WO2016074199A1

WO2016074199A1 - Dc grid protection method and system thereof

Info

Publication number: WO2016074199A1
Application number: PCT/CN2014/091027
Authority: WO
Inventors: Kai Liu; Youyi Li; Jianping Wang
Original assignee: Abb Technology Ltd
Priority date: 2014-11-13
Filing date: 2014-11-13
Publication date: 2016-05-19
Also published as: EP3218978A4; CN106463950B; EP3218978A1; CN106463950A

Abstract

A DC grid protection method and system are provided. The method comprises the following steps: fault detecting step (501): acquiring fault component travelling wave current value △i and fault component travelling wave voltage value △u of a DC line, wherein forward direction is defined extending from the protection to the DC line; fault determining step (502): if △i and △u at fault occurring time meet following criterions, performing fault detection determination step, the criterions include: product of △i and △u being less than a forward value which is bigger than zero and is associated with a predetermined threshold; and absolute value of △i or △u meeting a sensitivity requirement which is determined by the predetermined threshold; fault direction determination step (503): determining the fault as forward fault. Said scheme can overcome the problem in the prior art and implement the pilot protection in the DC system, especially in the HVDC system.

Description

DC GRID PROTECTION METHOD AND SYSTEM THEREOF

TECHNICAL FIELD

The present application relates to a DC grid protection method and a system thereof.

BACKGROUND ART

In existing HVDC systems， generally， protection based on travelling wave front of local measurements is used as main protection and classical current differential protection is used as backup protection. But the drawback of them is that， the main protection has bad sensitivity for high resistance fault and may mal-operate in LCC DC grid， and the backup protection has very slow operation speed.

In existing two-terminal HVDC system， the main protection for transmission line is mainly based on change rate and amplitude of directional travelling wave front. Such protection has an obvious advantage that it only uses local measurements and has very fast operation speed for metal fault.

But one of the disadvantages of such protections is that it has very low (bad) sensitivity for high resistance fault. Normally，＞200 Ohm fault resistance may lead to failure of operation because the amplitude of the wave front rely heavily on fault resistance. As a result， the high resistance fault has to be cleared by its backup current differential protection with very slow operation speed (e.g. ＞0.5s) . Obviously， it is not reasonable.

And furthermore， this protection is based on the physical feature of smoothing reactor in HVDC system which can slow down the change of current. In some kind of DC grid systems (e.g. some kind of series MTDC system) ， the travelling wave by external fault will not flow through the smoothing reactor， the HVDC protection mentioned above based on travelling wave will fail to operate or mal-trip. For the worst conditions， if the external DC fault occurs on the line with higher voltage level， its travelling wave front is even larger than that by internal fault. It will bring big trouble to the existing HVDC protection based on travelling wave.

Fig. 1 is a chart showing travelling wave fronts of internal and external DC fault in LCC DC grid.

As shown in the Fig. 1， the change rate of the wave fronts from internal fault and external fault are definitely the same at the beginning. And at the same time， the travelling wave front of external fault is even much larger than that of internal fault， because the external line’s voltage level is higher.

In the traditional travelling wave protection device， as shown in Fig. 2， three different measurements will start to determinate if the wave has sufficient amplitude for a specified time. The first measurement calculates the wave difference between just before the wave front and after 10 samples (0.2 ms) . The second and third calculate the wave difference between just before the wave front and after 25 and 35 samples (0.5 and 0.7 ms) . If all three measurements are greater than the threshold， a line fault is detected.

Considering of the wave front of external fault in Fig. 1 is even larger than that of internal fault and the change rate of both internal &external faults has the same rate， the existing HVDC main protection will mal-trip in LCC DC grid in such cases. In other words， existing HVDC travelling wave protection cannot be used directly in LCC DC grid.

In existing HVDC system， normally， the backup protection for transmission line is line current differential protection. Classical current differential algorithm is used in such protections. It operates when the main protections (travelling wave protection) cannot work (e.g. high resistance fault) .

A typical criterion of current differential protection is shown below，

|I_DL-I_{DL_FOS}|＞max (120A， 0.1×|I_DL+I_{DL_FOS}|/2)

In which I_DL is the current of local side， and I_{DL_FOS} is the current of remote side.

Another typical criterion of current differential protection is shown below，

||I_DL|-|I_{DL_FOS}||＞90A

Generally， the sensitivity of the current differential protection is quite good if with proper setting. But its operation speed is too slow. Its operate time is normally several hundreds of millisecond or even several seconds. The main reason is that the fault transient and charging current will influence this protection algorithm greatly. Thus， a long delay is necessary to ensure reliability.

Both main protection and backup may be influenced by high impedance faults.

1) Influence to travelling wave protection

Existing traveling wave criterion is：

|W_COMM|＝|Z_COMI_COM-U_COM|＞350kV

|W_POLE|＝|Z_DIFI_DIF-U_DIF|＞210kV

wherein Z_COM is common mode wave impedance， Z_DIF is differential mode wave impedance， W_POLE is pole wave， and W_COMM is ground wave.

I_COM is common mode current， U_COM is common mode voltage， I_DIF is differential mode current， U_DIF is differential mode voltage.

The protection detects the wave head by using the change rate of ground wave.

|dW_COMM/dt|＞396kV/ms

The DC voltage decreases with a small rate of change when the line grounds to the earth via large impedance， leading to the mis-operations of existing protection based on travelling wave.

If travelling wave protection mis-operates， control &protection system will delay to eliminate faults.

2) Influence to voltage change rate and low voltage protection

Voltage change rate criterion is：

D_UT＝dU _dl/dt ＜-396kV /ms &U_dl ＜200kV ， where U_dl is line voltage and D_UT is the corresponding changing rate.

Voltage change rate protection will mis-operate caused by small DC voltage decreasing， when the line grounds to the earth via large impedance.

3) Influence to current differential protection

The typical criterion of current differential protection is shown below，

|I_DL-I_{DL_FOS}|＞max(I_SET，k×|I_DL+I_{DL_FOS}|/2)

wherein I_SET is fixed current setting which is normally set to 120A， k is ratio coefficient which is normally set to 0.1.

In order to ensure operation on condition of large impedance fault， setting I_SET and k are normally set to a small value. Thus delay time has to be set long enough to avoid mis-operation caused by capacity charging current.

If fast protection (travelling wave protection) mis-operates， backup protection will delay to work. And delay time is too long to guarantee steady operation of power system.

Pilot protection can be a good candidate

Pilot protection based on direction comparison may be one of good candidates for HVDC/DC Grid protection which has many unique advantages such like：

1. Fast operation. Although the pilot protection needs communication between terminals， it has much faster speed than current differential protection， because it is based on the initial travelling wave front and will not be influenced by the HVDC control or transient charging current. Generally， its total operation speed may be ＜15 ms including communication delay and the time spend on algorithm may be ＜1 ms.

2. Low requirement for communication bandwidth. The pilot protection ONLY transfers directional information between terminals. It doesn’t need the measurements from remote terminals. The bandwidth requirement is much lower than that of current differential protection.

3. Low requirement for symmetry of the communication channel. The symmetrical of communication channel is very important for current differential protection. And asymmetrical channel even may lead to mai-trip. But for pilot protection， channel symmetry is not a problem， because it doesn’t need strict data symmetry.

4. Can be used for different types of DC line system. The pilot protection can be used for both two-terminal HVDC system and multi-terminal DC grid system including the LCC DC grid， and it is independent of smoothing reactor or DC filter.

Working principle of pilot direction protection

For the line pilot direction protection， the direction of the fault is determined by the initial travelling wave by both ends， after exchanging the direction information by optical fiber， the pilot direction protection will make the decision to operate or not. The pilot direction protection working principle is shown in Fig. 3a， Fig. 3b， Fig. 3c and Fig . 4.

Fig. 3a shows the situation when internal fault occurs， both Relay A and Relay B will determine the forward fault， the Relay A will send permission signal to Relay B， and Relay B will send permission signal to Relay A. The pilot direction protection will operate when local relay determines the forward fault and receives the permit signal from the remote side， as shown in Fig. 4.

Fig. 3b shows the situation when an external fault occurs at the backside of Relay A， under this condition， Relay A will determine inverse fault， and Relay B determine forward fault， both the relays will not operate. Fig. 3c shows the situation when an external fault occurs at the backside of Relay B， under this condition， Relay A will determine forward fault， and Relay B determine inverse fault， both the relays will not operate either.

Problem of pilot protection exists：

Equations

1 and 2 show the existing direction element based on travelling wave：

Forward fault： Δu′·Δi′＜0 (1)

Reverse fault： Δu′·Δi′＞0 (2)

Here， Δu′andΔi′is the voltage and current of initial travelling wave. The algorithm above is simple and clear. It can work in DC system in theory generally. When equation 1 is satisfied， forward direction fault is determined， and when equation 2 is satisfied， inverse direction fault is determined.

BUT， unfortunately， the algorithm above has a very obvious disadvantage that it will FAIL to operate for total-reflection boundary. For total-reflection boundary， the voltage may be zero (Δu′＝0) or current may be zero (Δi′＝0) . Therefore， Δu′·Δi′≈0. In other words， the polarity of voltage wave or polarity of current wave cannot be detected reliably. Thereby， Δu·Δi＝0. According to the algorithm above， the protection cannot detect the fault direction. And as a result， it will fail to operate even for internal fault in theory.

And it should be noted that in DC system， the condition which is close to total-reflection boundary is possible. For example， if there is large reactor or capacitor installed in the terminal， for the high frequency initial travelling wave front， the condition which is close to total-reflection will occur at the terminal boundary at the beginning.

Another important disadvantage of the algorithm above is that the mentioned algorithm is sensitive for measurement error. If the Δu′or Δi′is close to zero in some cases. Considering of the measurement error， Δu′or Δi′may be positive or negative randomly. The fault direction based onΔu′^＊Δi′may be wrong. And the protection maybe mal-operate or fail to operate.

So， although the algorithm mentioned above is generally correct in theory， it is not so reliable and practical in fact.

SUMMARY

Accordingly， one aspect of the present invention provides a DC grid protection method， comprising the following steps：

fault detecting step： acquiring fault component travelling wave current value and fault component travelling wave voltage value of a DC line， the fault component travelling wave current value at fault occurring time is Δi and the fault component travelling wave voltage value at fault occurring time is Δu， wherein forward direction is defined extending from the protection to the DC line；

fault determining step： if the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time meet following criterions， performing fault direction determination step， the criterions include：

product of the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time being less than a forward value which is bigger than zero and is associated with a predetermined threshold； and

absolute value of the fault component travelling wave current value Δi at fault occurring time or absolute value of the fault component travelling wave voltage value Δu at fault occurring time meeting a sensitivity requirement which is determined by the predetermined threshold；

fault direction determination step： determining the fault as forward fault.

Preferably， the forward value is

wherein i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.

Preferably， the forward value is

wherein k is a constant， 0＜k＜1， i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.

Preferably， the sensitivity requirement includes： |Δi| ＞i_set or |Δu| ＞Z_C ·i_set ， wherein i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.

Preferably， the fault determining step includes：

if the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time meet the following criterions， performing fault direction determination step， the criterion includes：

Preferably， further comprising fault eliminating step：

activating protect operation if forward fault is determined.

Conveniently， the predetermined threshold i_set is bigger than current noise of the line.

Preferably：

the protection includes local terminal protection and remote terminal protection， forward direction of the fault component travelling wave current value acquired by the local terminal protection is defined extending from the local terminal protection to the DC line， forward direction of the fault component travelling wave current value acquired by the remote terminal protection is defined extending from the remote terminal protection to the DC line；

fault direction determination step further includes：：

the local terminal protection determines the fault as forward fault， sends forward fault detected in local message to the remote terminal protection， and determines internal fault when receives forward fault detected in remote message from the remote terminal protection；

the remote terminal protection determines the fault as forward fault， sends forward fault detected in remote message to the local terminal protection， and determines internal fault when receives forward fault detected in local message from the local terminal protection.

Preferably， further comprising fault eliminating step：

activating protect operation in the local terminal protection if internal fault is determined；

activating protect operation in the remote terminal protection if internal fault is determined.

Another aspect of the present invention provides a computer program comprising computer program code adapted to perform all of the steps of any one of the above when run on a computer.

A further aspect of the present invention provides a computer program according to the above， embodied on a computer-readable medium.

Another aspect of the present invention provides a DC grid protection system， comprising the following modules：

fault detecting module： acquiring fault component travelling wave current value and fault component travelling wave voltage value of a DC line， the fault component travelling wave current value at fault occurring time is Δi and the fault component travelling wave voltage value at fault occurring time is Δu， wherein forward direction is defined extending from the protection to the DC line；

fault determining module： if the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time meet following criterions， performing fault direction determination module， the criterions include：

fault direction determination module： determining the fault as forward fault.

Preferably， the forward value is

Preferably， the sensitivity requirement includes： |Δi| ＞i_set or |Δu| ＞Z_C·i_set ， wherein i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.

Preferably， the fault determining module includes：

if the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time meet the following criterions， performing fault direction determination module， the criterion includes：

Preferably， further comprising：

fault eliminating module： activating protect operation if forward fault is determined.

Preferably：

fault direction determination module further includes：：

Preferably， further comprising， fault eliminating module：

The present invention determines the direction of the fault component travelling wave current value by the condition one： product of the fault component travelling wave current value Δi and the fault component travelling wave voltage value Δu less than a forward value which is bigger than zero and is associating with a predetermined threshold； and activates the protect operation when meet the condition two： absolute value of the fault component travelling wave current value Δi or absolute value of the fault component travelling wave voltage value Δu meets the sensitivity requirement which is determined by the predetermined threshold. Because the fault component travelling wave voltage value is superimposed in the total-reflection situation， so that the present invention can overcome the problem in the prior art and implements the pilot protection in the DC system， especially in the HVDC system.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 shows a chart showing travelling wave fronts of internal and external DC fault in LCC DC grid；

Fig. 2 shows a measurement schematic view of the traditional travelling wave protection device；

Fig. 3a illustrates schematically the situation when an internal fault occurs；

Fig. 3b illustrates schematically the situation when an external fault occurs at the backside of Relay A；

Fig. 3c illustrates schematically the situation when an external fault occurs at the backside of Relay B；

Fig. 4 illustrates schematically the pilot protection logic in accordance with the prior art；

Fig. 5 shows a flow-process diagram illustrating a DC grid protection method in accordance with the present invention；

Fig. 6 illustrates schematically the traveling wave of a line；

Fig. 7 illustrates schematically the situation when an inverse fault occurs；

Fig. 8 illustrates schematically the situation when a forward fault occurs；

Fig. 9 illustrates schematically the total-reflection situation；

Fig. 10 shows a structural module drawing of a DC grid protection system；

Fig. 11 illustrates schematically travelling wave based protection in accordance with the prior art；

Fig. 12 illustrates schematically the distributed line capacitance of the line when external fault occurs；

Fig. 13 illustrates schematically the distributed line capacitance of the line when internal fault occurs；

Fig. 14 shows the simulation model；

Fig. 15a shows the simulation result of the fault component travelling wave voltage value， the fault component travelling wave current value and the product of both in local terminal when the internal fault occurs；

Fig. 15b shows the simulation result of the fault component travelling wave voltage value， the fault component travelling wave current value and the product of both in remote terminal when the internal fault occurs；

Fig. 16a shows the simulation result of the fault component travelling wave voltage value， the fault component travelling wave current value and the product of both in local terminal when the external fault occurs；

Fig. 16b shows the simulation result of the fault component travelling wave voltage value， the fault component travelling wave current value and the product of both in remote terminal when the external fault occurs.

DETAILED DESCRIPTION OF THE EMBODIMENTS

Hereinafter， the present invention is further introduced in detail by the particular embodiments in combination with the figures.

Fig. 5 shows a flow-process diagram illustrating a DC grid protection method in accordance with the present invention， including the following steps：

step 501： acquiring fault component travelling wave current value and fault component travelling wave voltage value of a DC line， the fault component travelling wave current value at fault occurring time is Δi and the fault component travelling wave voltage value at fault occurring time is Δu， wherein forward direction is defined extending from the protection to the DC line；

step 502： if the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time meet following criterions， performing fault direction determination step， the criterions include：

step 503： determining the fault as forward fault.

The protection detects the fault by any method of the prior art in step 501. But the protection does not determine whether the fault is forward fault or internal fault in step 501. The grid may be divided into no-fault network and fault-component network when detected the fault， so the fault component travelling wave current value and the fault component travelling wave voltage value are the travelling wave current/voltage values in the fault-component network. The method of dividing the grid into no-fault network and fault-component network can implement with any method of the art.

Fig. 6 illustrates schematically the traveling wave of a line. The voltage and current value of any location in the line can be expressed by equation (3) according to the principle of electrotechnics：

wherein， ^xis the distance between point A and any given point along the line， x_A＝0； ^trepresents the time instant；

is the forward travelling wave voltage， where subscript 1 represents forward traveling wave， v is the speed of the traveling wave；

is the inverse travelling wave voltage， where subscript 2 represents inverse traveling wave；

is the forward travelling wave current；

is the inverse travelling wave current；

Z_C is wave impedance.

At the start point of the line， as shown by point A which represents the location of the protection in Fig. 6， x_A＝0.

Substitute x＝0 in the equation (3) ， we can obtain：

Wherein， the u (t) is the travelling wave voltage value at time t of point A and i (t) is the travelling wave current value at time t of point A. Using the same method in fault-component network， we can get the fault component travelling wave current value Δi (t) and the fault component travelling wave voltage value Δu (t) at time t and point A which is the location of the protection. Δu and Δi represent the fault component travelling wave current value and the fault component travelling wave voltage value when the fault detected at fault occurring time t.

In step 502， the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time should meet the following criterions：

criterion one： product of the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time less than a forward value which is bigger than zero and is associating with a predetermined threshold； and

criterion two： absolute value of the fault component travelling wave current value Δi at fault occurring time or absolute value of the fault component travelling wave voltage value Δu at fault occurring time meets the sensitivity requirement which is determined by the predetermined threshold.

The criterion one is used to determine whether the fault component travelling wave current value Δi at fault occurring time is forward or inverse. As flowing from the equipment to the protection is the forward direction of the fault component travelling wave current value Δi at fault occurring time， so Δi＞0 when the current flows from the equipment to the protection and Δi＜0 when the current flows from the protection to the equipment. The criterion one includes situation that the product is between 0 to the forward value. Because the forward value is associated with a predetermined threshold which determines the sensitivity requirement in criterion two， so the situation that the product is between 0 to the forward value can be excluded by setting a proper sensitivity requirement.

The criterion two is used to determine the sensitivity requirement. When the Δi or Δu meet the sensitivity requirement， it can trigger the step 503. Consider the total-reflection situation illustrated schematically in Fig. 9. When reflection occurs， the magnitude of Δu will increase or even be doubled according to the traveling wave reflection theory. This can help to accelerate the forward fault determination. And it can implement the pilot protection in DC system， especially in HVDC system.

In one embodiment， the forward value is

In this embodiment， the criterion one embodies as：

In one embodiment， the forward value is

In this embodiment， the criterion one embodies as：

k is a reliability factor， which 0.5 is a typical value. Introducing k as a reliability factor in criterion one can decrease the forward value so that it can detect the forward fault efficiently.

In one embodiment， the sensitivity requirement includes： |Δi| ＞i_set or |Δu| ＞Z_C·i_set， wherein i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.

In one embodiment， step 502 includes：

if the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time meet the following criterions， performing fault eliminating step， the criterion includes：

Fig. 7 illustrates schematically the situation when an inverse fault occurs， wherein the protection is relay 71.

Inverse fault， by definition， is the fault 72 occurred at the back side of point A as shown in Fig. 7. If designed properly， the protection relay should not operate under an inverse fault.

When an inverse fault occurs， as shown in Fig. 7， the relay 71 only senses forward travelling wave voltage Δu₁ (t) and forward travelling current wave Δi₁ (t) at the very beginning of the fault inception. And Δi₁ (t) ＝Δu₁ (t) /Z_C. In Fig. 7， forward direction is defined extending from the relay 71 to the DC line 73.

Therefore we have：

Δu·Δi＝Δi₁×Z_C×Δi₁＝Z_C×Δi₁ ² (6)

If Δi₁＞i_set， SO

Compare with the first criterion shown in equation 5， we can observe that the criterion for forward fault determination is not fulfilled because parameter k is always smaller than 1 as defined by the third criterion shown in equation 5. Therefore the protection relay will not mai-operate because of inverse fault.

If Δi₁＜i_set， we know from equation 5 that the protection relay will not operate either because of the second criterion |Δi| ＞i_set. Furthermore， for the inverse fault， Δu＝Δu₁＝Δi₁×Z_C， obviously， if Δi₁＜i_set， the Δu is less than Z_C·i_set， the criterion described by equation 5 that |Δu| ＞Z_C·i_set can’t be fulfilled as well.

The conclusion can be obtained that the protection relay 71 with the designed criteria will not mai-operate when an inverse fault occurs.

Fig. 8 illustrates schematically the situation when a forward fault occurs， wherein the protection is relay 81.

Forward fault， by definition， is the fault 82 occurred at the front side of point A as shown in Fig. 8. If designed properly， the protection relay 81 should operate under a forward fault. In Fig. 8， forward direction is defined extending from the relay 81 to the DC line 83. Therefore we have，

Δu·Δi＝-Δi₂×Z_C×Δi₂＝-Z_C×Δi₂ ² (7)

It can be obtained from equation 7 that when a forward fault occurs， Δu·Δi is always negative， which is far less than

because parameter k is always positive as defined by the third criterion in equation 5.

However， it should be noted that there will be always be multi stages of traveling wave along the line because of refraction and reflection when the traveling waves reach the converter station or the fault point， as illustrated in Fig. 9.

The closer the fault 92 location to the converter substation 93， the quicker reflection wave will arrive and be reflected. Therefore， for the close forward fault， there will be possibility that the relay 91 will sense multi-stages travelling wave， both inverse traveling wave and forward traveling wave， before forward fault determination.

Even though， the sigh of the Δu·Δi will not change because the magnitude of the sum of forward traveling waves will always be smaller than the sum of inverse traveling waves due to the energy loss caused by refraction. That is to say， the first criterion of equation 5 will be anyhow fulfilled for the forward fault. For inverse fault， this phenomenon is not mentioned because protection operation will be completed before the arrival of the inverse traveling wave due to the long length of the transmission line.

Therefore， the protection relay 81 will determine forward fault as long as the second criterion of equation 5 is fulfilled， i.e. |Δi| ＞i_set or |Δu| ＞Z_C·i_set.

It should be pointed out that even if |Δi| ＞i_set is not fulfilled， there is possibility that |Δu| ＞Z_C·i_set can be fulfilled. The reason is that when reflection occurs， the magnitude of Δu will increase or even be doubled according to the traveling wave reflection theory. This can help to accelerate the protection relay determining forward fault even if the magnitude of the current fault component does not reach the threshold.

From the above analysis， when an inverse fault occurs， the value of Δu·Δi is larger than a positive threshold，

； and when a forward fault occurs， the value of Δu·Δi is with negative signal， always less than zero. There is large gap in the value Δu·Δi between inverse fault and forward fault. So it is easy to set the setting coefficient k， set k＝0.5 is ok for most all of conditions， it is not related to the feature of the DC converter substation.

When an inverse fault occurs， the value of Δu·Δi is larger than a positive threshold. Since k is smaller than 1，

is smaller than

So even if the |Δi| ＞i_set (the second criterion of equation 5 is fulfilled) ， we still have

that is to say， the first criterion of equation 5 will not be fulfilled. When a forward fault occurs， the value of Δu·Δi is always negative. Since k is higher than 0，

is always positive. So the first criterion of equation 5 will always be fulfilled. It only depends on the second criterion to trigger the forward fault determination.

In one embodiment， further comprising fault eliminating step：

activating protect operation if forward fault is determined.

In one embodiment， the predetermined threshold i_set is bigger than current noise of the line.

The setting i_set is to ensure the reliability of the relay. If |Δi| is too small， the calculated value of Δu·Δi may be wrong because of noise. The only demand for setting i_set is that it should be set larger than noise. So we can know that the setting i_set will be a relatively small value， which is helpful to ensure the sensitivity of the direction element.

In one embodiment：

fault direction determination step further includes：：

The local terminal protection can communicate with the remote terminal protection in DC system. When both the local terminal protection and the remote terminal protection detected the forward fault， it means the fault is occurring between the local terminal protection and the remote terminal protection. This fault is what means internal fault.

In one embodiment， further comprising fault eliminating step：

Both the local terminal protection and the remote terminal protection should activate the internal fault protect operation respectively. And if one of the protections detected the inverse fault， it means the external fault is occurring. And in generally， for the external fault， no protections operate.

Fig. 10 shows a flow-process diagram illustrating a DC grid protection method in accordance with the present system， comprising the following modules：

fault detecting module 1001： acquiring fault component travelling wave current value and fault component travelling wave voltage value of a DC line， the fault component travelling wave current value at fault occurring time is Δi and the fault component travelling wave voltage value at fault occurring time is Δu， wherein forward direction is defined extending from the protection to the DC line；

fault determining module 1002： if the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time meet following criterions， performing fault direction determination module， the criterions include：

fault direction determination module 1003： determining the fault as forward fault.

Performance analysis of the present invention

The present invention uses the travelling wave based direction element in DC system.

The method of the present invention can avoid mai-operation due to lightning disturbance by introducing time delay or differential mode.

There are two scenarios when a lightning occurs. One scenario is that no fault triggered by the lightning； another scenario is that a fault occurs in the line because of lightning， may with high resistance or not.

For the first scenario， travelling waves will appear in the line when a lightning occurs in or somewhere near the line. In this case， the protection should not mal-operate because of travelling wave. One usual method to avoid mal-operate is to use the time delay， for example， existing travelling wave protection which has been used in HVDC lines for years uses time delay to avoid mai-operation from lightning， whose operation characteristics is shown as Fig. 11， which will operate only when four voltage sampling points are larger than the threshold. The present invention’s algorithm can also use such delay to make sure the reliability considering of the lightning wave is very short.

For the second scenario， the fault occurs in the line because of lightning. The protection should operate to isolate the fault. The fault caused by lightning may be metal fault or fault with high resistance.

Good sensitivity on high resistance faults

The present invention’s method has good sensitivity on high resistance faults because it is based on fault direction determination rather than amplitude determination.

The ability of classical protection principle based on travelling wave to tolerate high impedance is limited， because when a fault with high resistance occurs， theamplitude of travelling wave is decreased， the higher fault impedance is， the lower amplitude will be.

While the present invention’s method has high tolerance for high impedance because it is based on fault direction determination， not based on the amplitude of travelling wave. Even when a fault with high resistance occurs， the fault direction in both line ends can be determined clearly.

Small communication bandwidth

For the pilot direction protection， only direction information needs to be transferred， ‘1’ refers the positive direction， ‘0’ is the negative direction， i.e. only one bit is transferred in the communication channel， so small bandwidth is needed.

Immune to Channel Asymmetry

Current differential protection needs currents of two ends to be highly synchronized， which raises the requirement on channel symmetry besides GPS based synchronization method.

However unlike the phasor information used in current differential protection， the direction information used in the present invention’s pilot protection has little requirement on synchronization. From this point of view， the principle is immune to channel asymmetry.

Extensive adaptability

In this section， the adaptability will be analyzed from two aspects： working principle and operation speed.

Working principle vs. adaptability

From above analysis， it can be obtained that the direction element the present invention is only related to line parameter Z_C and the relationship between travelling wave voltage and travelling wave current. It has no special requirement on the topology and control of the DC system.

Operation speed vs. adaptability

The protection principle of the present invention is based on the initial travelling wave； therefore the typical operation time for direction element will be very fast， normally less than 1ms. Considering the communication time between 1ms-20ms depending on the line length， the pilot protection will operate within 2ms-21ms.

Therefore， we can choose to configure the present invention’s pilot protection as main protection or backup protection according to the requirement on the operation speed of different DC systems.

For example， for point to point DC line or DC grid based on LCC technology， and point to point DC line with VSC technology， the present invention’s pilot protection can be used as either main protection or backup protection.

For DC grid based on VSC technology， the present invention’s pilot protection can be used as backup protection， because the requirement on the operation speed is quite high， usually within 5ms. Ifthe length of the transmission line is short， the time delay caused by the communication can be reduced， and then the present invention’s pilot protection can also be used as main protection.

It should be pointed out that the performance of the present invention’s pilot protection， when configured as backup protection， is much better than the existing backup protection based on differential current， of which the operation time is usually longer than hundreds milliseconds.

Operation Speed Comparison with differential current protection

Different current protection operation speed analysis

It is well proved that the differential current is influenced by line capacitance. The HVDC line differential current protection is heavily influenced because the line is very long and with very large capacitance. The reason why the differential current protection has slow operation speed is analyzed below.

External fault

Fig. 12 shows the situation when an external fault occurs. When an external fault occurs， there is current flowing through the distributed line capacitance， which will increase the differential current. Because of complex charge and discharge process of line capacitance and inductance， there will be large amount of high frequency components appears in the current.

Normally there are two ways to avoid the influence of capacitor discharge current and these high frequency components： one way is to set the threshold ofdifferential current to be a very high value， the other way is to set a long delay to wait until these transient currents disappear. In most practical cases， the latter one is adopted， which explains why the HVDC line differential current has long delay.

Internal fault

Fig. 13 shows the situation when an internal fault occurs， similar to external fault， the differential current also consists rich high frequency components which will exist for a relatively long time. Therefore it also needs time to calculate differential current with acceptable accuracy， which increases the operation time as well.

The present invention’s protection principle is based on travelling wave process during the transient state after fault inception. The complete distributed transmission line model is considered to calculate the traveling wave. We actually utilize the high frequency component to do judgment rather than waiting for their disappearance. That is the essential reason that the present invention’s pilot protection can achieve much faster response speed compared with traditional current differential protection.

Hybrid OHL and cable for transmission

The protection principle works for hybrid OHL and cable theoretically， however further studies are needed to come to the final conclusions.

Resonances influence

There is distributed line capacitance along the line， because the HVDC line is very long， the line capacitance is large. When a fault occurs， large oscillation in voltage and current appears ( “resonances” ) ， some traditional protection principle will be influenced heavily， such as traditional current differential protection， low voltage protection， etc.

But the present invention’s direction element is based on travelling wave， which uses the feature of charge and discharge process of line capacitance and inductance. It is not influenced by “resonances” .

Above is theoretical analysis， further studies are needed to come to the final conclusions.

Simulation

Fig. 14， aLCC DC Grid system is used to demonstrate the present invention method. And it should be noted that the present invention’s pilot protection can be used for both complex system like DC Grid (LCC or VSC) and simple system like two-terminal HVDC system (HVDC classic and VSC HVDC) .

Model

The ±800 kV 4-terminal series MTDC consists of tow rectifier stations R1， R2 and tow inverter stations I1， I2. Total length of the transmission line is 2000 km， including two branch lines (500 km each) and one backbone lines (1000 km) . Each converter station has a configuration with one 12-pulse valve group. The lower converter stations R1， I1 have been connected to earth electrodes normally. The upper converter stations R2， I2 also have earth electrodes but disconnected from the transmission line normally. Each rectifier converter will have a nominal DC voltage of 400 kV across； each inverter converter will have a nominal DC voltage of 373 kV across (refer to Section 2.5) ， and the HV DC line voltage to ground is about 400 kV (for R1 and I1) or about 800 kV (for R2 and I2) .

The protection relay 141 is located at one terminal of the +400kV transmission line as shown in the figure 15. And the internal fault is at the end of +400kV line， the external fault is at the beginning of the +800kV line. And the pole-pole wave impedance Zc is 264 Ω in this case. Relay 141is the local terminal protection communicated with the relay 142 which is the remote local terminal protection.

Internal DC fault case

This is an internal fault with very large fault resistance (3000 Ω) . And the fault occurs on +800kV pole line at 4s. The fatal component travelling wave voltage Δu， current Δi and power Δu·Δi are shown in the figure 15a and 1 5b respectively.

It is very clear that the polarity of the travelling wave current and travelling wave voltage is different. As a result， their product is negative. Thereby， it meets the operate requirement in equation 5， both protection will detect it’s a forward fault and finally the pilot protection will detect the internal fault and trip.

External DC fault case

This is an external fault with very large fault resistance (3000 Ω) . And the fault occurs on +400kV pole line at 2s. The fatal component travelling wave voltage Δu， current Δi and power Δu·Δi are shown in the figure 16a and 16b respectively.

It is very clear that the polarity of the travelling wave current and travelling wave voltage is the same for local terminal and different for remote terminal. As a result， remote terminal will detect a forward fault with expectations because the product of travelling wave current and voltage is negative， it easily satisfy the equation 5. And for local terminal， the product of current and voltage is calculated as shown below，

In the calculation above， considering it’s an external fault， travelling wave voltage and current only consists of forward directional travelling wave， it means Δu＝Z_C·Δi. Obviously， the operate value Δu·Δ is bigger than threshold； local terminal will detect a reverse fault with expectations.

Thereby， after exchange the directional information by communication channel， the pilot protection will detect the external fault and will not trip reliably.

Here are the advantages of the present invention：

1. It can be used for different kinds of DC system including two-terminal HVDC systems and multi-terminal DC Grid systems.

2. It has good sensitivity for high resistance fault.

3. It has good sensitivity for total-reflection boundary conditions.

4. Low requirement for communication bandwidth and communication channel symmetry compared with current differential protection.

5. Fast operation speed. Generally， the time for algorithm ＜1 ms and total operate time ＜15 ms including the communication.

6. Very reliable. Measurement error and big disturbance has been taken into account inherently in the algorithm.

7. It is immune to HVDC control.

8. It can be used for main protection for DC system， especially for the cases that existing main protection based on travelling wave cannot work well (e.g. high resistance fault， some types of DC Grid) .

The above-identified embodiments are only used for representing several examples of the present invention， which are illustrated in detail， but shall not be understood to limit the protection scope of the present patent. It should be noted that， several modifications and/or improvements may be made for the skilled in the art， without going beyond the technical concept of the present invention， all of which fall into the protection scope of the present invention. Therefore， the protection scope of the present invention is dependent on the accompanied Claims.

Claims

A DC grid protection method， comprising the following steps：

fault detecting step： acquiring fault component travelling wave current value and fault component travelling wave voltage value of a DC line， the fault component travelling wave current value at fault occurring time is Δi and the fault component travelling wave voltage value at fault occurring time is Δu， wherein forward direction is defined extending from the protection to the DC line；

fault determining step： if the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time meet following criterions， performing fault direction determination step， the criterions include：

product of the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time being less than a forward value which is bigger than zero and is associated with a predetermined threshold； and

absolute value of the fault component travelling wave current value Δi at fault occurring time or absolute value of the fault component travelling wave voltage value Δu at fault occurring time meeting a sensitivity requirement which is determined by the predetermined threshold；

fault direction determination step： determining the fault as forward fault.
The method according to claim 1， wherein the forward value is
wherein i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.
The method according to claim 1， wherein the forward value is
wherein k is a constant， 0 ＜ k ＜ 1， i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.
The method according to claim 1， wherein the sensitivity requirement includes： | Δi |＞ i_set or | Δu |＞ Z_C·i_set， wherein i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.
The method according to claim 1， wherein the fault determining step includes：

if the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time meet the following criterions， performing fault direction determination step， the criterion includes：

wherein k is a constant， 0＜k＜1， i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.
The method according to claim 1， further comprising fault eliminating step：

activating protect operation if forward fault is determined.
The method according to any one of claim 1 to 6， wherein the predetermined threshold i_set is bigger than current noise of the line.
The method according to claim 1， wherein：

the protection includes local terminal protection and remote terminal protection， forward direction of the fault component travelling wave current value acquired by the local terminal protection is defined extending from the local terminal protection to the DC line， forward direction of the fault component travelling wave current value acquired by the remote terminal protection is defined extending from the remote terminal protection to the DC line；

fault direction determination step further includes：：

the local terminal protection determines the fault as forward fault， sends forward fault detected in local message to the remote terminal protection， and determines internal fault when receives forward fault detected in remote message from the remote terminal protection；

the remote terminal protection determines the fault as forward fault， sends forward fault detected in remote message to the local terminal protection， and determines internal fault when receives forward fault detected in local message from the local terminal protection.
The method according to claim 8， further comprising fault eliminating step：

activating protect operation in the local terminal protection if internal fault is determined；

activating protect operation in the remote terminal protection if internal fault is determined.
A computer program comprising computer program code adapted to perform all of the steps of any one of the preceding claims when run on a computer.
A computer program according to claim 8， embodied on a computer-readable medium.
A DC grid protection system， comprising the following modules：

fault detecting module： acquiring fault component travelling wave current value and fault component travelling wave voltage value of a DC line， the fault component travelling wave current value at fault occurring time is Δi and the fault component travelling wave voltage value at fault occurring time is Δu， wherein forward direction is defined extending from the protection to the DC line；

fault determining module： if the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time meet following criterions， performing fault direction determination module， the criterions include：

product of the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time being less than a forward value which is bigger than zero and is associated with a predetermined threshold； and

absolute value of the fault component travelling wave current value Δi at fault occurring time or absolute value of the fault component travelling wave voltage value Δu at fault occurring time meeting a sensitivity requirement which is determined by the predetermined threshold；

fault direction determination module： determining the fault as forward fault.
The system according to claim 12， wherein the forward value is
wherein i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.
The system according to claim 12， wherein the forward value is
wherein k is a constant， 0 ＜ k ＜ 1， i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.
The system according to claim 12， wherein the sensitivity requirement includes： |Δi |＞ i_set or | Δu |＞ Z_C·i_set， wherein i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.
The system according to claim 12， wherein the fault determining module includes：

if the fault component travelling wave current value Δi at fault occurring time and the fault component travelling wave voltage value Δu at fault occurring time meet the following criterions， performing fault direction determination module， the criterion includes：

wherein k is a constant， 0＜k＜1， i_set is the predetermined threshold which is bigger than zero， and Z_C is wave impedance of the line.
The system according to claim 12， further comprising：

fault eliminating module： activating protect operation if forward fault is determined.
The system according to any one of claim 12 to 17， wherein the predetermined threshold i_set is bigger than current noise of the line.
The system according to claim 12， wherein：

the protection includes local terminal protection and remote terminal protection， forward direction of the fault component travelling wave current value acquired by the local terminal protection is defined extending from the local terminal protection to the DC line， forward direction of the fault component travelling wave current value acquired by the remote terminal protection is defined extending from the remote terminal protection to the DC line；

fault direction determination module further includes：：

the local terminal protection determines the fault as forward fault， sends forward fault detected in local message to the remote terminal protection， and determines internal fault when receives forward fault detected in remote message from the remote terminal protection；

the remote terminal protection determines the fault as forward fault， sends forward fault detected in remote message to the local terminal protection， and determines internal fault when receives forward fault detected in local message from the local terminal protection.
The system according to claim 19， further comprising， fault eliminating module：

activating protect operation in the local terminal protection if internal fault is determined；

activating protect operation in the remote terminal protection if internal fault is determined.