Learnings

Outcomes

This project is the first trial in the GB to demonstrate a working industry standard VSM prototype in a highly realistic testing environment.

The findings of the tests indicate that VSM is a promising technology that can certainly be part of the suite of tools that can be used to address the upcoming challenges associated with the decline of synchronous generation on the system.

The project also highlighted the importance of establishing minimum specifications for the behaviour of VSM/Grid Forming Converters which reinforces the work which has been completed as part of the Grid Code Modification proposal (GC0137).

In addition to the specification, a well-defined testing strategy will need to be formulated to verify the detailed performance of VSM to ensure that the specified performance is being met. This project has demonstrated the different avenues that can be adopted to verify the actual performance of VSM against their technical specification, which is an important consideration during compliance testing. As a result, the findings have been shared with the Great Britain Grid Forming (GBGF) Best Practice Group and the Stability Pathfinder team within NGESO who are actively working on developing guidance in this area.

Lessons Learnt

• In general, from the tests conducted and the engagement with stakeholders, it was observed that VSM technology has reached a relatively mature stage (around TRL 5), where commercial prototypes are generally available from some manufacturers.
• Existing NIA and other innovation projects mainly focus on investigating VSM technology with TRL of 4 or below. The project is the first exercise in GB to test an industrial standard VSM prototype in a real network environment – demonstrating the technology’s TRL level of 6. While there are aspects of performance that could be improved, the tests conducted demonstrated that it is feasible for the concept of VSM to be implemented in industry-standard converters and battery systems, thus they could potentially be used in a real system.
• From the tests completed, the VSM unit could deliver inertial response during frequency disturbances and providing reactive power support during voltage disturbances, which aligns well with the behaviour of a synchronous generator.
• During the fault tests, it was observed that the VSM unit could initially experience a drop in its output current and have a delay in fault current injection, an undesirable behaviour where the objective is to increase system stability. The VSM performance is largely dependent on the way it is designed which means that the performance can be tuned to meet the desired requirements. This reinforces the need for a standard set of minimum performance criteria for VSM technology to be developed, this is what the Grid Code GC0137 Working Group has delivered. This will help equipment manufacturers to design their equipment to meet these requirements and help NGESO when assessing the performance of VSM or other grid forming technologies during the compliance tests.
• As the performance of the VSM can be modified by implementing software modifications e.g. tuning the parameters of the controller, the VSM unit could offer more flexibility in the damping and inertial response characteristics compared with synchronous generators whose response is largely dependent on the machine design parameters.
• It was observed that from a testing perspective, it is preferable to have the VSM unit operating at a non-zero active power operating point during the tests. This is because when the current is very low, the power quality of current can be very poor, which could affect the observation and evaluation of the behaviour during the tests.
• The project demonstrated that VSM can also help to energise the network following a Black Start Event and support the recovery of supplies.
• It was identified in the project that there could be limitations as to what can be tested in real systems as opposed to what can be tested in the simulation environment. It is therefore important that that the right test systems are developed that would enable physical tests to be carried out as much as possible.
• Compliance tests are important to verify the actual behaviour of VSM systems and to calibrate the simulation models that have been developed.
• The studies demonstrates that the developed experimental approach based on Network Frequency Perturbance method provides a valuable tool for network operators and manufacturers for experimentally evaluating the inertial and damping performance.
• Overall, it is believed that VSM technology can play an important role in maintaining the stable operation of low-inertia grids. Further work is likely to be required to enable large scale deployment of the technology e.g. to identify the collective behaviour of VSM controllers when a large number of these devices are commissioned on the system.

Review of benefits case

The success criteria of the project were based on two main aspects:

1. Inverters can be modified and used to allow inverter connected assets to provide services traditionally provided by synchronous assets generation, e.g. inertia, fast fault current injection and black start.
2. A battery system installed at the Power Networks Demonstration Centre can successfully demonstrate the application of this technology to give confidence to the ESO that this is approach could be used to support system security.

The tests conducted have shown a promising response in line with VSM expected behaviour with good correlation between the simulated and actual tests results where applicable. The inertial response in the form of fast active power injection/absorption following changes in frequency and fast reactive current injection/absorption to support the voltage during faults or voltage step changes were generally witnessed in both the simulated and real tests. The section below provides a high level overview of the test results observed from WP1, WP2 and WP3.

• Voltage step tests: up to ±10% voltage steps have been applied and a proportional reactive power injection (maximum of 110kVAr) or absorption (maximum of 130kVA) was provided by the VSM during the step voltage changes.
• Frequency step tests: overall a proportional active power injection or absorption is provided by the VSM for step frequency drops or rises respectively. A maximum active power injection of 200kW was observed for a -2Hz frequency step. An opposite reactive power response is observed due to changes in the voltage during the frequency steps. There were no observable changes to active and reactive power during slow frequency ramps.
• Standard fault ride-through tests: the VSM was able to ride-through all symmetrical fault events except for cases where the test voltage recovered to 1.4pu. In this case, the VSM did not provide pre-fault output levels. Furthermore, current oscillations were observed during some of these events which require further investigation. The VSM also rides-through asymmetrical fault events. However, output current oscillations were observed more frequently during these faults, which makes it difficult to discern the response of the VSM (particularly reactive power response) during these events.
• Severe fault ride-through tests: the VSM rides-through all the events down to 50% voltage depression. However, output current oscillations occurred at this voltage depression, similar to those observed in some of the standard fault ride-through tests.
• Real fault tests: a single-phase fault was applied between the Triphase and VSM and was met with a proportional reactive power injection by the VSM. The fault duration, however, was too short to observe any support to the voltage from the VSM.
• Phase angle step tests: the VSM was able to ride-through voltage angle shifts of 2.5°-60°. Higher angle shifts applied from 90° and above resulted in the Triphase tripping on overcurrent.
• Black start tests: all black start tests were successful. the VSM based Battery and Inverter system was able to supply loads applied both pre and post network energisation. The maximum loading applied was 160kVA. Small frequency step changes of around 0.1Hz were observed with each load step applied. Furthermore, a three-phase PV inverter was synchronised to the VSM energised network and was able to export 5kW, thus reducing the active power output of the battery system by the same amount.
• Steep Frequency Ramp Test: The VSM can provide an inertial response in a similar manner as a Synchronous Generator and a Synchronous Condenser i.e. with a comparable response time and similar characteristics. However, relatively large levels of mismatch between the theoretical and the experimental values were observed.
• PHiL tests with low inertia: A GB transmission network model has been developed in the RTDS with an inertia level of 95 GVA to emulate a low inertia scenario. A frequency disturbance event with 1.8 GW active power mismatch has been simulated and the VSM capacity in the PHiL simulation is varied between 1 GVA to 4 GVA. With the addition of VSM capacity in the network, it was observed that there was a noticeable improvement in the ROCOF due to the additional inertia provided by the VSM.
• Frequency Sweep of the VSM Response: Three different methods - “Peak Response”, “Asymptote in Low Frequency Range” and “Curve Fitting” have been illustrated to determine the inertia constant and damping constant. Each method was demonstrated to have its own set of advantages, drawbacks and impacts on the accuracy of the calculations for the inertia and damping constants.

Dissemination

The project team has presented the project learnings in a workshop as part of the GFC Testing Working Group led by the Fraunhofer Institute for Solar Energy, where manufacturers, ESOs and battery suppliers from UK and Germany were also present.

The project team have also published two research papers. 1) “Comparative Evaluation of Dynamic Performance of a Virtual Synchronous Machine and Synchronous Machines” which was presented at the 9th Renewable Power Generation Conference in March 2021 and 2) “Experimental assessment and validation of inertial behaviour of virtual synchronous machines”.

The project team is also planning to work with Fraunhofer ISE, BELECTRIC and other stakeholders to engage with international professional organisations, e.g. CIGRE, to form working groups for sharing the learnings from this project and other relating projects conducted by the stakeholders.

Following the completion of final project report in August 2021, the project team was invited to present the main findings of the project to the Grid Forming Converters Best Practice group which was formed to provide guidance on the technical specification which was developed as part of GC0137 for Grid Forming Converters.

Furthermore, the lessons from this project were shared with the ESO Stability Pathfinder team to support the creation of the compliance guidance notes which will be used to verify the performance of grid forming converters who are successful in the stability pathfinder tenders.