maximum power through AFE 1004 may be of order 45kW; the capacity of the battery pack 1020 may typically be in the range 50-100kWh (with the preceding power values) with, for example, a 35kW peak power transfer from the battery pack onto bus 1006. The solar feed 1026 may potentially provide up to 50kW. Where implemented a power transfer facility from one unit to another of up to, say, 10kW may be provided.

Referring now to Figure 10b, this shows an example physical implementation of the architecture of Figure 10a in combination with a combined heating and cooling system of the type previously described. Thus referring to Figure 10b, and where like elements to those previously described are indicated by like reference numerals, this shows a battery energy storage system 1020 comprising a plurality of batteries, a cold water storage tank 1032, a stratified thermal storage tank 232 (instrumented with temperature sensors), a compressor VFD unit 1014, three smaller VFD units 1016, 1018, 1034 for pumps and the like, and a multi-port interface unit 1036 providing the functions of interface units 1004, 1022 and 1024. Referring now to Figure 1 1 a, this shows a flow diagram of a procedure for responding to an external demand control signal indicating a load management event. The demand control signal may be provided, for example, from an electricity generator or network operator or from a controller controlling multiple heat pump systems of the type described. The demand control signal may, in embodiments, be a regularly updated signal, being updated for example every 5 or 10 minutes. Alternatively the signal may be provided in response to network conditions, when it may also indicate an expected duration of the event.

The procedure of Figure 1 1 a may run on PLC 1028. Thus at step 1 100 the procedure/PLC receives the demand control signal, and at step 1 102 determines whether or not the battery energy storage system (BESS) will lose its working charge (according to some criterion) before the event ends. The duration of the event may be specified with the signal, or the duration may, for example, be taken to be the interval to the next update of the demand control signal, or it may be known in some other way. In embodiments the determination may be based upon a measurement or estimate of the electrical energy stored in the BESS and upon a measured or assumed power consumption of the system.

If it is determined that the battery power is a potentially insufficient to run the heat pump system for at least the presently known duration of the load management event then the procedure interrogates data relating to the stored thermal energy (step 1 104). If the stored thermal energy is less than a threshold, for example if the thermal store is less than 50% full, then the procedure sets a cap on the speed of the compressor motor to a value "MINIMUM". This cap may be implemented by communicating the cap to the thermal system control procedure shown in Figure 6, where it acts upon the steps involving a change in compressor motor speed. Thus by way of example the compressor motor may normally have a maximum speed of 1500 rpm and the value of "MINIMUM" may be set to 800 rpm. This has the effect of providing an immediate limitation on the power consumed by the heat pump system. However in embodiments of the power control architecture this does not necessarily reduce the power being drawn from the battery pack because this is controlled (to a degree independently) by the power control signal applied to the bi-directional convertor 1022. Thus because the battery pack is predicted to lose its working charge whilst the load management event is ongoing, the procedure also controls the bi-directional converter 1022 to reduce the power from the battery pack provided to bus 1006. In embodiments, rather than immediately shut off this power it is preferable to progressively reduce the power provided from the battery pack to the bus/compressor towards zero (step 1 108). As the power from the battery pack reduces so the power from the grid mains potentially increases (depending upon where the thermal control loops sets the compressor). However the overall compressor speed is nonetheless limited, limiting power drawn from the grid mains even when the power from the battery pack is at zero (when the battery pack may, nevertheless, have some residual charge). In effect by progressively reducing the battery power the system is aiming to extend the use of the battery power over the load management event, or at least until the next update of the demand control signal. In embodiments the progressive reduction of power from the battery pack may be achieved by a cyclical, stepwise power reduction in which the cycle time is smaller than the duration of the load management event (if known) and/or smaller than the time interval to the next update of the demand control signal.

The above described procedure is implemented when both electrical and thermal storage are low. Where, at step 1 104, it is determined that the stored thermal energy as above a threshold, for example because the thermal storage tank is equal to or greater than 50% full, then again the compressor maximum speed is capped (step 1 1 10). However in this case the cap is set at a higher level than the previous MINIMUM, for example MINIMUM plus Δ. Continuing the previous example, this cap may be, for example, l OOOrpm. This effectively anticipates that there will be less need to draw upon the stored electrical energy because the heat pump can rely more upon the stored thermal energy. Then, as before, the procedure progressively reduces the power being supplied by the battery pack onto bus 1006 (step 1 1 12).

Where at step 1 102 the procedure determines that the battery pack has sufficient charge to run the system, at least until the next demand control signal update (or for the duration of the load management event), the procedure uncaps the compressor speed (step 1 1 14) if necessary. Preferably the procedure then sets the power flow from the battery pack to the bus 1006 to a maximum (step 1 1 16), or at least controls the interface 1022 so that the power which is able to flow from the battery pack onto the bus is the maximum permitted by design (noting that the actual power may depend upon the actual power drawn from bus 1006 at a particular time by the compressor and other motors). In embodiments, in this set of circumstances the operation of the procedure does not depend upon the level of stored thermal energy.

Figure 1 1 b shows a variant of the procedure of Figure 1 1 a which is preferred in some implementations of the system. Thus in response to an external signal 1 100 the procedure first checks the level of thermal storage 1 152. If this is adequate, say =>50%, no restriction is placed on the compressor 1 154, and the procedure loops to maintain a check on the level of the thermal store. However if the thermal store is low, say <50%, the procedure then checks the level of stored electrical energy 1 156. If this is adequate, for example because it is predicted not to lose its working charge before the end of the event, then again there is no restriction on the compressor 1 154. However if the battery system has insufficient stored charge then the compressor operation is restricted 1 158, for example by imposing a cap on the compressor motor speed, and the level of stored electrical energy is rechecked 1 160. If this is now adequate then the grid draw may be maintained 1 162, but if not the gird draw may be increased 1 164.

In some embodiments the grid draw is permitted to increase if, after restricting the compressor, the system is still unable to draw sufficient power from the BESS. In other embodiments, however, the power drawn from the grid may be an overriding control parameter, in which case the system (heat pump) operation may be further curtailed or may cease entirely when the both the stored thermal energy and stored electrical energy run out. Referring to Figure 12, this shows a procedure for operation of the power control system in the absence of a demand control signal, that is where there is no load management event indicated. Thus when the procedure determines that the demand control signal is absent (step 1200) a determination is then made as to whether the grid mains power is "green", as previously described, and/or low cost. This determination may be made based upon an external signal, for example indicating that the electricity from the supplier is low-carbon electricity and/or the determination may be made based upon the time of day/week/month/year according to a known or expected pattern or tariff. Where the electricity is green then, in embodiments irrespective of the stored thermal energy, the procedure uncaps the compressor speed (step 1204) and, essentially, attempts to charge the battery pack at the maximum rate. Thus the procedure may control the bi-directional converter 1022 to provide power from bus 1006 to the battery pack 1020 and may set the power flow at a maximum (step 1206). The system then progressively charges the battery pack (battery energy storage system - BESS) until this is full (step 1208). The particular charging algorithm may depend upon the battery technology but typically charging will slow as the battery pack nears capacity. In a variant of this procedure optionally the system may, for example after step 1204, identify when the electricity supply is ultra-low carbon, that is below a further threshold of carbon usage in generating the supply. Charging of the BESS may then be conditional on the supply carbon level being below this further threshold; if not the system may neither charge nor discharge the BESS.

If the grid mains supply is determined at step 1202 to be either or both of expensive and not green then the procedure still uncaps the compressor speed (step 1210) and sets the power flow through the bi-directional converter to the battery pack to be a permitted maximum (step 1212). However in this instance the system controls the charging of the battery pack to restrict the charging rate (step 1214). The restriction of charging rate may be performed in many ways depending, in part, upon the battery technology. However typically this may be achieved by reducing the charging rate to merely trickle charge the battery once the battery is above a threshold stored charge, for example when the battery pack is more than ¾ full. This approach enables the battery pack to be sufficiently charged to efficiency handle load management events whilst still making effective use of the availability at other times of cheaper/greener grid mains power.

In a variant of this procedure, rather than charging at a reduced rate (steps 1212, 1214), the system may drain energy from the BESS when the supply is "not green" (and/or expensive).

Still further additionally or alternatively the system may receive a signal indicating that a grid surplus is present. In this case, where the BESS is less than fully charged the system may increase grid draw to charge the BESS. Referring now to Figure 13, this shows, in simplified form, examples of power flow during operation of the above described procedures. The upper diagram illustrates charging of the battery pack during a period when no load management event is present; the lower diagram shows an example of power flows during a load management event. For the sake of comparison in both cases the same power (18kW) is assumed to be used by the compressor. The diagram illustrates in simplified form that, particularly when green electricity is available from the grid the mains, power consumption is increased and used to charge the battery. By contrast in a load management event the mains power consumption is reduced as compared with that used by the compressor, the balance of power being provided from the battery pack. It will be appreciated, however, from the foregoing description that in embodiments of the system the detailed operation is more subtle, in particular taking account of the stored thermal energy and/or coolth.

By way of further illustration, consider the example of a system in which an external demand signal requires the heat pump to operate for a known duration, say 20 minutes, before reverting back to local control. During this period the demand signal may also, implicitly or explicitly, require a target level of rid support from the system, which may be expressed as a target upper limit of power to be drawn from the grid. For example it might be required that throughout a 20 minute duration the heatpump should reduce its present 22kW grid draw to say 12kW of grid draw - which may be achieved by draining the battery at a rate of 10KW). Alternatively it may be the case that the heat pump is required (by the demand control signal or another signal) to increase its grid draw for, say, a 30 minute period from its current rate of say 22kW to a rate of say 35kW of grid draw - which may be achieved by filling the battery at a rate of 13kW. In embodiments, when responding to an external control signal the battery energy storage system is always used as part of the power management system, in in a battery drain mode, which reduces grid-draw, or in a battery fill mode, in which case the grid draw is increased.

In a state where the system is responding to an external control signal for a duration, when the battery is being drained, the battery may also be seen to be insufficiently charged part way through the event in such a way that the originally targeted grid draw cannot be suppressed for the duration of the event. If at the same time the Thermal Energy Store (TES) is at a low enough level that compressor loading would normally be triggered, then the compressor is set at a low output and is prevented from loading the electrical power supply unduly. The principle then is that the suppressed grid draw that is required from the heat pump system for the (entire) duration of the event is maintained not just by draining the battery but also by reducing the power consumption of the heat pump. Thus in the above example where the battery is required to operate in drain mode at a rate of 10kW for an entire 20 minutes it might be that the battery drain rate is reduced at some point, for example "minute 14", to 5kW and that the compressor's power draw is then also reduced by 5kW. In this way the system can ensure that the grid draw of the overall heat pump between,, say "minute 14" and "minute 20", remains at the targeted 12kW.

Broadly speaking we have described a heat pump system which includes a battery energy storage system and bi-directional power converter, as well as a thermal storage system that, in embodiments, collects heated water from a condenser or gas cooler. A measure of the stored thermal energy is used by a procedure which governs the behaviour of the refrigerant compressor to provide thermal control. In addition power drawn by the heat pump from the grid mains, power through the bi-directional converter, and power used by the compressor, is controlled according to a power management procedure which is dependent upon the status (stored charge) of the battery energy storage system, upon the stored thermal energy (typically in the form of heated water), and upon the status of an incoming signal from a higher level energy management system such as the UK National Grid's Demand Side Response (DSR) system. In embodiments the power control procedure operates, in part by controlling the thermal control procedure, more particularly by limiting the compressor power.

In broad terms the control techniques we have described facilitate management of the power drawn from the grid mains according to the (external) demand control signal so that the level of power drawn from the grid can be increased or decreased without substantially adversely affecting the performance of the heat pump system. In embodiments, dependent upon the stored electrical energy, power consumption from the grid may be controlled by capping the flow rate of the compressor(s) in a way which overrides other signals (for example from the thermal control system) which might otherwise cause the compressor to increase its flow rate. For example, where the amount of stored thermal energy inside, say, an accumulator at the outlet (hot) side of the heat pump is sufficiently low that it would cause the compressor to increase its flow rate, the ramp-up signal resulting from the low level of stored heat energy is suppressed and/or replaced by a ramp-down signal. The compressor speed (and hence heating/cooling capacity) may be capped for a period of time (there may be a pause after steps 1 106, 1 1 10 described above), and then the power drawn from the battery energy storage system may also be gradually reduced, in embodiments until substantially all power flowing to the compressor is derived from the grid mains without substantial battery assistance.

Fleet control and other aspects

Referring now to Figure 14a, this shows an embodiment of a heat pump system 1400 including a power control architecture 1000 as previously described. The heat pump system 1400 may be configured as a combined heating and cooling system as previously described, or the heating circuit may be omitted and the system may be configured for a refrigeration-only application.

Figure 14b shows an energy management system 1450 comprising a plurality or fleet of heat pump systems 1400, 1 ...n, configured for control by a fleet control server 1452. In the illustrated example the heat pump systems 1400 are coupled to fleet control server 1452 via the Internet 1456, although other forms of wired and/or wireless connection may be employed. The fleet control server 1452 has a user interface 1453 for controlling the heat pump systems in a coordinated manner via server 1452, for example to globally limit (or reduce to zero) the power drawn from the grid by a group of the heat pump systems. Thus the user interface may allow a user (supervisor) to group one or more heat pump systems and to provide the same or a corresponding demand control signal to each heat pump system in a group. This may be controlled in real time by the user and/or server 1452 may enable a user to define one or more time windows during which grid power drain is limited, and/or server 1452 may enable a user to define that grid power drain is limited when a demand control signal 1458 from an electricity generator/distributor 1454 indicates that a reduced demand is desirable. As the skilled person will be aware, server 1452 typically comprises a computer system with at least one communications interface, and a processor running under control of a program stored in memory to perform the above-described functions. Figure 15a shows an example of a refrigeration-only heat pump system 1500 installed in a building. The heat pump system 1500 includes a user interface 1502 to enable a user to define how system 1500 uses power from grid mains supply 1510, 1512, and in particular to enable a user to define a demand control signal for heat pump system 1500. As illustrated, optionally heat pump system 1500 may deliver power back into the grid as well as drawing power from the grid. The skilled person will appreciate that this is similarly a feature of the heat pump systems previously described above. Figure 15b shows an example of a refrigeration-only heat pump system 1500 installed in an electric vehicle 1550 to provide cooling for a refrigerated chamber 1552, for example for temperature-controlled cargo. In the illustrated example the heat pump system 1500 has a grid-detect module 1554 to detect when the vehicle is able to derive power from the grid mains supply and when it needs to rely upon a vehicle battery (which may partially or completely comprise a rechargeable battery for the drive system of vehicle 1550). When the heat pump system 1500 needs to rely solely upon the vehicle battery a demand signal is generated to inhibit grid mains drain accordingly.

As previously described, in embodiments the heat pump system 1500 may comprise power electronics assembly including variable frequency drive (VFD) unit configured to drive a variable speed C02 refrigerant compressor. A bidirectional DC:DC battery charger enables transfer of part of all of its power output into the VFD from a battery storage system at a rate dependent upon one or more signals received from a programmable logic controller (PLC) or other processor/controller.

In embodiments the PLC is configured to send ramp-up and/or ramp-down signals to the VFD. The operating speed of the compressor may be restricted at a speed below a normally permissible maximum, and hence the power being transferred out of the battery by the bidirectional converter may also be reduced, when a set of factors are simultaneously present and detected by the PLC: The charge in the battery storage system is sufficiently low that battery energy must be conserved if the overall assembly is to keep running off-grid (or partially off-grid) for a given duration; one or more temperatures in one or more refrigerated chambers have risen above a set point that would normally trigger compressor ramp-up; and there is a temporary shortage or complete lack of mains power from the public AC grid. A similar set of factors may control a refrigeration-only heat pump system when installed elsewhere, for example in a building as described with reference to Figure 15a.

No doubt many other effective alternatives will occur to the skilled person. It will be understood that the invention is not limited to the described embodiments and encompasses modifications apparent to those skilled in the art lying within the spirit and scope of the claims appended hereto.