ELECTRICAL ARCHITECTURE COMPRISING AT LEAST ONE LINEAR PHOTOVOLTAIC INSTALLATION FORMED BY SEVERAL GROUPS OF PHOTOVOLTAIC PANELS AND BY A DC NETWORK, CONNECTED TO AN AC TRANSMISSION NETWORK AND/OR AN AC DISTRIBUTION NETWORK WITH ARBITRATION OF THE POWER INJECTED FROM THE DC NETWORK TO THE AC NETWORK

Abstract

An electrical architecture including at least one linear photovoltaic installation formed by several groups of photovoltaic panels and by a DC network, connected to an AC transmission network and/or an AC distribution network with arbitration of the power injected from the DC network to the AC network. The system substantially entails putting in place an architecture with at least one linear PV installation with a DC network and interconnecting this subassembly at at least two separate points of interconnection with a preferably existing AC power grid. Each point of interconnection to a node of the AC network is a voltage source converter VSC which can inject 0 to 100% of the maximum power P of the linear PV installations.

Claims

1. An electrical architecture comprising: at least one linear installation comprising at least one group of photovoltaic panels suitable for producing a maximum total power P, and a direct current network comprising at least one bus, to which the group(s) of PV panels are electrically parallel-connected, each via a DC/DC converter, an alternating current transmission and/or distribution network, at least two voltage source converters, one of the two converters connecting the DC bus to a first node of the AC network, the other of the two converters connecting the DC bus to a second node of the AC network, separate from the first node, each of the VSCs being suitable for injecting 0 to 100% of the power P into the AC network, a control system suitable for allocating the injection of power between the VSCs according to needs and/or operating conditions of the AC network, so as to reduce the total losses of the latter and/or improve the quality of service of the AC network.

2. The architecture according to claim 1, the VSCs being modular multi-level converters.

3. The architecture according to claim 1, the converters being controlled according to a control mode for injected power and for the voltage at the point of connection of the network or according to a control mode for active and reactive power injected at the AC network.

4. The architecture according to claim 1, the bus, to which the group(s) of PV panels are directly electrically parallel-connected, being a medium voltage DC bus.

5. The architecture according to claim 4, comprising several geographically distributed loads, such as high power electric vehicle charging stations or electrolysers for supplying hydrogen-operated vehicles, each connected via a DC/DC converter to the MVDC bus.

6. The architecture according to claim 4, comprising several geographically distributed electrical storage means, including batteries, each connected via a DC/DC converter to the MVDC bus.

7. The architecture according to claim 4, comprising other geographically distributed current sources each connected via a DC/DC converter to the MVDC bus.

8. The architecture according to claim 4, the DC network of the linear installation comprising at least one high voltage DC bus connected to the medium voltage DC bus and to a voltage source converter connected to a node of the AC network.

9. The architecture according to claim 1, the control system being connected to the real-time data acquisition and control system of the AC network.

10. The architecture according to claim 1, comprising voltage and/or frequency measurement means at the first and second nodes, connected to the control system such that it allocates the injection of power between the VSCs according to the measurements carried out.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0066] FIG. 1 schematically illustrates an electrical architecture according to the invention, as may be installed in the south-east region of France.

[0067] FIG. 2 is a block-diagram view of the electrical architecture according to the invention.

[0068] FIG. 3 is a schematic view illustrating an architecture according to the invention restricted to two connections per VSC between the DC network and the AC network.

[0069] FIG. 4 is a schematic view of the “New England IEEE 39-bus” test network on which tests of power injection from the DC network according to the electrical architecture of the invention have been carried out.

DETAILED DESCRIPTION

[0070] FIG. 1 shows an architecture according to the invention, as may be installed in the south-east region of France.

[0071] As illustrated, this electrical architecture comprises a plurality 1 of groups of photovoltaic (PV) panels 10 suitable for producing a maximum total power P. These groups of panels 10 are distributed over hundreds of kilometres in this region by being arranged over ground-based surfaces stretched out over man-made terrain, such as along the edges of railway lines, motorways, etc.

[0072] All these groups of panels are electrically parallel-connected by at least one bus 20, 21 of a direct current (DC) network 2 to form a single linear PV installation. This linear PV installation is itself connected to the alternating current (AC) transmission 3 and/or distribution 4 network already existing in this region. All the groups of PV panels are therefore interconnected by at least one DC bus 20, 21. FIG. 2 shows in greater detail the components and possible connection modes according to one embodiment of such an architecture.

[0073] In this FIG. 2, the groups of PV panels are distributed along a line L on the ground. The plurality 1 of these PV groups can produce a maximum total power P.

[0074] The linear PV installation comprises the plurality of the groups of panels 10 and the DC network 2, which in this case comprises several MVDC buses 20, to which the groups of panels 10 are electrically parallel-connected, each via a DC/DC converter 11.

[0075] The DC network also comprises several HVDC buses 21 within the linear PV installation, each connected to one or more MVDC buses 20. The HVDC buses 21 can transmit the DC current over the entire DC network.

[0076] Each MVDC bus 20 or HVDC bus 21 of the linear installation is connected to a node of an AC transmission network 3 or to a node of an AC distribution network via a voltage source converter 22.

[0077] Preferably, the VSCs 22 are modular multi-level converters (MMCs).

[0078] According to the invention, each of the VSCs 22 is suitable for injecting 0 to 100% of the power P into the AC network.

[0079] Also according to the invention, a control system for the architecture can distribute the injection of power between all the VSCs 22 according to needs and/or operating conditions of the AC network, so as to reduce the total losses of the latter and/or improve the quality of service over the latter.

[0080] As also illustrated in this FIG. 2, the DC network of the architecture comprises several geographically distributed loads, such as high power electric vehicle charging stations (EVs) or electrolysers (H.sub.2) for hydrogen vehicles. Each of these loads is directly connected to the MVDC bus 20 via a DC/DC converter 23.

[0081] The DC network can also support several geographically distributed electrical storage means, such as batteries. Each of these storage means is also directly connected to the MVDC bus 20 via a DC/DC converter 23.

[0082] To illustrate the operation of the architecture according to the invention, FIG. 3 shows only two VSCs 22 which perform the coupling between the DC network of the linear installation, connected to the plurality 1 of PV groups 10 and the AC network 3, 4. More specifically, the AC network schematically comprises four nodes N1 to N4, the node N1 being the coupling node of the VSC 21, the node N2 being the coupling node of the VSC 22.

[0083] In this configuration, the power injected by each of the converters VSC1 and VSC2 is dependent on the need of the AC network. The transit power P12 between N1 and N2 is in the direction of the arrow indicated in FIG. 3. If the line between N1 and N2 is highly loaded, there is an overload risk.

[0084] In that case, according to the invention, the control system for the architecture can reduce the injected power P1 at VSC1 and increase the injected power P2 at VSC2 in order to reduce the overload on the line N1-N2. This control system is preferably the one integrated in the linear PV installation and which drives the operation of the linear PV installation. This has the advantage of local arbitration autonomy for the injected power P1 and P2.

[0085] Alternatively, there may be a control system external to the linear PV installation, for example connected with the SCADA control system of the AC network.

[0086] Generally, the power between P1 and P2 can be adjusted to reduce the total losses of the AC network or improve its quality of service.

[0087] The proposed architecture according to the invention has been validated on the IEEE New England 39-bus AC transmission test network. This AC network, shown schematically in FIG. 4, is a simplification of the New England network in north-eastern USA. It includes 39 nodes, of which 10 are production nodes (numbered 30 to 39), and 46 lines. In particular, at node 29, there is a generator G9.

[0088] The total production and consumption of the New England AC network are 6147 MW and 6097 MW respectively.

[0089] Software marketed under the name “PowerFactory”, by the company DIgSILENT, has been used to perform the validations on this New England network.

[0090] The validation test assumptions were made considering six groups of PV panels all electrically parallel-connected to a DC bus of the linear PV installation. The maximum power of each PV group is 50 MW, i.e. a maximum total power P of 300 MW for the linear installation.

[0091] The DC bus is connected to two VSCs which are connected to the New England AC network at nodes 26 and 29, referenced VSC_26 and VSC_29 respectively.

[0092] In other words, in the tests, these two VSCs, VSC_26 and VSC_29, inject the production of the six PV groups into the New England AC network.

[0093] An overview of the test configuration is shown in FIG. 4.

[0094] For the tests, the power of these two VSCs is variable as an arbitrary injection, according to needs and/or operating condition of the AC network.

[0095] More specifically, to see the impact of optimal power distribution between the two, VSC_26 and VSC_29, five cases have been tested according to the production power of the generator G9.

[0096] For each of these five cases, both VSC_26 and VSC_29 inject thereto the maximum total power P. For each of the cases, to reduce the transit of power over the lines 29-26 and 29-28, some or all of the power P is injected into both VSC_26 and VSC_29 and the total losses of the AC network are evaluated.

[0097] From one case to the next, the power injected into each of VSC_26 and VSC_29, denoted by P_VSC_26 and P_VSC_29 respectively, according to the production power of the generator G9 (P_G9), is therefore evaluated.

[0098] Table 1 below summarises all the cases, numbered 1 to 5, the power levels and the losses evaluated, with a total power of the linear PV system of 300 MW.

TABLE-US-00001 TABLE 1 P_G9 P_VSC_26 P_VSC_29 Losses Case (MW) (MW) (MW) (MW) Case 1 730 300 0 57.53 Case 2 630 200 100 51.26 Case 3 530 100 200 48.7 Case 4 430 0 300 50.2 Case 5 430 270 30 44.16

[0099] From this Table 1, the following is apparent: [0100] for Case 5, the total losses are minimal; [0101] the total losses are reduced by about 23% from Case 1 to Case 5; [0102] with optimum distribution of power injected between the two VSCs, P_VSC_26 and P_VSC_29, the losses over the AC network can be optimally reduced.

[0103] Consequently, these tests show that if the distribution of power injected on the various VSC coupling points between the linear PV installation DC network, connected to a plurality of groups of PV panels in electrical parallel, and an AC network is optimal, the total losses over the AC network are reduced.

[0104] The invention is not limited to the examples which have just been described; in particular, characteristics of the examples illustrated, among variants that are not illustrated, can be combined with one another.

[0105] Other variants and embodiments can be considered without thereby departing from the scope of the invention.

REFERENCES CITED

[0106] [1]: T. Athay, R. Podmore, and S. Virmani. “A Practical Method for the Direct Analysis of Transient Stability”. In: IEEE Transactions on Power Apparatus and Systems PAS-98 (2 Mar. 1979), pp. 573-584. [0107] [2]: M. A. Pai. “Energy Function Analysis for Power System Stability”. The Kluwer International Series in Engineering and Computer Science. Power Electronics and Power Systems. Boston: Kluwer Academic Publishers, 1989.