RfG – Fast Fault Current Injection Update Antony Johnson / Peter Simango National Grid – Network Capability March 2017

Disclaimer Note this presentation covers the study work completed to date in respect of evaluating the requirements for fast fault current injection. A test network has been set up and the models tested. A multi machine study has also been set up and initial studies started however these need to be completed before results and conclusions can be drawn Stakeholders are therefore requested to be aware that such analysis does take time and the results presented in this presentation reflect a snap shot of the work completed to date. A further update with a more complete set of results will therefore be provided at the April meeting.

Summary Test Network and proof of concepts Test Network Performance – Synchronous Machines Test Network Performance – Power Park Modules Variations in Converter based reactive current injection Virtual Synchronous Machine Multi Machine Study Set up and initial work completed to date. Summary / Next Steps

Test Network( Fig1)

Test Network Behaviour of Synchronous Plant Assumptions A three phase fault was applied at 400kV Ref Bus The fault infeed at 400kV Ref Bus was taken as 10000MVA Synchronous plant varied at the 33kV busbar varied from 1MVA – 50MVA The same starting point was assumed for asynchronous plant

Test Network - Results Behaviour of Synchronous Plant (1MVA)

Test Network - Results Behaviour of Synchronous Plant (25MVA Group)

Test Network - Results Behaviour of Synchronous Plant (50MVA Group)

Test Network – Synchronous Plant Summary of Results To achieve a minimum retained voltage of 10% at the 33kV busbar, at least 25MVA of synchronous plant is required at the 33kV busbar The current injection from this group of synchronous plant synchronous plant is approximately 4.3pu The retained voltage at 33kV is around 12% during fault conditions The greater the number of machines the higher the retained voltage on the 33kV busbar

Test Network Behaviour of Power Park Modules Assumptions The same number (rating) of Power Park Modules as synchronous machines was used The Asynchronous machine was modelled as a static generator with Phase Locked Loop (PLL) control Standard Power Factory Static Generator model applied The blocking voltage varied between 10% and zero Voltage and current injection plots obtained

Test Network - Results Behaviour of Power Park Modules (Switching threshold = 0.1pu)

Test Network - Results Behaviour of Power Park Modules (Switching threshold of =0) (25MVA)

Power Park Modules – Reactive Current Injection Reactive Current (p.u) Static Generator Max Value Ramp Rate Time Delay Time (ms)

Static Generator Parameters

Test Network - Results Power Park Modules (Max 1 Test Network - Results Power Park Modules (Max 1.5pu injection 25MVA group)

Test Network - Results Power Park Modules (Max 1 Test Network - Results Power Park Modules (Max 1.5pu injection 50MVA group)

Test Network - Results Power Park Modules (Max 1 Test Network - Results Power Park Modules (Max 1.5 injection 100MVA group)

Test Network – Static Generator with Fast fault current injection The more the machines the better the terminal voltage It is possible to delay the injection The higher the injection the less the number of machines required to achieve the desired retained voltage For Power Park Modules the maximum reactive current injection (eg 1.5pu) is critical Max Current Injection (pu) 1.5 No of Synchonous machines (1MVA each ) 1 25 50 100 33kV Retained Connection Voltage[pu] 0.043 .0.833 0.17

Test Network Comparison for the three cases No of machines (1MVA each ) 1 25 50 100 Static Generator Retained Connection Point Voltage 0.011 0.023 0.033 Sync. machine retained Connection Point Voltage 0.13 0.24 0.38 Stat. Gen With FFCI retained Connection Point Voltage 0.043 0.0833 0.17 When the number of Power Park Modules is very low the fault current contribution is insignificant A group of synchronous machines will offer significantly more voltage support compared to the same number Power Park Modules

Test Network - Results VSM Model The network is the same as that shown in Figure 1 VSM technology uses the same static generator but the controller has been modified to reflect VSM capabilities

Test Network - Results VSM 25MVA

Test Network - Results VSM 50MVA

Test Network - Results VSM 100MVA

Test Network – VSM Performance The more the machines the better the retained Connection Point voltage The VSM offers slightly better performance than a standard PPM Converter using PLL VSM Size 25 50 100 33kV Retained Connection Point Voltage[pu] .0035 0.071 0.141

Multi Machine Study approach and Methodology Full GB Transmission Network Includes DNO Networks Specific area of interest will focus on an area of the network known to have a high volume of Embedded Generation: South West Base case study Intact network conditions System conditions – Max / Min Demand All embedded generation initially modelled as negative demand Solid Three phase short circuit fault applied at various points across the South Western part of the Transmission System Voltage profile assessed across the Transmission System and DNO system during and after the above faults

Area Under Study

Multi Machine System Study Assumptions Demand as negative Gen Base Case Fault Condition: Solid Three phase double circuit fault between Indian Queens and Taunton substation

GB System Study Embedded Generation modelled as Negative Demand The Voltage at the point of fault on the Transmission System is zero A number of busbars have a voltage above 10% due to network interconnection The voltage at Hayle 33kV busbar is 0.012pu.

GB System Study Result Summary Embedded Generator modelled as 25MVA synchronous machine )

GB System Study Result Summary Embedded Generation modelled as 25MVA synchronous machine The Voltage at the point of fault on the Transmission System is zero The Connection Point Voltage at Hayle 33 kV Substation has increased from 0.012pu to 0.163pu This improvement has cascaded to some of the busbars around the network

GB System Study - Results (Embedded Gen Modelled as static Gen)

GB System Study - Summary Embedded Generation modelled as 25MVA Static generator The Voltage at the point of fault on the Transmission System is zero The Connection Point Voltage at Hayle 33kV substation has increased from 0.012pu to 0.051pu

Multi Machine System Studies Progress to Date Case 1 Base Case with Embedded Synchronous Generation explicitly modelled in DNO network and specific transmission faults applied. – Ongoing Case 2 Base Case with Embedded Power Park Modules explicitly modelled in DNO Network and specific transmission faults applied - Ongoing Case 2A As per case 2 with variations in reactive current performance – To be completed Case 2B As per case 2 but the converters installed within the Power Park Modules are modelled with Virtual Synchronous Machine (VSM) capability - To be completed

Summary / Next Steps Test network setup to test machine and controller models Multi machine network set up and studies ongoing It is too early to draw any concrete conclusions Initial results would indicate that low fault current, particularly from Converter based plant provides little voltage support during Transmission System faults The greater the volume of Embedded Generation the higher the retained voltage Distribution and volume of Embedded Generation at this stage would appear to be a significant factor in determining the parameters for fast fault current injection and voltage against time curves. A further update will be provided at the next meeting in April