WO2009019306A1

WO2009019306A1 - Method for controlling electrical loads in a low-voltage network

Info

Publication number: WO2009019306A1
Application number: PCT/EP2008/060405
Authority: WO
Inventors: Gunnar Kaestle
Original assignee: Gunnar Kaestle
Priority date: 2007-08-07
Filing date: 2008-08-07
Publication date: 2009-02-12
Also published as: DE102007037277A1

Abstract

The invention relates to a low-voltage network NS having controllable current loads connected thereto. Said loads can provide control energy. Control of the effective power drawn by the current loads is done by an active variation of the network voltage by means of a transformer T having a variable transformer ratio.

Description

Method for controlling electrical consumers in the low-voltage network

The invention relates to a method for operating a low-voltage network to which one or more controllable power consumers are connected. The invention is particularly concerned with the provision of balancing or control energy in the power grid by loads on the low voltage grid.

Control energy is needed to keep the generation of electrical energy at the same level of consumption at the same time. This is necessary for stable operation of power networks. A distinction is made here between primary control power, secondary control power and tertiary control power (so-called minute reserve). Normally, control energy is supplied by throttled thermal power plants and quickly approachable and retractable power plants such. B. gas turbines and hydroelectric plants.

Primary control power serves to maintain the balance of power generation and consumption. It is usually automatically retrieved by an active power frequency statics generator sets of power plants. In this case, the active power is changed as a function of the grid frequency on the basis of the statics, a linear controller characteristic. If the mains frequency drops, the active power is increased and vice versa. Secondary control power is used to restore the setpoint frequency of, for example, the 50 Hz in the UCTE network and is fully automatic by means of a network controller of the respective control zone (automatic generation control) with integrative behavior. Activation of the Minute reserve is usually done manually, eg. By telephone call, and has the economic optimization of the interconnected operation for the purpose.

Electrical consumers occur, inter alia, in the low-voltage network. As controllable loads with displaceable power requirement, for example, the compressor motors of heat pumps, air conditioners and refrigerators are mentioned. Due to thermally inert components or because of refrigeration or heat accumulators, they can shift their transit times within certain limits without sacrificing comfort and thus act as functional power storages. The coupling of multiple consumers to a controllable network is known as Demand Side Management (DSM). DSM offers advantages through the synergy effects for the entire system. By the external influence of the individual plants cooperative tasks can be perceived, eg. As for the reduction of power peaks and a concomitant reduction in the price of performance or optimization of network use, but also for the collective delivery of control energy.

One way of coordinating the network of electrical loads may be through a central control room, from which the control and optimization of the DSM takes place. Conventional concepts for DSM require a communication link for each consumer in addition to the power supply. This extra effort for control technology pays off for large-scale, industrial-scale services, but less so for small consumers with a few kW connected load. The costs of communication and control in the LV network are greater than those of the higher voltage levels due to the fine-grained structure of the DSM system and the size distribution of small consumers.

Also known are methods that address a large number of consumers by means of ripple control signals. This is done for example in electric night storage heaters, which receive a release in this way. A disadvantage is the induced load stroke, due to the simultaneity of the distributed consumption and the lack of feedback due to the unidirectional signal flow direction. In US 4,363,974 a methodology is presented in which encoded by means of an on-load tap changer voltage signals, ie a pattern of increments and gradations, commands to electrical equipment. The decoder on the consumers then switches these off and on again for load management and can also control capacitor banks for reactive power compensation. The frequent switching operations reduce the service life of the on-load tap-changer and increase the maintenance effort.

Intelligent loads, which switch on and off depending on the frequency (see US Pat. No. 4,317,049) and voltage deviations, have also already been proposed and implemented in prototypes. One example is the conference contribution Distributed intelligence in power networks by Klaus-Wilhelm Köln to the World Climate & Energy Event (RIO 5), Rio de Janeiro, 2005. Furthermore, reference is made to the publications of the "Dynamic Demand" initiative from the United Kingdom (eg the article "A dynamically-controlled refrigerator" of October 2005, http://www.dynamicdemand.co.uk/pdf_fridge_test.pdf) and Referred to "GridWise" from the US.

The object of the invention is to provide a simple and robust method for operating a low-voltage network that integrates a plurality of controllable power consumers in a stable and self-regulating network operation.

To solve this technical problem, it is proposed according to claim 1 that the regulation of the active power, which is obtained from at least one controllable consumer from the low-voltage network, by an active variation of the mains voltage.

The connection that the active power of a consumer related to the grid depends on the voltage level at its connection point is known per se, such as, for example, B. the voltage-dependent power ohmic loads (P ~ Uc). In contrast, decoupling of the active power consumption from the grid sizes is noticeable due to the ever increasing spread of switched-mode power supplies, electronic ballasts and converter-fed drives. Within the scope of the invention, It is recognized that, in principle, it is also possible to have a linear characteristic curve in the sense of active power voltage statics for controllable loads with directly or indirectly coupled storage, such as, for example, heat pumps, battery chargers or delivery pumps.

The principle of evaluating electrical energy as a function of voltage and thus also influencing consumption has been considered, for example, in the area of an automatic electricity trading system ("The viable route to the regenerative energy industry", Christoph Müller, Pogonon-Verlag, 1992) ). Principally similar approaches are also known from the poster contribution "Netz ohne Kraftwerk" (by Klaus-Wilhelm Köln, 23rd Symposium Photovoltaic Solar Energy, Bad Staffelstein, March 2008).

Based on this general knowledge and the above cognizance, the present invention proposes, in a first basic idea, to regulate the behavior of the (controllable) consumers according to the voltage level Ui at the connection point i. The consumers are controlled so that the power consumption of the consumer increases with increasing mains voltage; with decreasing mains voltage the power decreases.

Preferably, in controllable consumers to a static used, d. H. the rated load is varied depending on the voltage over a linear characteristic. Thus, for example, a battery charger, which calculates a charging power of 3 kW for a recharge, continuously shift this charging power up and down, depending on the level at which the average value of the mains voltage is exceeded or fallen short of. For clocking loads, a linear shift of the turn-on and turn-off thresholds can occur, i. Depending on the voltage, for example, the switching hysteresis of a two-position controller for refrigeration compressors of freezer chests is adjusted - at high voltage the compressor switches on earlier and off earlier at low voltage.

Fig. 1 illustrates the basic idea of the invention: If the voltage Ui drops at the connection node i, the load reduces the reference power Pi and vice versa. Preferably, the consumers are designed as thermally acting devices with a heat or cold buffer; but there are also engines and similar consumers controllable type possible, which essentially depends on the work done in a defined period, such. As the energy that fills the battery of an electric vehicle during a nightly storage cycle again.

In the known systems and methods, the mains voltage in the low-voltage network increases or decreases as a function of the active power taken from the respective network segment and fed in decentrally. In contrast, the invention proposes an active variation of the mains voltage, i. H. For regulating the mains voltage, a variable reference value for the voltage is specified by a superordinate point, for example a control room with automatically operating master computers.

According to the invention, the low-voltage line L itself and, if appropriate, the medium-voltage line serving this purpose serve as the physical layer of the data connection for the network guidance. The usable date, which conveys these as a data bus, is the voltage level. In contrast to the inductive property of medium and high voltage lines, low voltage lines have a predominantly ohmic characteristic. It can therefore be concluded from the voltage drop to the load state of the network segment.

The above-described consumer behavior in the case of grid management results in a uniform flow of energy over the network segment or the low-voltage line L supplying transformer T. The dynamic adjustment of the energy flow through the transformer, as they are changing, inter alia, for the supply of control energy or the departure Generation profiles required according to the invention by the active variation of the voltage level on the low voltage level of the transformer. The variation will expediently take place in the time range of minutes, on the one hand to minimize the switching operations on the tap changer, on the other hand, a fast primary control is easier to implement via the frequency signal. The variable secondary voltage serves as a signal carrier for the request of the regulated power. It should expediently the tolerance band of the standard voltages, z. B. Do not leave according to DIN IEC 60038, as otherwise damage to connected devices may occur and decentralized feeders switch off automatically, thereby provoking network instability.

For active variation of the mains voltage, a transformer with a variable transmission ratio, for example a series regulator with transformer-coupled additional voltage or else an on-load tap changer, is preferably used. The voltage variation can be done in local network stations, but also at a higher voltage level in the substations. As additional resources for voltage regulation in distribution networks are mains voltage regulator in the form of autotransformers, decentralized power generation units with active and reactive power control or medium-voltage DC couplings to call.

If a voltage-variable transformer slightly lowers the voltage, then in the network segment supplied by it, the displaceable LV loads are stimulated to reduced energy consumption, which leads to a reduced network load. Conversely, raising the voltage causes a higher active power reference and thus a larger network load. As an immediate data-technological node, only one voltage-variable transformer has to be integrated into a higher-level control system, which drastically reduces the communication effort. The shift of the reference value for the active variation of the mains voltage is preferably carried out via the on-load tap-changer on the power transformer, but also possible is the variation by means of voltage-variable distribution transformer at the local network station.

The presented method of the invention deviates from the arrangements known in the prior art by violating the iron principle of the energy technology for the operation of AC networks to couple the active power P with that of the frequency f and the reactive power Q with the mains voltage U. In electric power systems, the distributed primary regulation of voltage and frequency by means of local proportional controllers based on Pf and QU statics in the past to a stable Pf and QU-Statiken in the past led to a stable and robust network operation, so that a deviating methodology with the combination of the active power P with the voltage U seems to be inappropriate. Influencing the voltage at the medium voltage level in order to move voltage-sensitive consumers in the low-voltage network for the purpose of tertiary control energy supply to increased or reduced active power consumption, contradicts the practiced practice of regulating the active power as a function of the grid frequency.

The inventive use of the voltage level as an information carrier for the active power reference in low-voltage networks has the advantage that consumers show a self-organizational behavior to the day. They communicate with each other via the voltage as a local signal, so that loads which are equipped with disposable power due to available storage capacities can automatically at least partially compensate for the fluctuating power requirements of non-controllable loads. Furthermore, the proposed method of complexity reduction (plug & play), since autonomous controls interact with the network based on the product parameters of the current, such as voltage and possibly also frequency. The voltage can be influenced as local size as opposed to the frequency in a dedicated network segment. The requirements for the communication effort are greatly reduced, since only sensors for voltage and possibly frequency are needed. The transmitter for the message "less or more load" is already available as equipment with the tap-changer on the power transformer.As a uniform communication standard, only voltage-quality standards such as EN 50160 must be adhered to also integrates participants who have no stress-sensitive behavior.

The method according to the invention is explained below using the example of a heat pump. In Fig. 2, the energy flows are shown within a building B, further the composite to other buildings B via a common low-voltage line L, which is fed by a distribution transformer T with variable transmission ratio. In addition to a partially heat-able heat pump H1, a clocking electric heating cartridge H2 is also shown. If necessary, it serves to provide the desired high temperatures for hot water preparation, as a reserve for the event of a fault and as a peak load, should the performance of the heat pump unit no longer suffice at the lowest outside temperatures. The heating current 1 is composed of the proportion for the heat pump 2 and the heating element 3. Heat pump H1 and heating element H2 generate heat flows at low 4 and high 5 temperature level, which are cached in the heat storage S. The thermal load 6 absorbs this heat. On the electrical side of the heating current 1, the other consumer streams 7 and possibly existing feeds 8 - here for example with a photovoltaic system P - the power balance 9 to the network. Neighboring houses B connected to the same low-voltage power L may also have bidirectional power flows.

Thanks to the heat storage S, a temporal decoupling between the heat demand and production is possible. This allows a network management of electrical loads, d. H. The plant operation of the consumers H1 and H2 depends on the conditions in the power grid. As a result, the reference power is reduced in the event of a heavy network load and the heat pump charges the storage tank during periods of low load.

One form of network control based on locally measurable variables is the voltage control. Since the ohmic resistance layer dominates the inductance coating for low-voltage lines, the heat pump can close the measurement of the voltage at the connection point to the load of the low-voltage network segment. The system control measures the voltage curve and determines typical trajectories depending on the season and the day of the week. Furthermore, it learns the typical requirement profile of the heat requirement as a function of weather conditions, in particular the outside temperature as well as the year and time of day on the basis of measured past heat reference data. Using the requirement profile, the plant controller can create a forecast that estimates the electricity purchase within a forecast horizon. This period depends on the size of the energy storage (here as a heat storage) and the average load this period. In each forecast interval, an adaptive controller adjusts the mode of operation of the heater depending on the expected voltage curve and the energy requirement as well as stored tariff information. Advantageously, selects a self-learning system control for the location and slope of the active power voltage statics in the heat pump H1 or switching thresholds of the hysteresis switch in the heating element H2 such a set of parameters that is preferably related to low grid load power, ie if there is a voltage peak. The value of the current at the transmission network level usually correlates positively with the measured network load at the low-voltage level.

Furthermore, voltage-sensitive consumers enable improved utilization of the network resources, since grid expansion measures due to voltage band violations can be prevented or delayed. Thus, in the example of Fig. 2 in the summer with a strong penetration of PV systems P with peak voltage at midday expected. Heat pumps H1 and H2 electric boilers are preferably in action when the voltage is raised, charging the heat accumulator at noon and thus counteracting the increase in voltage automatically.

Fig. 3 illustrates how live loads in the low voltage grid G-2 are controlled from a central location. In this case, the central point of the on-load tap-changer of the power transformer T1, which is connected to the high-voltage network GO, feeds the medium-voltage grid G-1 and also supplies the low-voltage grid G-2 via the local grid transformer T2. In the respective network levels flow positive and negative active power P and reactive power Q. The system boundary S includes the controllable load, which is connected to the low voltage node G2-i. The global grid frequency f _G o and the voltage at the connection point U _G2 -ι are measured. In the example, by means of the parameters f _ref (setpoint for the frequency [Hz]) and K _f (power factor of the frequency [W / Hz]), a primary control of the referenced active power PQ ₂ -, shown. In addition, the active power consumption P _G2 -ι is linked to the voltage via a parameter pair U _ref (setpoint for the voltage [V]) and Ku (power factor of the voltage [V / Hz]). Here, a processing of the input signal U _G2 -ι (Mains voltage at low voltage node i [V]) by suitable estimators or filters to hide fast voltage changes. The signal to be used is the slow voltage changes in the time range of minutes, as defined, for example, in EN 50160. The voltage regulator on the on-load tap-changer of the power transformer can now deviate from a fixed reference value for U _GI - _O and select a value slightly above or below it. This change in voltage propagates through the medium-voltage and low-voltage links to the consumer, where it is stimulated to draw more or less power.

Furthermore, in the case of a changing network size, as shown e.g. can occur in the separation of the synchronous zone into several smaller network segments as a result of a major disruption or as a result of the transition to the island operation of an area network, as appropriate, the above listed active power frequency and active power voltage statics as a function of network size vary. Thus, in an emergency of a major fault (grid panic mode) or in off-grid operation, the contribution of very small loads to the frequency support (primary control) is more important than during normal operation in the synchronous zone network. In this case, the capacity for tertiary treatment will be transferred to the primary scheme. In small networks, consumers are more sensitive to frequency changes and less susceptible to voltage changes. As soon as the grid has regained its size, it is switched over again, so that the tertiary control based on the voltage versus the primary control dominates on the basis of the frequency and thus reduces the load change reactions of the power consuming unit and the associated wear phenomena.

Incidentally, the same methodology can also be implemented for the reactive power Q at the low-voltage level, eg. B. when consumers are freely adjustable in the choice of their reactive power current, as is the case, for example, for inverter-controlled drives, which have a four-quadrant operation. In this case, the setpoint value for the reactive power Q is dependent both on a voltage statics and on a frequency statics and the parameterization of these statics varies (slope and setpoint value for the voltage U) with the network size. One way to determine changing network sizes is the impedance-hopping method, which is used to detect small island meshes. The emergence of large network islands, such as a distribution network with sufficient self-generation can be communicated for example by a (radio) ripple control system.

Depending on the time of the year and the time of day, the controllable performance changes. So z. For example, in a network area with a large number of heat pumps, the displaceable output in winter tends to be higher than in summer, and in the case of high density of air conditioning systems it is the other way around. This results in a variable elasticity of the power consumption compared to the voltage.

Due to the uncertainty about the internal states of the consumers, in particular the energy content of the thermal storage, the control technology behind the voltage regulator can not assume a deterministic behavior of a single load on the given voltage change. However, the system response is statistically detectable in a variety of live loads.

At a central point in the network, such. B. on a transformer, the power flow Po can be measured and brought into relation with the reference voltage voltage Uo. Voltage changes DU cause power changes DP in the subordinate distribution network segment. In maps, measured annual and time-dependent dP / dU voltage elasticities of the active power are stored. These values can be used to determine which control energy pulse can be generated. A higher-level control system can thus also fulfill DSM tasks via the voltage control equipment in order to optimize network operation or power consumption, for example. On the one hand, this can be done by generating power strokes DP, on the other hand, upper limits _Pmax or quantities derived therefrom can be set to specific operating means, such as eg transformers or lines.

Claims

claims

1. A method for controlling electrical consumers in the low-voltage network, characterized in that electrical energy is obtained from at least one consumer from the low-voltage network, the consumer increases the active power reference with increasing mains voltage and decreases with decreasing mains voltage and this active power reference by varying the reference value for the mains voltage within of the tolerance band of the standard voltages is influenced by a superordinate body.

2. The method according to claim 1, characterized in that the regulation of the active power reference takes place in the time domain of a tertiary control.

3. The method according to claim 1 or 2, characterized in that the consumer behavior by a linear regulator characteristic in the form of a

Active power-voltage statics is determined.

4. The method according to claim 3, characterized in that the consumer behavior in addition to the active power voltage statics is determined by an active power frequency statics and both statics are modified depending on the directly or indirectly determined network size.

5. The method according to any one of claims 1 to 4, characterized marked, that the variation of the mains voltage by on-load tap-changer or longitudinal regulator with additional transformers.

6. The method according to any one of claims 1 to 4, characterized in that the variation of the mains voltage is effected by voltage regulator in the network.

7. The method according to any one of claims 1 to 4, characterized in that the variation of the mains voltage is performed by decentralized power generation units with active or reactive power control.

8. The method according to any one of claims 1 to 4, marked thereby, that the variation of the mains voltage by medium voltage

DC couplings done.

9. The method according to any one of claims 1-8, comprising the steps of: • Measuring the power flow P ₀ in a network segment

• Determining the regulated mains voltage U ₀

• Calculation of a time-dependent dP / dU characteristic

10. The method according to claim 9, characterized in that a higher-level control technology based on dP / dU characteristic curves engages in the voltage regulation of the distribution network.

A system for calculating the voltage elasticity of the power of a distribution network segment, characterized in that the system comprises a data processing device, a voltage regulator and means for determining the power flow and the mains voltage.

12. System according to claim 11, characterized in that a higher-order control system specifies a power stroke DP and the power stroke DP shifts the reference value of the voltage control via the calculated voltage elasticity dP / dU.

13. System according to claim 11, characterized in that a higher-level control system sets a maximum power Pmax and the Reference value of the voltage regulation is lowered according to the calculated voltage elasticity, provided that the measured power flow approaches the predetermined maximum power Pmax.