HUB

Abstract

A hub for transferring power between DC systems. The hub comprises N modules, each for connection to a respective DC system of voltage Vidc and for exchanging power P1 with the respective DC system, and a common p-phase AC bus for connecting the N modules. Each module comprises a DC/AC converter for transforming the respective DC voltage Vidc into a respective p-phase AC voltage Viac of frequency ωs.

Description

The present invention relates to an electronic hub for transferring power between a plurality of DC systems.

DC (direct current) power transmission is significantly better than AC (alternating current) transmission when a single line at a single voltage level is considered. However, DC transmission is not widespread because of the difficulties in voltage stepping and fault isolation. There are many point-to-point HVDC (High Voltage Direct Current) links worldwide which are justified where performance benefits outweigh additional converter costs. In the offshore environment (cable systems), AC transmission has extremely poor performance that necessitates HVDC use. VSC (Voltage Source Converter) is a major improvement in HVDC technology, and has been use in the field for over 10 years at around 20 projects worldwide. VSC HVDC technology has already been implemented for connecting two offshore wind farms and two oil platforms in Europe.

The offshore environment is the primary driver for advancing HVDC to meshed DC grids. However, there are others. Worldwide, there are many HVDC lines operating solely as point-to-point links. There would be significant performance/operational and cost benefits if these lines could be interconnected or tapped on the DC side. Similar arguments apply for developing an EU-wide DC Supergrid and a North Sea offshore grid [1].

In the last 10 years, major advances have been made in high-power electronics. There is growing confidence that the current status of high power electronics will enable the development of converter systems that will lead to cost-effective DC transmission networks with similar reliability and better performance than AC grids. DC transmission grids have been extensively studied in the past 5 years and they are becoming accepted technology. Manufacturers are already offering 300 kV, 9 kA semiconductor DC CB (circuit breakers) [2] and they confidently tender for simple DC grids. Nevertheless, there are significant outstanding challenges with DC circuit breakers, which relate to costs and protection coordination over large distances. The cost of semiconductor DC CB is high, at the same order of magnitude as full VSC converters or DC/DC converters. Adopting specialised electronic DC circuit breakers implies that DC voltage stepping (DC/DC conversion) will be achieved by other electronic components, which results in a further cost penalty.

Existing AC grids were designed on the assumption that simple and inexpensive AC circuit breakers can be placed at the ends of each transmission line. Similarly, simple electro-mechanical transformers have facilitated ready voltage matching. Ring topologies and highly-meshed networks give a cost-effective solution when circuit breakers are inexpensive. However, DC networks cannot adopt such topologies because of high costs and losses associated with voltage stepping and fault isolation components.

High power DC/DC converters take the role of transformers in AC grids. Unlike AC transformers, DC converters are based on semiconductors and are thus highly controllable. This controllability makes other functions possible in addition to voltage stepping. Previous studies have shown that a DC/DC converter can take the role of a traditional transformer, circuit breaker and power regulator in a single component [3]. It would be desirable to have an electronic DC substation connecting multiple DC lines and independently managing all DC grid control operation and protection functions. The cost of such a DC substation would be high, but performance would be considerably better than an AC substation.

Variable speed machines such as wind generators have low voltage output and use multiple AC/DC conversion stages which can be optimized if DC grids are used. Some manufacturers offer MW class wind generators with DC output [4]. Medium voltage DC collection grids have advantages with large wind power parks and PV (photo-voltaic) arrays [5].

Reference [6] presents a DC/DC converter topology that could achieve very high voltage stepping ratio without intermediate AC transformers. The topology can operate at very high switching frequencies and can be built for high current and high voltage. However, the study primarily focuses on a two-port converter, i.e., interconnecting two DC systems. Numerous 2-port DC/DC converters would be required for a complex DC grid, which results in increased costs. For complex DC grids, it is more cost-effective to use multi-port DC/DC converters (hubs) in order to reduce the number of conversion stages and components.

Reference [6] also presents a multi-terminal DC/DC converter with the capability to transfer power between n DC systems of voltage V₁ and m DC systems of voltage V₂. However, this does not allow for the interconnection of DC systems with more than 2 different voltages. Moreover, the multi-terminal converter disclosed in reference [6] does not allow for flexible connection and disconnection of DC systems.

According to one aspect of the present invention, there is provided a p-phase electronic hub for transferring power between N DC systems of DC voltage V_idc (i=1, 2, . . . N), respectively, the hub comprising:—

N modules, each for connection to a respective DC system of voltage V_idc and for exchanging power P_i with the respective DC system; and

a common p-phase AC bus for connecting the N modules;

wherein each module comprises:—

a DC/AC converter for transforming the respective DC voltage V_idc into a respective p-phase AC voltage v_iac of frequency ω_s, root mean square line-neutral magnitude V_iacm and angle α_i; and

an LC circuit for each phase p, for transferring power between the DC/AC converter of the module and the p-phase AC bus, wherein each LC circuit comprises an inductor L_i, a capacitor C_i for supplying reactive power, and a circuit breaker CB_i for disconnecting the module, connected together at their first terminals, the second terminal of each inductor L_i being connected to the respective phase of the respective AC voltage v_iac, the second terminal of the circuit breaker CB_i being connected to the respective phase of the common AC bus. Preferably, the second terminal of capacitor C_i is connected to a central point of the respective DC voltage V_iac in the case where p=1. The second terminals of the capacitors C_i associated with each phase p are preferably connected together in the case where p>1.

With this configuration, each module incorporates LC circuitry, and the modules are connected to one another by means of a common p-phase AC bus. Accordingly, it is straightforward to connect an additional DC system to the hub simply by connecting an additional module to the p-phase AC bus. Modification of the LC circuits of the modules already connected to the hub is not required. Similarly, it is straightforward to disconnect a DC system from the hub without modification of the remaining modules, by disconnecting a module from the p-phase AC bus.

In particular, the circuit breaker(s) CB_i located at the bus side of capacitors(s) C_i, enable straightforward connection and disconnection of a module and its LC circuit(s) from the hub, such that the basic operating principles of the remaining modules are not affected.

This is in contrast to the DC/DC converter of reference [6], in which each terminal or module comprises a DC/AC converter, and these modules are connected by an inner LCL circuit for each phase. Thus, if a module were to be added or removed, the inner LCL circuit(s) would need to be redesigned in order for the converter to operate.

The hub of the present invention is for transferring power between N DC systems, where N is any positive integer greater than 1. Power may be injected into the hub or absorbed from the hub by each DC system, provided that the total power flowing into the hub is equal to the total power flowing out, which is achieved with appropriate control.

The AC voltages v_iac generated by the DC/AC converters each have p phases, where p is any positive integer greater than 0.

The hub comprises a common p-phase AC bus (which may also be referred to as p common AC buses, where p is the number of phases). The p-phase AC bus comprises one electrical pathway or bus for each phase. However, in some embodiments, the hub may comprise one or more additional pathways or busses. For example, in the case of a two phase hub, three pathways may be provided. One for each phase plus a neutral bus. In the case where p=1 (ie, the single phase topology) the return AC current path is preferably provided by connecting a common module point with the DC voltage central point.

Since the LC circuits of the hub operate with alternating current, commonly available mechanical circuit breakers for use with AC systems may be used as the circuit breakers _Cbi.

Preferably, in use, the voltage of capacitor(s) C_i is regulated such that its fundamental root mean square line-neutral magnitude is V_cm, where V_cm>V_iacm0 for all i, wherein V_iacm0 is the maximum fundamental root mean square line-neutral magnitude of v_iac.

According to a second aspect of the present invention, there is provided a p-phase electronic hub for transferring power between N DC systems of DC voltage V_idc (i=1, 2, . . . N), respectively, the hub comprising:—

N modules, each for connection to a respective DC system of voltage V_idc, and for exchanging power P_i with the respective DC system; and

a common p-phase AC bus for connecting the N modules; wherein each module comprises:—

a DC/AC converter for transforming the respective DC voltage V_idc into a respective p-phase AC voltage v_iac of frequency ω_s, root mean square line-neutral magnitude V_iacm and angle α_i, such that

V _iac =V_iacm cos(α_i)+jV_iacm sin(α_i)=V_iacd+jV_iacq

where V_iac is the phasor of v_iac; and

an LC circuit for each phase p, for transferring power between the DC/AC converter of the module and the common p-phase AC bus, wherein each LC circuit comprises an inductor L_i, a capacitor C_i for supplying reactive power and a circuit breaker CB_i for disconnecting the module, connected together at their first terminals, the second terminal of each inductor L_i being connected to the respective phase of the respective AC voltage V_iac, the second terminal of the circuit breaker CB_i being connected to the respective phase of the common AC bus, the second terminal of capacitor C_i being connected to a central point of respective DC voltage V_idc in the case where p=1, and the second terminals of capacitors C_i associated with each phase p being connected together in the case where p>1;

wherein, in use, the voltage of capacitor(s) C_i is regulated to have a root mean square line-neutral magnitude V_cm, where V_cm>V_iacm for all i;

wherein the value of each inductor L_i is selected according to the formula:—

$L_i=\frac{V_{\mathrm{iacq}}V_{\mathrm{cm}}}{P_{\mathrm{im}}}$

and wherein the value of each capacitor C_i is selected according to the formula:—

$C_i=\frac{V_{\mathrm{cm}}-V_{\mathrm{iacd}}}{{\omega }_s^2V_{\mathrm{cm}}L_i}$

wherein V_iacd and V_iacq are any positive values that satisfy the inequality V_iacd²+V_iacq²≦V_iacm0², wherein V_iacm0 is the maximum value of V_iacm.

In a preferred embodiment, the value of each inductor L_i is selected according to the formula:—

$L_i=\frac{V_{\mathrm{iacm}0}\sqrt{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}}{P_{\mathrm{im}}{\omega }_s}$

and the value of each capacitor C_i is selected according to the formula:—

$C_i=\frac{1}{{\omega }_sV_{\mathrm{cm}}^2}\frac{P_{\mathrm{im}}\sqrt{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}}{V_{\mathrm{iacm}0}}$

where V_iacm0 is the maximum value of V_iacm.

These formulas for L_i and C_i are designed to achieve zero reactive power at module i, at maximum power transfer. This in turn minimises current magnitude and therefore minimises switching and conduction losses at full power.

Preferably, the power exchanged by each module with the respective DC system is controllable by varying V_iacm and α_i of the AC voltage v_iac generated by the DC/AC converter of the respective module.

Preferably, at least one of the N modules is configured for regulating the capacitor voltage v_c to have a fundamental root mean square line-neutral magnitude V_cm.

Preferably, the other(s) of said N modules is/are configured for regulating the power exchanged by that module with the respective DC system.

Where p>1, the second terminals of the capacitors C_i associated with each AC/DC converter may be connected in ring between phases. In particular, where p=3, the second terminals of the capacitors C_i associated with each AC/DC converter may be connected in delta between phases. Other capacitor connection can also supply reactive power in symmetrical and balanced manner.

According to an aspect of the invention there is provided an electronic hub for transferring power between n DC systems respectively of DC voltage V_idc (i=1, 2, . . . n), where n is any positive integer larger than 1, comprising:

n AC/DC converters for respectively transforming DC voltages V_idc into respective p phase AC voltages v_iac of frequency ω_s, root mean square line-neutral magnitude V_iacm and angle α_i;

a LC circuit for each phase for each AC/DC converter, wherein each LC circuit comprises an inductors L_i, a capacitor C_i for supplying reactive power and a circuit breaker CBi connected together at their first terminals, the second terminal of each inductor L_i being connected to the respective phase of the respective AC voltage v_i, the second terminal of CB_i connected to a common hub AC bus for particular phase, wherein the second terminals of all capacitors are connected to respective AC bus and the second terminals of C_i are connected to a common point (star point) or they can be connected in delta between other phases or if p=1 to central point of DC voltage,

wherein the value of the capacitor C_i and each inductor L_i are selected to enable required power transfer at respective DC system, at required DC voltage, to minimize reactive power exchange between AC/DC converters and to minimize hub losses.

According to an aspect of the invention there is provided an electronic hub for transferring power between n DC systems respectively of DC voltage V_idc (i=1, 2, . . . n), where n is any positive integer larger than 1, comprising:

n AC/DC converters for respectively transforming DC voltages V_idc into respective p phase AC voltages V_iac of frequency ω_s, root mean square line-neutral magnitude V_iacm and angle α_i, also expressed as

V _iac =V_iacm cos(α_i′)+jV_iacm sin(α_i′)=V_iacd+jV_iacq

where V_iac is the phasor of v_iac;and exchanging power P_i between i-th DC system the hub comprising:

a LC circuit for each phase for each AC/DC converter, wherein each LC circuit comprises an inductors L_i, a capacitor C_i for supplying reactive power and a circuit breaker CB_i connected together at their first terminals, the second terminal of each inductor Li being connected to the respective phase of the respective AC voltage V_iac, the second terminal of CB_i connected to a common hub AC bus for particular phase, wherein the second terminals of all capacitors are connected to a common point (star point) or they can be connected in delta between other phases, or if p=1 to central point of DC voltage,

wherein the voltage of capacitors is regulated at nominal value V_cm, V_cm being larger than V_iacm0,

wherein the value of each inductor L_i is selected according to the formula:

$L_i=\frac{V_{\mathrm{iacq}}V_{\mathrm{cm}}}{P_{\mathrm{im}}}$

the value of capacitor C_i is selected according to the formula:

$C_i=\frac{V_{\mathrm{cm}}-V_{\mathrm{iacd}}}{{\omega }_s^2V_{\mathrm{cm}}L_i}$

wherein V_iacd and V_iacq are any positive values that satisfy the inequality V_iacd²+V_iacq²≦V_iacm0², wherein V_iacm0 is the maximum value of V_iacm.

According to an aspect of the invention there is provided an electronic hub for transferring power between n DC systems respectively of DC voltage V_idc (i=1, 2, . . . n), where n is any positive integer larger than 1, comprising:

n AC/DC converters for respectively transforming DC voltages V_idc into respective p phase AC voltages v_iac of frequency ω_s, fundamental root mean square line-neutral magnitude V_iacm and angle α_i, and exchanging power P_i between i-th DC system the hub comprising:

a LC circuit for each phase for each AC/DC converter, wherein each LC circuit comprised an inductor L_i, a capacitor C_i for supplying reactive power and a circuit breaker CB_i connected together at their first terminals, the second terminal of each inductor L_i being connected to the respective phase of the respective AC voltage v_iac, the second terminal of CB_i connected to a common hub AC bus for particular phase, wherein the second terminals of all capacitors are connected to a common point (star point) or they can be connected in delta between other phases or if p=1 to central point of DC voltage,

wherein the voltage of capacitors is regulated at nominal value V_cm, V_cm being larger than V_iacm0,

wherein the value of each inductor L_i is selected according to the formula:

$L_i\le \frac{V_{\mathrm{iacm}0}\sqrt{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}}{P_{\mathrm{im}}{\omega }_s}$

the value of capacitor C_i is selected according to the formula:

$C_i\ge \frac{1}{{\omega }_sV_{\mathrm{cm}}^2}\frac{P_{\mathrm{im}}\sqrt{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}}{V_{\mathrm{iacm}0}}$

According to an aspect of the invention there is provided an electronic hub for transferring power between n DC systems respectively of DC voltage V_idc (i=1, 2, . . . n), where n is any positive integer larger than 1, comprising:

n AC/DC converters for respectively transforming DC voltages V_idc into respective p phase AC voltages v_iac of frequency ω_s, fundamental root mean square line-neutral magnitude V_iacm and angle α_i, and exchanging power P_i between i-th DC system the hub comprising:

a LC circuit for each phase for each AC/DC converter, wherein each LC circuit comprises an inductors L_i, a capacitor C_i for supplying reactive power and a circuit breaker CB_i connected together at their first terminals, the second terminal of each inductor L_i being connected to the respective phase of the respective AC voltage v_iac, the second terminal of CB_i connected to a common hub AC bus for particular phase, wherein the second terminals of all capacitors are connected to a common point (star point) or they can be connected in delta between other phases or if p=1 to central point of DC voltage,

wherein the voltage of capacitors is regulated at nominal value V_cm, V_cm being larger than V_iacm0,

wherein the value of each inductor L_i is selected according to the formula:

$L_i\le \frac{V_{\mathrm{iacm}0}\sqrt{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}}{P_{\mathrm{im}}{\omega }_s}$

the value of capacitor C_i is selected according to the formula:

$C_i=\frac{1}{{\omega }_sV_{\mathrm{cm}}^2}\frac{P_{\mathrm{im}}\sqrt{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}}{V_{\mathrm{iacm}0}}$

According to an aspect of the invention there is provided an electronic hub for transferring power between n DC systems respectively of DC voltage V_idc (i=1, 2, . . . n), where n is any positive integer larger than 1, comprising:

n AC/DC converters for respectively transforming DC voltages V_idc into respective p phase AC voltages v_iac of frequency ω_s, root mean square line-neutral magnitude V_iacm and angle α_i, and exchanging power P_i between i-th DC system the hub comprising:

a LC circuit for each phase for each AC/DC converter, wherein each LC circuit comprises an inductor L_i, a capacitor C_i for supplying reactive power and a circuit breaker CB_i connected together at their first terminals, the second terminal of each inductor L_i being connected to the respective phase of the respective AC voltage v_iac, the second terminal of CB_i connected to a common hub AC bus for particular phase, wherein the second terminals of all capacitors are connected to a common point (star point) or they can be connected in delta between other phases or if p=1 to central point of DC voltage,

wherein the voltage of capacitors is regulated at nominal fundamental root mean square line-neutral magnitude of V_cm, V_cm being larger than V_iacm0;

and wherein the value of each inductor L_i and capacitor C_i are selected according to any of the above claims and:

V_iacm and angle α_i are used to control power flow at i-th AC/DC module.

According to an aspect of the invention there is provided an electronic hub for transferring power between n DC systems respectively of DC voltage V_idc (i=1, 2, . . . n), where n is any positive integer larger than 1, comprising:

n AC/DC converters for respectively transforming DC voltages V_idc into respective p phase AC voltages v_iac of frequency ω_s, root mean square line-neutral magnitude V_iacm and angle α_i, and exchanging power P_i between i-th DC system the hub comprising:

a LC circuit for each phase for each AC/DC converter, wherein each LC circuit comprises an inductor L_i, a capacitor C_i for supplying reactive power and a circuit breaker CB_i connected together at their first terminals, the second terminal of each inductor L_i being connected to the respective phase of the respective AC voltage V_iac, the second terminal of CB_i connected to a common hub AC bus for particular phase, wherein the second terminals of all capacitors are connected to a common point (star point) or they can be connected in delta between other phases or if p=1 to central point of DC voltage,

wherein the voltage of capacitors is regulated at nominal value V_cm, V_cm being larger than V_iacm0,

and wherein the value of each inductor L_i and capacitor C_i are selected according to any of the above claims and:

one or more AC/DC converters are used to regulate V_c, at nominal value V_cm, and all other modules regulate local powers P_i with additional control loops responding to Vc variation (droop feedback control).

The present invention will now be described with reference to the accompanying drawings in which:—

FIG. 1 shows a DC grid connected by a 5-module DC hub;

FIG. 2 shows a single phase DC hub topology with N modules;

FIG. 3 shows a two phase DC hub topology with N modules;

FIG. 4 shows a two phase DC hub topology with N modules and a neutral bus;

FIG. 5 shows a three phase DC hub topology with N modules and also shows capacitor delta connection;

FIG. 6( a) illustrates unidirectional modulation for generating an AC voltage from a DC voltage;

FIG. 6( b) illustrates bidirectional modulation for generating an AC voltage from a DC voltage;

FIG. 7 shows the phasor diagram for a 3-module test system;

FIG. 8 shows a control schematic for a Voltage Module of a DC hub;

FIG. 9 shows a control schematic for a Power Module of a DC hub;

FIGS. 10( a)-10(l) show simulation results for a 7-module DC hub, wherein FIGS. 10(a)-10(g) show active power, FIGS. 10(h)-10(k) show reactive power, and FIG. 10(l) shows the capacitor voltages;

FIGS. 11( a)-11(g) show simulation results which illustrate the effect of isolating and connecting a DC transmission line to a 7-module DC hub, wherein FIGS. 11(a)-11(c) show active power, FIGS. 11(d)-11(f) show reactive power and FIG. 11(g) shows capacitor voltage;

FIGS. 12( a)-12(h) show simulation results which illustrate droop control of the 7-module hub, wherein FIG. 12(a) shows the capacitor voltage of the hub and FIGS. 12(b)-12(h) show the power exchanged at each module;

FIGS. 13( a)-13(c) show simulation results which illustrate the ratio of fault current over rated current for a DC fault at the same module (a DC fault at each module is illustrated);

FIGS. 14 (a)-14(h) show simulation results which illustrate the response of a DC hub to a DC fault condition (a permanent DC fault occurs at module 6 at 2.0 s, and module 6 is tripped by its circuit breaker CB₆ at 2.05 s), wherein FIGS. 14(a)-14(g) illustrate AC current of each module and FIG. 14(h) illustrated capacitor voltage.

The term “hub” is used herein in the sense of a common connection point for DC systems.

The term “terminal” is used herein to refer to a contact on an electrical device at which current enters or leaves.

The term “bus” is used to refer to one or more electrical pathways.

FIG. 1 shows a DC-grid which uses a 5-module DC hub. In this example, power is generated by three wind generators, and injected into the hub as DC voltages of ±60 kV, ±80 kV and ±120 kV on DC cables 1, 2 and 3 respectively. Power is drawn from the hub on DC cables 4 and 5 at voltages of ±320 kV and ±300 kV respectively.

FIGS. 2 to 5 show the topology of DC hubs which embody the present invention.

FIG. 2 shows a single phase n-module DC/DC converter or “DC hub” which embodies the present invention.

The hub comprises N modules, represented in FIG. 2 by 4 modules labelled 1, 2, i and N; and a common single phase AC bus with two pathways or busses, Bus_A and Bus_G.

Each module i connects to an external DC system of voltage V_idc, with a transmission line represented by bipolar voltage source V_idc.

Each module i comprises two switches S_1—i, S_2—i arranged as a half-bridge to form a DC/AC converter, an inductor L_i, a capacitor C_i, an AC circuit breaker CB_i.

The DC/AC converter of each module i is connected to transform voltage V_idc into a single phase AC voltage v_iac. In each module, inductor L_i, capacitor C_i and circuit breaker CB_i are connected together at their first terminals; the second terminal of inductor L_i is connected to the AC voltage v_iac; the second terminal of the circuit breaker CB_i is connected to Bus_A; and the second terminal of the capacitor C_i is connected to a Bus_G and to the central point of the bipolar DC voltage V_idc.

The module may be considered as having two AC terminals for connection to Bus_A and Bus_G respectively, such that the circuit breaker CB_i is connected to Bus_A via a first AC terminal, and the capacitor C_i is connected to Bus_G via a second AC terminal.

Further capacitors C_id are optionally provided to filter the harmonics and improve power quality.

FIG. 3 shows a 2-phase DC hub with N modules. The topology is similar to that described in relation to FIG. 2. However, in the topology shown in FIG. 3, the DC/AC converter in each module i comprises four switches S_1—i, S_2—i, S_3—i, S_4—i arranged as a full bridge (two legs), connected to transform DC voltage V_idc into a two phase voltage v_iac.

Further, the common 2-phase AC bus comprises one pathway or bus associated with each phase, Bus_A and Bus_B. Further, each module i comprises an additional inductor L_i, an additional capacitor C_i and an additional circuit breaker CB_i, compared to the topology in FIG. 2, so that there is one of each per phase.

For each phase, one inductor L_i, one capacitor C_i and one circuit breaker CB_i are connected together at their first terminals, the second terminal of the inductor L_i is connected to the respective phase of the AC voltage v_iac, and the second terminal of the circuit breaker CB_i is connected to the respective phase of the common 2-phase AC bus. The second terminals of the capacitors C_i for both phases are connected together at a common point.

Other aspects of the topology shown in FIG. 3 are as described in relation to FIG. 2.

FIG. 4 shows another 2-phase DC hub with N modules. The topology is similar to that described in relation to FIG. 3, except that the common 2-phase AC bus further comprises a neutral bus, Bus_—0. In each module i, the second terminals of the two capacitors C_i are connected to Bus_—0. This topology might better share unbalanced conditions between modules.

Other aspects of the topology shown in FIG. 4 are as described in relation to FIG. 3.

FIG. 5 shows a 3-phase DC hub with N modules. The topology is similar to that described in relation to FIGS. 2 and 3. However, in the topology shown in FIG. 5, the DC/AC converter in each module i is formed by six switches S_1—i-S_6—i arranged in three legs, connected to transform DC voltage V_idc into a three phase voltage v_iac. The common 3-phase AC bus comprises three pathways or busses, Bus_A, Bus_B and Bus_C. Thus, there is one bus associated with each phase. Each module i comprises three inductors L_i, three capacitors C_i and three circuit breakers CB_i. For each phase, one inductor L_i, one capacitor C_i and one circuit breaker CB_i are connected together at their first terminals, the second terminal of the inductor L_i is connected to a respective phase of the AC voltage V_iac, and the second terminal of the circuit breaker CB_i is connected to the respective phase of the common 3-phase AC bus. The second terminals of the capacitors C_i in all three phases are connected together at a common point. The alternative delta capacitor connection is also shown.

Other aspects of the topology shown in FIG. 5 are as described in relation to FIG. 2.

FIGS. 2 to 5 show DC hubs with N modules, in accordance with embodiments of the present invention. In each figure, 4 modules are depicted. However, it will be appreciated that, in general, N may be any positive integer greater than one. Moreover, it will be appreciated that other variations of the topologies depicted in FIGS. 2 to 5 are possible, for example, with a different number of phases.

In general, for a DC hub with p phases, where p is any positive integer greater than 0, each module i of the hub will comprise a DC/AC converter having 2p switches arranged in p legs for transforming a DC voltage V_idc into a p phase voltage V_iac, p inductors L_i, p capacitors C_i and p circuit breakers CB_i. The hub will comprise a common p-phase AC bus, with one electrical pathway or bus for each phase p. However, additional busses may be present. In each module, and for each phase, one inductor L_i, one capacitor C_i and one circuit breaker CB_i are connected together at their first terminals, the second terminal of the inductor L_i is connected to a respective phase of the AC voltage V_iac, and the second terminal of the circuit breaker CB_i is connected to the common AC bus associated with that phase. The second terminals of the three capacitors C_i in the module are connected together at a common point. Alternatively, the capacitors may be connected in delta between other phases. Each module i takes the angle of common voltage v_c as the reference zero phase and generates AC voltage v_iac given by:

v _iac=√{square root over (2)}V_iac cos(2πf_st+α_i)  (1)

where V_iac, f_s and α_i are respectively the fundamental root mean square line-neutral magnitude, switching frequency and phase angle of v_iac. The switching frequency f_s is fixed and common for all modules.

An AC voltage can be generated from a given DC voltage in various ways. For example, using unidirectional pulse width modulation (PWM) as illustrated in FIG. 6a, or bidirectional PWM as illustrated in FIG. 6b. Bidirectional PWM is thought to be particularly suitable with bipolar HVDC.

Derivation of the basic circuit equation is described below. Taking the waveform from FIG. 6b as an example, using Fourier series expansion and neglecting all harmonics, the root means square (RMS) line-neutral AC voltage magnitude V_iacm of fundamental component at operating frequency is:

$\begin{array}{cc}{V}_{\mathrm{iacm}}=\frac{4}{\pi \sqrt{2}}V_{\mathrm{idc}}\left(2\sin {\theta }_i-1\right)& \left(2\right)\end{array}$

In a dq frame with the d axis aligned to v the AC voltage vectors of the instantaneous voltage v_iac is expressed as:

$\begin{array}{cc}\stackrel{\_}{V_{\mathrm{iacm}}}=\frac{4}{\pi \sqrt{2}}V_{\mathrm{idc}}\left(2\sin {\theta }_i-1\right)\mathrm{\angle \alpha }i=V_{\mathrm{iacd}}+jV_{\mathrm{iacq}}& \left(3\right)\end{array}$

where V_iac is the phasor, V_iac is the RMS magnitude and α_i is the phase angle of the voltage v_iac. The subscripts d and q denote corresponding phasor components in the dq frame, which are calculated using the following equations:

V _iacd =V _iacm0 M _id

V _iacq =V _iacm0 M _iq

M _i=√{square root over (M_id²+M_iq²)}  (4)

where M_id, M_iq are D-Q components of a generalised control signal. The control signals M_id and M_iq can be readily linked with α_i and θ_i depending on the chosen AC waveform. V_iacm0=4V_idc/(π√{square root over (2)}) is the maximum RMS magnitude of v_iac.

The current equation of the inductor L_i is:

jω _s L _i I _iac = V_iac− V_c  (5)

and the voltage equation of the capacitor is:

$\begin{array}{cc}{\mathrm{j\omega }}_s\stackrel{\_}{V_c}\sum _{i=1}^NC_i=\sum _{i=1}^N\stackrel{\_}{I_{\mathrm{iac}}}& \left(6\right)\end{array}$

where ω_s=2πf_s.

For simplicity, the d-axis of the dq frame is assumed to be aligned with the capacitor voltage vector V_c. Therefore, V_cq=0, and:

V _c =V_cd+jV_cq=V_cd=V_c  (7)

where V_c is the RMS line neutral AC voltage magnitude of the capacitor voltage v_c.

Given the assumption of equation (7) and using equations (3) and (5), the AC current of module i can be expressed as:

$\begin{array}{cc}{I}_{\mathrm{iac}}=\frac{\stackrel{\_}{V_{\mathrm{iac}}}-\stackrel{\_}{V_c}}{{\mathrm{j\omega }}_sL_i}& \\ I_{\mathrm{iacd}}+jI_{\mathrm{iacq}}=\frac{V_{\mathrm{iacd}}+jV_{\mathrm{iacq}}-V_c}{{\mathrm{j\omega }}_sL_i}& \left(8\right)\\ I_{\mathrm{iacd}}+jI_{\mathrm{iacq}}=\frac{V_{\mathrm{iacq}}}{{\omega }_sL_i}-j\frac{V_{\mathrm{iacd}}-V_c}{{\omega }_sL_i}& \left(9\right)\end{array}$

Given the assumption of equation (7), the capacitor voltage in equation (6) becomes:

$\begin{array}{cc}{\mathrm{j\omega }}_sV_c\sum _{i=1}^NC_i=\sum _{i=1}^N\left(I_{\mathrm{iacd}}+jI_{\mathrm{iacq}}\right)& \left(10\right)\end{array}$

FIG. 7 shows the phasor diagram of a 3-module test system (a DC hub with 3 modules) where three AC voltages and currents are shown. It can be seen that the d axis is aligned with V_c. It can also be seen that the current phasors for each module i are in phase with the voltage phasors for that module. This is based on the assumption of zero reactive current at each module.

Design of the LC circuit in each module is described below. The complex power per phase (S_i) at terminal i is:

$\begin{array}{cc}\begin{array}{c}{S}_i=\left(V_{\mathrm{iacd}}+jV_{\mathrm{iacq}}\right)\left(I_{\mathrm{iacd}}+jI_{\mathrm{iacq}}\right)\\ =V_{\mathrm{iacd}}I_{\mathrm{iacd}}+V_{\mathrm{iacq}}I_{\mathrm{iacq}}+j\left(V_{\mathrm{iacq}}I_{\mathrm{iacd}}-V_{\mathrm{iacd}}I_{\mathrm{iacq}}\right)\end{array}& \left(11\right)\end{array}$

The real power per phase P_i is the real part of (11):

P _i =V _iacd I _iacd +V _iacd I _iacd  (12)

Substituting the expression for current from equation (9) in equation (12) gives:

$\begin{array}{cc}\begin{array}{c}{P}_i=V_{\mathrm{iacd}}\frac{V_{\mathrm{iacq}}}{{\omega }_sL_i}-V_{\mathrm{iacq}}\frac{V_{\mathrm{iacd}}-V_c}{{\omega }_sL_i}\\ =\frac{V_{\mathrm{iacq}}V_c}{{\omega }_sL_i}\end{array}& \left(13\right)\end{array}$

From equation (13), a general formula for designing the inductor L_i for each phase can be deduced:

$\begin{array}{cc}{L}_i=\frac{V_{\mathrm{iacq}}V_{\mathrm{cm}}}{P_{\mathrm{im}}}& \left(14\right)\end{array}$

where V_cm is the rated fundamental root mean square line-neutral magnitude of the capacitor voltage, P_im is the maximum power per phase and V_iacq is pre-selected according to other design requirements.

The capacitor for each phase C_i is designed to compensate the reactive current generated by L_i at maximum power condition. From equations (9) and (10):

$\begin{array}{cc}{\omega }_sV_{\mathrm{cm}}C_i=\frac{V_{\mathrm{cm}}-V_{\mathrm{iacd}}}{{\omega }_sL_i}& \left(15\right)\\ C_i=\frac{V_{\mathrm{cm}}-V_{\mathrm{iacd}}}{{\omega }_s^2V_{\mathrm{cm}}L_i}& \left(16\right)\end{array}$

where V_iacd is pre-selected according to other design requirements.

Equations (14) and (16) are general formulas for calculating L_i and C_i, wherein V_iacd and V_iacq are any positive values that also satisfy the inequality V_iacd²+V_iacq²≦V_iacm0².

The hub is preferably designed to achieve zero reactive power at module i, when that module is operating at maximum power. This minimises current magnitude and therefore minimises switching and conduction losses. The condition for zero reactive power at module i implies:

$\begin{array}{cc}\angle \stackrel{\_}{V_{\mathrm{iac}}}=\angle \stackrel{\_}{I_{\mathrm{iac}}}\frac{V_{\mathrm{iacq}}}{V_{\mathrm{iacd}}}=\frac{I_{\mathrm{iacq}}}{I_{\mathrm{iacd}}}& \left(17\right)\end{array}$

Substituting the expression for current from equation (9) in equation (17) gives:

$\begin{array}{cc}\frac{V_{\mathrm{iacq}}}{V_{\mathrm{iacd}}}=\frac{V_{\mathrm{iacd}}-V_c}{-V_{\mathrm{iacq}}}-V_{\mathrm{iacq}}^2=V_{\mathrm{iacd}}^2-V_{\mathrm{iacd}}V_c& \left(18\right)\\ V_{\mathrm{iacm}}^2=V_{\mathrm{iacd}}V_c& \left(19\right)\end{array}$

Substituting equation (4) into equation (19) gives the zero reactive power condition expressed in terms of control signals:

(M_id²+M_iq²)V_iacm0=M_idV_c  (20)

Rearranging equation (3) and squaring gives:

P _i ²ω_s²L_i²=V_iacq²V_c²  (21)

Taking the square of equation (19) and adding to equation (21) gives:

P _i ²ω_s²L_i²+V_iac⁴=V_iacq²V_c²+V_iacd²V_c²

P _i ²ω_s²L_i²+V_iac⁴=V_iac²V_c²  (22)

The inductor L_i is designed to enable maximum (rated) power transfer at the respective module. Substituting V_iac with V_iacm0, P_i with P_im and V_c with V_cm, gives:

P _im ²ω_s²L_i²+V_iacm0⁴=V_iacm0²V_cm²  (23)

where P_im is the maximum power per phase of module i and V_cm is the maximum RMS line neutral AC voltage magnitude of the capacitor voltage v_c. V_cm may be selected as a first design step, since the design below is valid for any V_cm. Equation (23) can be rearranged in the following form:

P _im ²ω_s²L_i²=V_iacm0²(V_cm²−V_iacm0²)  (24)

Equation (24) indicates that V_cm should be larger than any V_iacm0, otherwise the term P_im²ω_s²L_i² will be less than zero, which never holds. Equation (24) allows L_i to be calculated under the condition of zero reactive power, according to:

$\begin{array}{cc}{L}_i=\frac{V_{\mathrm{iacm}0}\sqrt{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}}{P_{\mathrm{im}}{\omega }_s}& \left(25\right)\end{array}$

Using the capacitor voltage equation (6), and assuming a maximum capacitor voltage Vc=Vcm,

$\begin{array}{cc}{\mathrm{j\omega }}_sV_{\mathrm{cm}}\sum _{i=1}^NC_i=\sum _{i=1}^NI_{\mathrm{iacd}}+j\sum _{i=1}^NI_{\mathrm{iacq}}& \left(26\right)\end{array}$

Separating equation (26) into real and imaginary components gives:

$\begin{array}{cc}{\mathrm{j\omega }}_sV_{\mathrm{cm}}\sum _{i=1}^NC_i=j\sum _{i=1}^NI_{\mathrm{iacq}}& \left(27\right)\\ \sum _{i=1}^NI_{\mathrm{iacd}}=0& \left(28\right)\end{array}$

Equation (27) indicates that the total reactive current from all modules (I_iacq) should be balanced by the reactive current generated by the capacitors. One way to achieve this is to let each local C_i balance the reactive current generated by each I_iacq, ie:

jω _s V _cm C _i =jI _iacq  (29)

Substituting the expression for current from equation (9) into equation (29) gives:

$\begin{array}{cc}{\omega }_sV_{\mathrm{cm}}C_i=\frac{V_{\mathrm{cm}}-V_{\mathrm{iacd}}}{{\omega }_sL_i}& \left(30\right)\end{array}$

Multiplying both sides of equation (30) by V_cm and considering the requirement for zero reactive power at maximum power, gives:

$\begin{array}{cc}{\omega }_sV_{\mathrm{cm}}^2C_i=\frac{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}{{\omega }_sL_i}& \left(31\right)\\ C_i=\frac{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}{{\omega }_s^2V_{\mathrm{cm}}^2L_i}& \left(32\right)\end{array}$

Substituting L_i from equation (25) in equation (32) gives an equation for the size of capacitor C_i in terms terminal power P_im, terminal voltage V_iacm0 and V_cm.

$\begin{array}{cc}{C}_i=\frac{1}{{\omega }_sV_{\mathrm{cm}}^2}\frac{P_{\mathrm{im}}\sqrt{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}}{V_{\mathrm{iacm}0}}& \left(33\right)\end{array}$

Equations (25) and (33) allow calculation of values for L_i and C_i that minimize current of each module at maximum power. However, a lower value for L_i will also transmit power P_im, if according to equation (13), V_iacq is reduced. Lower V_iacq implies under-utilisation of converters. However, lower V_iacq may be required to provide some control margin or to account for internal losses. Thus, a value for L_i that is lower than the value given by equation (25) can also be used. Furthermore, according to equation (16), if V_iacd is kept unchanged, C_i needs to be increased if L_i is reduced. Thus, values of L_i and C_i that satisfy the following two inequalities can also be used to transmit power P_i to the hub.

$\begin{array}{cc}{L}_i\le \frac{V_{\mathrm{iacm}0}\sqrt{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}}{P_{\mathrm{im}}{\omega }_s}& \left(34\right)\end{array}$

the value of capacitor Ci is selected according to the formula:

$\begin{array}{cc}{C}_i\ge \frac{V_{\mathrm{cm}}^2-V_{\mathrm{iacm}0}^2}{{\omega }_s^2V_{\mathrm{cm}}^2L_i}& \left(35\right)\end{array}$

By summing equation (33) from i=0 to i=N, the total capacitor of the hub is:

$\begin{array}{cc}C=\sum _{i=1}^NC_i=\frac{-\sum _{i=1}^N\frac{V_{\mathrm{iacm}0}^2}{L_i}}{{\omega }_s^2V_{\mathrm{cm}}^2}+\frac{\sum _{i=1}^N\frac{1}{L_i}}{{\omega }_s^2}& \left(36\right)\end{array}$

As discussed in more detail below, the hub of the present invention is both expandable and flexible in the sense that modules may be connected and disconnected, without disruption to the overall operation of the hub.

AC voltage magnitude V_cm and switching frequency ω_s are fixed. If the hub is designed according to the methods outlined above, it can be seen from equation (25) that L_i only depends on V_iacm0 and P_im. That is to say, it does not depend on parameters associated with any other module than the one to which it belongs. Further, no matter how many other modules are connected or disconnected, the local power flow at a module is solely dependent on V_cm, V_iacm and L_i. Thus, modules can be readily connected to or disconnected from the DC hub without the need to change the inductor L_i or capacitor C_i in any of the other modules.

When a module i is added, the required additional capacitor C_i can be determined according to equation (33). It is only required that the AC voltage magnitude of the new module is lower than the magnitude of the capacitor voltage V_cm, according to equation (24).

If a module is tripped from the DC hub, the required reduction in capacitor will be according to equation (33). Therefore, a circuit breaker CB_i is included, located at the bus side of C_i, which enables module i to be connected or disconnected without affecting the basic operating principles of the remaining terminals.

In designing the DC hub, the maximum power and rated DC voltage of each module are known a priori, operating frequency and capacitor AC voltage are independently selected, and the inductance and capacitance of each terminal need to be determined.

According to equations (25) and (33), the only independent variable in the designing stage is the AC voltage of the capacitor, assuming f_s is initially fixed. After selecting V_cm, inductances and capacitance will be calculated according to equations (25) and (33) using the maximum power condition for each module i. These equations enable zero reactive power at each module under maximum power for that module.

According to equation (25), V_cm should be larger than V_iacm0. Thus, V_cm should be greater than the maximum fundamental root mean square line-neutral voltage magnitude at any terminal. Although, in principle, any V_cm can be selected, a high value for V_cm has cost penalties. Very low V_cm may cause control difficulties. The best overall performance is typically obtained if V_cm is chosen to be around 20% higher than the maximum V_iacm0. After selecting V_cm, each L_i and C_i is calculated according to equations (25) and (33).

Control of the hub is described below. In the previous analysis, hub design is based on rated power levels. At each module i, active and reactive power can be controlled independently using the two control signals M_id and M_iq. According to equation (13), M_iq (V_iacq) is used to control P_i. Similarly, according to equation (11), M_id is used to control reactive power.

In an N terminal hub according to an embodiment of the present invention, one module may be used to maintain power balance within the hub by regulating capacitor voltage v_c, whilst the other modules control their local power. The module in which v_c is controlled may be referred to as the “Voltage Module”, whilst the other modules may be referred to as “Power Modules”.

In normal operation, each module can independently demand local power. However, the sum of all power over all modules connected to the hub must be zero, as the power input of the system must be equal to the power output. Accordingly, a central master controller is provided for moderating demanding powers and calculating power orders which are assigned to controllers at each of the modules. For example, if a module k operates as the Voltage Module, and the other modules operate as Power Modules, then the master controller ensures that the sum of all power orders for the Power Modules stays within the rating of the Voltage Module. That is to say:

$\begin{array}{cc}\sum _{i=1,i\ne k}^NP_i^{\mathrm{ref}}\le P_{\mathrm{km}}^{\mathrm{ref}}& \left(37\right)\end{array}$

where injecting power into the common AC bus is defined as the positive direction.

From the capacitor voltage equations (6):

$\begin{array}{cc}\begin{array}{c}\mathrm{j\omega }\left(V_{\mathrm{cd}}+{\mathrm{jV}}_{\mathrm{cq}}\right)C=\sum _{i=1}^N\left(I_{\mathrm{iacd}}+jI_{\mathrm{iacq}}\right)\\ =\sum _{i=1}^N\frac{\left(V_{\mathrm{iacd}}+jV_{\mathrm{iacq}}\right)-\left(V_{\mathrm{cd}}+jV_{\mathrm{cq}}\right)}{{\mathrm{j\omega }}_sL_i}\end{array}& \left(38\right)\\ -{\omega }_sV_{\mathrm{cq}}C=\sum _{i=1}^N\frac{V_{\mathrm{iacq}}-V_{\mathrm{cq}}}{{\omega }_sL_i}-{\omega }_sV_{\mathrm{cd}}C=\sum _{i=1}^N\frac{V_{\mathrm{iacd}}-V_{\mathrm{cd}}}{{\omega }_sL_i}& \left(39\right)\end{array}$

Equation (39) may be arranged in the following form:

$\begin{array}{cc}\left(\sum _{i=1}^N\frac{1}{{\omega }_sL_i}-{\omega }_sC\right)V_{\mathrm{cq}}=\sum _{i=1}^N\frac{V_{\mathrm{iacq}}}{{\omega }_sL_i}\left(\sum _{i=1}^N\frac{1}{{\omega }_sL_i}-{\omega }_sC\right)V_{\mathrm{cd}}=\sum _{i=1}^N\frac{V_{\mathrm{iacd}}}{{\omega }_sL_i}& \left(40\right)\end{array}$

A variable K_c may be defined as:

$\begin{array}{cc}{K}_c=\sum _{i=1}^N\frac{1}{{\omega }_sL_i}-{\omega }_sC& \left(41\right)\end{array}$

Substituting equation (36) into equation (41) gives:

$\begin{array}{cc}\frac{K_c}{{\omega }_s}=\sum _{i=1}^N\frac{1}{{\omega }_s^2L_i}-C=\frac{\sum _{i=1}^N\frac{V_{\mathrm{iacm}0}^2}{L_i}}{{\omega }_s^2V_{\mathrm{cm}}^2}& \left(42\right)\end{array}$

Equation (42) indicates that K_c>0. If a module k is selected to maintain V_c, equations (40) can be rewritten as follows:

$\begin{array}{cc}{K}_cV_{\mathrm{cq}}=\sum _{i=1,i\ne k}^N\frac{V_{\mathrm{iacq}}}{{\omega }_sL_i}+\frac{V_{\mathrm{kacq}}}{{\omega }_sL_k}=\sum _{i=1,i\ne k}^N\frac{V_{\mathrm{iacq}}}{{\omega }_sL_i}+\frac{V_{\mathrm{kacm}}}{{\omega }_sL_k}& \left(43\right)\\ K_cV_{\mathrm{cd}}=\sum _{i=1,i\ne k}^N\frac{V_{\mathrm{iacd}}}{{\omega }_sL_i}+\frac{V_{\mathrm{kacd}}}{{\omega }_sL_k}=\sum _{i=1,i\ne k}^N\frac{V_{\mathrm{iacd}}}{{\omega }_sL_i}+\frac{V_{\mathrm{kacm}}}{{\omega }_sL_k}& \left(44\right)\end{array}$

M_kq and M_kd are used to maintain V_cq and V_cd respectively.

FIG. 8 shows a control schematic for the Voltage Module k. The reference angle for the firing logic of the voltage terminal comes from a voltage controlled oscillator (VCO). Thus θ_s=2πf_st, where f_s is the switching frequency of the hub.

Equation (13) may be rewritten:

$\begin{array}{cc}{P}_i=\frac{V_{\mathrm{iacm}0}M_{\mathrm{iq}}V_c}{{\omega }_sL_i}& \left(45\right)\end{array}$

Thus, P_i can be controlled by manipulating the q-axis modulation index M_iq. Also, from equation (45), it can be seen that V_cm must be at the rated value at all times in order to enable rated power transfer at each terminal. At rated power, P_i—rate, the q-axis control index is:

$\begin{array}{cc}{M}_{\mathrm{iq\_rate}}=\frac{P_{\mathrm{i\_rate}}{\omega }_sL_i}{V_{\mathrm{iacm}0}V_{\mathrm{cm}}}& \left(46\right)\end{array}$

According to the zero reactive power condition of equation (20):

V _iacd ² +V _iacq ² =V _iacd V _c  (47)

Referring to the active power equation (13), V_iacq should be reduced to reduce the active power P_i. To maintain zero reactive power at module i, equation (47) indicates that V_iacd should also be reduced, following the reduction of V_iacq.

However, referring to the capacitor voltage equation (44), if V_iacd for each of the Power Modules is reduced, M_kd of the Voltage Module would need to be increased. This may not be allowed, as it may exceed its limit, with the result that V_cd would be lower than its rated value. If V_cd is decreased, the capability of the Power Modules to transmit active power is reduced. It is possible for some modules to transmit partial power while other terminals transmit rated power. If V_cd is lowered, the requirement for the hub to transmit active power is not satisfied.

To ensure that V_c is maintained at its rated value V_cm, all V_iacd are maintained at their rated values. Thus:

M _id =M _{id — rate}=√{square root over (1−M_iq—rate²)}  (48)

FIG. 9 shows a control schematic for the Power Modules. Each Power Module is used to control its transmitted active power to the power reference P_i,ref^pu. The power reference could be manually set or come from an upper layer supervising system. M_id is maintained at its rated value according to equation (48). The reference angle for firing logic θ comes from a voltage controlled oscillator (VCO). Thus, θ_s=2πf_st, where f_s is the switching frequency of the hub.

In FIG. 9, the power control loop also employs droop control feedback (K_droop). The droop control is required to ensure that:

1) The hub maintains some voltage control, and thus prevents collapse if the Voltage Module is lost;

2) V_c regulation is shared among all modules in order to avoid saturation of the Voltage Module;

3) The hub voltage is maintained close to rated values in case the master hub control is lost.

The values for the droop gains are calculated to ensure that capacitor voltage is within the required range in the worst case operating conditions, such as tripping of the Voltage Module, or unfavourable power demand at every Power Module.

Fault current analysis of a hub which embodies the present invention is discussed below. The capacitor voltage during normal operation is, from (5) and (6):

$\begin{array}{cc}\stackrel{\_}{V_c}=V_{\mathrm{cd}}+jV_{\mathrm{cq}}=\frac{\sum _{i=1}^N\frac{\stackrel{\_}{V_{\mathrm{iac}}}}{L_i}}{D}=\frac{\sum _{i=1}^N\frac{M_{\mathrm{id}}V_{\mathrm{iacm}0}+jM_{\mathrm{iq}}V_{\mathrm{iacm}0}}{L_i}}{D}\mathrm{where}& \left(49\right)\\ D=\sum _{i=1}^N\frac{1}{L_i}-{\omega }_s^2C& \left(50\right)\end{array}$

Current at module i is:

$\begin{array}{cc}{I}_{\mathrm{iacd}}+jI_{\mathrm{iacq}}=\frac{V_{\mathrm{cq}}-V_{\mathrm{iacq}}}{{\omega }_sL_i}+j\frac{V_{\mathrm{cd}}-V_{\mathrm{iacd}}}{{\omega }_sL_i}=\frac{V_{\mathrm{cq}}-M_{\mathrm{iq}}V_{\mathrm{iacm}0}}{{\omega }_sL_i}+j\frac{V_{\mathrm{cd}}-M_{\mathrm{id}}V_{\mathrm{iacm}0}}{{\omega }_sL_i}& \left(51\right)\end{array}$

If a DC fault happens at module i, the voltage becomes V_iacm0=0. Substituting this condition in equation (50) gives the expression of V_c when a module i is at fault. Substituting V_iacm0=0 and equation (50) in equation (51) gives the expression for current at the faulted terminals and at other terminals for a fault at module i.

Although direct analytical formula for magnitude of fault currents cannot be derived, extensive simulations have been performed using detailed simulation models. In all tested cases, with realistic parameters, the fault current is around 1.1-1.3 pu of the rated current. This is a significant conclusion because semiconductors can withstand this small overcurrent, and there is no need to trip terminals or the hub for DC faults.

Detailed simulation of a 7-module DC hub in PSCAD/EMTDC is used to validate the design and control of the hub. DC voltage rating and power rating of the hub is shown in Table 1. Modules 1-6 are Power Modules and module 7 is the Voltage Module.

The capacitor voltage V_cm is 755.9 kV (V_cm/V_7acm0=1.2, V_7acmo is the maximum value among each V_{i acm0}) and the switching frequency is 1250 Hz.

In table 1, power injected into the hub is defined as the positive power direction (1).

TABLE 1
Termi-
nal	1	2	3	4	5	6	7
VDC	100	200	300	400	500	600	700
(kV)
P	200	300	200	400	350	500	600
(MW)
Li (H)	0.086	0.112	0.243	0.152	0.199	0.145	0.103
Ci (μF)	0.186	0.136	0.058	0.082	0.053	0.055	0.048
Power	1	−1	1	1	−1	1	−1
Direc-
tion

FIGS. 10( a)-10(l) show simulation results for the 7-module hub. Step changes of the power orders of modules 1-6 are applied sequentially at t=4 s with an interval of 2 s.

FIGS. 10( a)-10(g) show that each Power Module can transmit rated power. At each module, the power follows the power order correctly and power levels at other modules are unaffected. In FIGS. 10(a)-10(g), the active power references P_ref1-P_ref7 and the measured active power P_1pu-P_7pu are per unit values taking the rated power of each module in Table 1 as the base value.

FIGS. 10( h)-10(k) show the curves of reactive power. From 2 s to 4 s, all the modules operate at full power, and it can be seen that zero reactive power is achieved at this full power condition. Similar to FIGS. 10(a)-10(g), each of Q_1pu-Q_7pu are per-unit values taking the rated power of each module in Table 1 as the base value.

FIG. 10( l) shows the capacitor voltages. The magnitude (V_cpu) and d component of capacitor voltage (V_cdpu) is maintained at 1 pu whilst the q component of capacitor voltage (V_cqpu) is maintained at zero in steady state at all operating points. Each of the V_cpu, V_cdpu and V_cqpu are per-unit values taking V_cm as the base value.

FIGS. 11( a)-11(g) show the effect of isolating and connecting a DC transmission line (with an associated module) to the hub. Module 1 is tripped from the hub at 3.0 s and reconnected to the hub at 4.0 s. Before module 1 is tripped, each Power Module operates at its rated power and the Voltage Module maintains v_c. When module 1 is tripped, P_1pu reduces to 0.0 as expected. Active power of the Voltage Module (module 7) automatically changes to balance the active power of the hub. It is observed that active power of the other power modules (modules 2-6) remains unchanged regardless of the connection status of module 1. V_c is also unchanged in steady state. The reactive power of the other power modules (modules 2-6) remains zero, regardless of the tripping or connection of module 1.

FIGS. 12( a)-12(h) show the performance of droop control. Before 2.0 s, all the Power Modules (modules 1-6) operate at full power. Power direction of module 5 is the same as the Voltage Module (module 7). At 2.0 s, the sign of power reference at module 5 (P_ref5) is reversed (i.e. module 5 changes from absorbing to injecting). As a result, there is an imbalance between the power orders. The droop control decreases the magnitude of power order at the modules injecting power into the hub, and increases the magnitude of power orders of the modules absorbing power from the hub.

FIG. 12( a) shows that the capacitor voltage of the hub V_cq is non-zero as expected. The magnitude of v_c is maintained at almost 1 pu.

FIGS. 12( b)-12(h) show the active power of all the modules. It can be seen that the injected power of the other modules is reduced.

FIGS. 13( a)-13(c) show the ratios of steady-state current magnitudes over respective rated current for DC faults at each module, considering a range of V_cm/V_7acm0 (selected V_cm/V_7acm0=1.2 for the 7-module DC hub shown in Table 1). It can be seen that fault current stays within 2 times the current under normal conditions, and that it is typically below 1.3 pu.

FIGS. 14( a)-14(h) show the system response to a DC fault. At 2.0 s, a permanent DC fault happens at module 6. At 2.05 s, module 6 is isolated by its circuit breaker CB₆. FIGS. 14(a)-14(g) show the AC current of each module. It can be seen that AC current for all the modules during the DC fault stays within 1.3 pu of its rated value. In FIGS. 14(a)-14(g), each current is a per-unit value taking the respective rated current of each module as the base value. FIG. 14(h) shows that no over-voltage will occur during faults. FIG. 14(h) also shows that, after the fault is cleared, the DC hub operates normally with 6 modules, and capacitor voltage is well controlled.

The circuit breakers CB_i employed by the present invention may be commonly available mechanical AC circuit breakers such as vacuum breakers, gas insulated breakers or any other type used with AC systems. The switches employed in the DC/AC converters may be any suitable switches, such as Insulated Gate Bipolar Transistors (IGBT).

Delta connection and star connection of capacitors C_i are specifically described herein. However, it will be appreciated by those skilled in the art that capacitors C_i may be connected together in any suitable way to give the required reactive power in balanced and symmetrical manner. All such connections can, however, be reduced to equivalent star connection giving the same reactive power.

REFERENCES

• [1] D. Van Hertem and M. Ghandhari, “Multi-terminal VSC HVDC for the European supergrid: Obstacles,” Renewable and Sustainable Energy Reviews, vol. 14, no. 9, pp. 3156-3163, 2010.
• [2] JURGEN HÄFNER, BJÖRN JACOBSON “Proactive Hybrid HVDC Breakers—A key innovation for reliable HVDC grids”, CIGRE symposium, Bologna, September 2011.
• [3] D Jovcic and B. T Ooi, “Developing DC transmission network using DC transformers” IEEE Transactions on Power Delivery, Vol. 25, issue 4, October 2010, pp 2535-2543.
• [4] S. Loddick “Active Stator, a new generator Topology for direct drive permanent magnet generators” 9th IET International Conference on AC and DC Power Transmission, ACDC 2010; London October 2010.
• [5] J Robinson, D Jovcic and G Joos, “Analysis and Design of an Offshore wind farm using MV DC grid” IEEE Transactions on Power Delivery, Vol. 25, issue 4, October 2010, pp 2164-2173.
• [6] D. Jovcic, International patent application no PCT/GB2012/051486

Claims

1. (canceled)

2. A p-phase electronic hub for transferring power between N DC systems of DC voltage Vidc (i=1, 2, . . . N), respectively, the hub comprising:— N modules, each for connection to a respective DC system of voltage Vidc, and for exchanging power Pi with the respective DC system; anda common p-phase AC bus for connecting the N modules;

3. A p-phase electronic hub as claimed in claim 2, wherein the voltage of capacitor(s) Ci is regulated at a value Vcm, where Vcm>Viacm0 and where Viacm0 is the maximum value of Viacm for all i; wherein the value of each inductor Li is selected according to the formula:—

4. A p-phase electronic hub as claimed in claim 2 wherein the voltage of capacitor(s) Ci is regulated at a value Vcm, where Vcm>Viacm0 and where Viacm0 is the maximum value of Viacm for all i; wherein the value of each inductor Li is selected according to the formula:—

5. A p-phase electronic hub as claimed in claim 2 wherein the power Pim exchanged by each module with the respective DC system is controllable by varying Viacm and αi of the respective AC voltage Viac generated by the DC/AC converter of the respective module.

6. A p-phase electronic hub as claimed in claim 2 wherein at least one of the N modules is configured for regulating the capacitor voltage Vc at value Vcm, whilst the other(s) of said N modules is/are configured for regulating the power exchanged by that module with the respective DC system.

7. A p-phase electronic hub for transferring power between N DC systems of DC voltage Vidc (i=1, 2, . . . N), respectively, the hub comprising:— N modules, each for connection to a respective DC system of voltage Vidc and for exchanging power Pi with the respective DC system; anda common p-phase AC bus for connecting the N modules;

8. A p-phase electronic hub as claimed in claim 7 wherein the voltage of capacitor(s) Ci is regulated at a value Vcm, where Vcm>Viacm0 and where Viacm0 is the maximum value of Viacm for all i; wherein the value of each inductor Li is selected according to the formula:—

9. A p-phase electronic hub as claimed in claim 7 wherein the voltage of capacitor(s) Ci is regulated at a value Vcm, where Vcm>Viacm0 and where Viacm0 is the maximum value of Viacm for all i; wherein the value of each inductor Li is selected according to the formula:—

10. A p-phase electronic hub as claimed in claim 7 wherein the power Pim exchanged by each module with the respective DC system is controllable by varying Viacm and αi of the respective AC voltage Viac generated by the DC/AC converter of the respective module.

11. A p-phase electronic hub as claimed in claim 7 wherein at least one of the N modules is configured for regulating the capacitor voltage Vc at value Vcm, whilst the other(s) of said N modules is/are configured for regulating the power exchanged by that module with the respective DC system.