Learnings

Outcomes

The main outcomes of the project are:

• The successful design, development, and testing of an EDGE-FCLi, a novel prototype solid-state fault current limiter. The device has been developed from TRL4 “Bench Scale Research” to TRL6 “Large Scale Deployment to a commercial scale device” as per the original aims of the project.
• The selection of the University of Warwick 33/11kV primary substation as an optimal location for the installation of the EDGE-FCLi.
  The EDGE-FCLi was connected in series with the university’s three existing 1.4MW CHP generators fed directly onto the 11kV primary busbar.
• The successful installation, commissioning, energisation, and trial of the EDGE-FCLi on the 11kV network at the trial site.
• The production of policy documentation that capture how to operate, control, inspect and maintain the EDGE-FCLi:
  o STANDARD TECHNIQUE: OC1Z Operation and Control of GridON 11kV Embedded Distributed Generation Electronic Fault Current Limiting Interrupter (EDGE-FCLi)
  o STANDARD TECHNIQUE: SP2CAE Inspection and Maintenance of GridON 11kV Embedded Distributed Generation Electronic Fault Current Limiting Interrupter (EDGE-FCLi)
• The dissemination of learning from the project Through regular six-monthly reports and several presentations to the electricity distribution industry. Refer to the closedown report for a full summary list of all six-monthly reports and industry presentations given forthe EDGE-FCLi.
• The production of an EDGE-FCLi Performance report that summarises the performance of the EDGE-FCLi through the testing and trial phases of the project.
  The production of the closedown report to summarise and capture the learning items generated throughout the project along with all associated closedown activities.

Lessons Learnt

The following learning relates to the last 12 months of the project. The project is now closed, and a full record of all project learning is available in the six monthly reports and associated closedown report.

HV cable VLF ‘pressure’ testing

A routine requirement prior to the connection of a new piece of equipment is to carry out a Very Low Frequency (VLF) ‘pressure’ test on the HV cables after they are installed, but before they are terminated. For polymeric insulated 11kV cables this involves applying 10.8kV 0.1Hz sinewave between the conductor and earth for each cable. The purpose is to check that the integrity of the insulation is not compromised before final connection.
The HV cables had been terminated at the EDGE-FCLi but left unterminated at the switchgear cable box at the UoW site. It was therefore proposed to carry out the VLF test with the EDGE-FCLi in the circuit to avoid having to disconnect the terminations at the EDGE-FCLi. However, the manufacturer advised that this was not to be done to avoid any damage to the internal power electronic circuitry. As a result, the site project team carried out the VLF with the EDGE-FCLi in the circuit and reconnected the HV terminations after the test was successfully completed. This issue did not impact the overall connection and energisation activities.

Alarm screen configuration
A key element in connecting innovation equipment is consideration of how the alarm and trip signals from the device interface with the central control system. Our control engineers have two main alarm screens. The primary screen displays all alarms that are set to be
‘acknowledgeable’ i.e., the engineer needs to manually accept the alarm remove it from the page.
The primary screen is typically used for high priority alarms that require control actions from the engineer. The second screen displays manual or auto-acknowledged alarms. It is used as an event viewer. Auto-acknowledge alarms are typically low priority alarms that go
straight to the secondary screen to avoid presenting a burden of alarms to the engineer. The project team worked closely with the control systems team to ensure that the alarm priorities were correctly defined in the control system configuration prior to the final commissioning and energisation of the EDGE-FCLi.

Control polarity during hot commissioning

There was a key learning point on site during the hot commissioning of the EDGE-FCLi prior to the energisation onto the 11kV network at UoW. When carrying out the functional testing of the control commands from our control centre, it was observed that the Idle and Recovery commands were reversed i.e., the IDLE command would trigger a Recovery action and vice versa. This was tracked to a wiring issue in the HMI panel that marshals all the signal wiring between the EDGE-FCLi and the substation equipment. This was quickly resolved during the commissioning and caused only minor delays; however, the point reinforces the need for robust testing and commissioning processes at site to identify issues and rectify them before final energisation. We recommend that the post-FAT or pre-commissioning procedures are updated to check and confirm that any subsequent wiring modifications have been carried out
correctly.

BIT Trip 1

The EDGE-FCli disconnected itself from the network on 19 May 2021. After the initial investigation on the cause of the trip event, it was found the device tripped due to the Built In Test (BIT) functionality incorrectly interpreting that there had been a power electronics
failure within phase 3 of the unit. The reason for the BIT failure and trip was that the BIT protection function was triggered incorrectly when there was insufficient phase current magnitude to the specified requirement. The current was in the range of 25A when it is
required to be in the range of 40A as per the design. This failure was attributed to noise in the transducer measurement system under field trial conditions. This also explains why this condition was not observed in the factory testing prior to the installation at the UoW
site.
To resolve this issue the software was modified to raise the phase current threshold at which the BIT function is triggered to above 40A to avoid the impact of noise on the triggering logic. After a period of software implementation and verification by the manufacturer, the
software was successfully uploaded to the device and the EDGE-FCLi was reconnected to the grid on 19 July 2021.

BIT Trip 2

The EDGE-FCLi was successfully reconnected following the software modification implemented to resolve BIT Trip 1. However, the device only continued to operate until a further trip event occurred on 28 July 2021. After an initial investigation it was again observed
that the device tripped due to a failure of the BIT protection function. The EDGE-FCLi control system logs various parameters in its internal memory and these log files were issued to the manufacturer to undergo a more detailed investigation into the cause of the error. While this investigation was ongoing the EDGE-FCLi was left disconnected from the 11kV network, and the site was restored to its normal configuration.
The manufacturer carried out an analysis of the detailed log files and reported that in the lead up to the error and discovered that the EDGE-FCLi successfully passed several BIT initiations; however, the timing duration of a BIT triggered on phase 2 of the device was much smaller than the nominal duration programmed into the software. This then caused the second trip as described above.
Subsequently, the manufacturer has carried out extensive testing and simulations in their laboratory to try and replicate the fault.
However, none of the investigations have yielded any root cause. The last step was to send the Micro Controller Unit (MCU) printed circuit board in the EDGE-FCLi back to the GridON factory for detailed inspection. Following receipt of the unit in Israel, GridON carried out an extensive inspection of the printed circuit board and found two metal filings lodged on the MCU electronics were the source of the BIT fault. The first filing was found to have no impact on the control system operation. However, the second filing broached two signal pins on the printed circuit board and analysis showed that this caused a resistive coupling between the two pins, spuriously triggering the BIT function and causing the issue at the site.
This is a highly unusual and unlikely event to occur to a piece of operational hardware and it is not known conclusively where the metal debris originally came from. It is probable, however, that the debris was generated from the drilling of the gland plate at the bottom of
the LV cubicle to allow the multicore and LV cables to be terminated into the device. The MCU electronics have now been retrofitted with a plastic cover to avoid further debris interacting with the electronics. In future it may be advisable to drill the cable gland plates away from the LV cubicle to reduce the likelihood of stray metal debris finding its way on the sensitive electronics.

Telecontrol

An important aspect of the learning from the trial is related to the telecontrol configuration used on the EDGE-FCLi. The control engineers responsible for operating the device remotely have communicated feedback to the project team on improvements that could
be made to this interface.
The control engineers’ main concern was that the “IDLE” and “RECOVERY” mode labelling on the control screens was significantly different to traditional network equipment, which could cause some misunderstanding when trying to operate the device remotely. After
a period of review, the “IDLE” and “RECOVERY” commands could realistically map to new labels “OFF” and “ON” respectively on the control screens. This is a much simpler description of the operational modes from a control perspective and aligns with the existing
terminology for network equipment.
The learning point for future reference would be to engage with the control engineers in the detailed design phase of the project that deals with formalising the operational regime of the device that is being connected. In that way, there is greater visibility of the telecontrol aspects earlier in the project delivery and reducing the likelihood of changes to the telecontrol configuration after device energisation.

Generator circuit breaker tripping

If the EDGE-FCLi detects a network fault, experiences a device malfunction, or loses its auxiliary LV supply, the unit disconnects itself from the network by tripping its feeding circuit breakers (CB22, CB24) and CB26 that supplies the CHP generators at UoW.
When the project team was carrying out the remedial works to prepare the device for reconnection following BIT Trip 2, it was observed that the automatic tripping of CB26 was an unnecessary action, especially when cycling the LV auxiliary supply to the EDGEFCLi during testing and commissioning activities. There is a risk that the CHP generators can be accidentally disconnected unnecessarily leading to a customer outage. After a review of the protection scheme, the tripping of CB26 does not serve any identifiable technical purpose and therefore the decision was made to remove the associated trip links from the circuit breaker trip circuit.
Note: The following sections are only required for those projects which have been completed since 1st April 2013, or since the previous Project Progress information was reported.