Tearing Up the Rulebook – Do We Really Need All of These Ancillary Services?
By Robin Preece
The short answer is ‘Yes – for now’ but we must stay future-focussed and plan the next stage in power systems operation. Otherwise, we risk constraining new devices by forcing them to replicate the services of the synchronous machines we are used to and do not fully exploit the benefits and flexibility they offer.
We have built power systems on a backbone of synchronous machines. They have been the workhorse of the modern power system for decades and you can travel around the world finding similar operational principles that are all based on the central idea that synchronous machines provide the glue (or more correctly, synchronising torque) that holds the system together. Even this discussion within this newsletter is predicated on the basis that new technologies should provide services that emulate or replicate the properties – inherent or controlled – of synchronous machines.
We are so interested in ideas like virtual inertia because we (correctly) place a lot of value on the large reserves of synchronous rotational kinetic energy that is directly coupled to the system. The generic use of the word ‘inertia’ when what we really value is instantly available energy is a debate for another article. Ideas like virtual synchronous machines are attractive as they take new power electronic (PE) converters and transform them – through their controls – so they behave in a familiar way like the synchronous machines we are used to. But we must remember that these devices are not the same – and that any emulation of services they provide is fundamentally limited. We are just sticking a synchronous machine mask onto a converter, and that mask might slip at the worst time.
One gigawatt’s worth of synchronous machines probably has about 3−6 GW.s (i.e. GJ) of rotational energy stored mechanically in the rotors. A one-gigawatt VSC-HVDC link has approximately one hundredth of this energy stored electrostatically in the capacitance of the converters and cables. These devices simply cannot support the system in the same way. Obviously, not all of the energy can be extracted from either system during a disturbance, but if we expect our virtual synchronous machines to fill in the holes that are left when we start decommissioning synchronous machines then we will be left embarrassed and upset when the energy runs out in milliseconds, and our systems collapse.
But these devices are also more flexible. To a large extent, power electronics interfaced loads can operate quite happily at voltages and frequencies that would fall well outside our current constraints on operational standards. We are left in the somewhat unusual situation where the increase in PE-interfaced devices is leading to reduced ‘inertia’, therefore there is higher rate-of-change-of-frequency (RoCoF) and higher subsequent frequency deviations, so we make our power electronics provide virtual inertia, but the PE devices do not really care about the increased frequency deviations anyway. They are treating symptoms that don’t affect them. But we still have lots of devices that are affected I hear you cry – and this is true. In many ways, the transition to this future state will be harder to manage than the future state itself. This is a future that will take a good few decades to reach before we have extreme dominance from PE devices. This will be a long and complex transition. This is the challenge that the European MIGRATE project has been focussed on. MIGRATE has developed research in many relevant areas from fundamental modelling through to demonstration of real-time monitoring and control solutions. With 24 project partners from industry and academia, MIGRATE has shown how greater penetration of PE devices can be achieved without compromising system stability. Of course, this is just one piece in the puzzle.
So what should we do? Tear up the rulebook of operational standards and limits and start again? Probably not, right. But we certainly need to look carefully at the standards, the permissible RoCoF, the under-frequency load shedding thresholds, the voltage ride-through requirements, and countless other details – and determine how these will need to change in the face of the huge challenges our power systems are experiencing. This needs to start with understanding the technological limitations. What can these devices do and when can they do it? How long can they do it for? What grid-connection requirements are no longer relevant or are limiting technological progress and decarbonisation? What new requirements do we need to add? Do we need new standards about controller interaction and device compatibility? Do our approaches towards harmonic penetration within our systems need updating? Are you noticing that this is an article with more questions than answers?
The good news is that this is a vibrant research area and that the technological progress is rapid. Ancillary service markets spring up and devices dive in to provide the services. There is a wealth of research identifying the technological limits of these technologies and developing easily integrated control schemes that support the system in innovative ways. This newsletter will no doubt be showcasing many of these. We also need to develop the language we use to discuss and measure the performance and operation of our future non-synchronous systems. When frequency is no longer analogous to synchronous machine speed, do we care about RoCoF or are there other metrics that are more important? More questions, no definitive answers…yet. But the report from the IEEE Task Force on Stability definitions and characterization of dynamic behaviour in systems with high penetration of power electronic interfaced technologies is eagerly awaited.
So perhaps we do not need to tear up the rule book. But we almost certainly need to take out a few pages, changes some numbers, and probably write some new pages. This has to be a collective effort, it needs understanding of device capability and operability as well as system requirements. But we shouldn’t just force our square converters into the round holes left by decommissioned synchronous machines.
Dr Robin Preece is a Senior Lecturer in Future Power Systems in the School of Electrical and Electronic Engineering at the University of Manchester, where he has been an academic since July 2014. Since then, he has helped to secure over £4 million in funding for the University of Manchester. Dr Preece has published more than 60 international peer-reviewed publications including numerous publications in the IEEE Transactions on Power Systems and on Power Delivery, the flagship journals of the IEEE Power and Energy Society. He has presented his research at major international conferences hosted by the IET, IEEE, IFAC, and Cigré, and is actively involved in numerous international working groups on the stability of future power systems.