Francisco de León
For about a hundred years following the war of the currents (between Edison and Tesla), settled at the end of the 19th century in favor of ac, transmission and distribution (T&D) systems have followed the well-established architecture still in use today. The bulk of the power is generated using large synchronous machines (driven by steam and water turbines) that are connected to step-up transformers operating at 50 or 60 Hz. Then, extra high-voltage and ultrahigh-voltage ac transmission lines carry electric power over long distances to the consumption centers. Next, a series of cascaded step-down transformers reduce the voltage for the different functions (such as subtransmission and distribution) and applications (such as industrial, commercial, and residential). Since the 1950s, various high-voltage dc (HVdc) projects around the world have demonstrated that dc transmission is economically viable for high-power long-distance transmission.
For about the past 20 years, T&D systems have been going through a radical transformation. A serious push is being made to have green and distributed power generation (solar and wind) as competitive sources of energy. Hundreds of HVdc lines are in service worldwide. The length at which dc transmission becomes more economical than ac transmission is shortening at a high speed. Hundreds of demonstrations of smart grid technologies exist around the world. These technologies are creating a new paradigm in the design of T&D systems that is being pulled by the constant improvement of power electronics technologies, such as new converter topologies based on modular multilevel converters, the constant increase of current and voltage ratings of electronic switches, and higher switching frequencies. Power electronics technologies are slowly penetrating the power system to perform novel control strategies.
Currently, only a few of the flexible ac transmission systems technologies have succeeded economically for generalized applications in power systems. For example, static var compensators and static synchronous compensators are ubiquitous today. However, from my privileged high-level view as editor-in-chief of IEEE Transactions on Power Delivery, I see great efforts among researchers pushing toward a future with distributed (renewable) generation mostly conveyed to the final consumer via dc lines. Perhaps even most consumers will also be producers of electricity, and the dc lines will be operative only during contingencies and other power unbalances.
In the near future, I foresee an increasing penetration of dc power transmission sharing the right-of-way with (legacy) ac lines, but eventually, because of their higher efficiency, I predict that dc transmission will dominate. Photovoltaic cells, batteries, electric vehicles, and even motors controlled with variable-frequency drives can be connected directly in dc. There are examples using dc buses to distribute power already, but I think we will eventually have a dc system with “ac links.” This topology would be exactly the reverse of today ac systems with some dc links.
Substantial research is being done around the world on distribution-level dc lines. Today, most large dc systems operate transmitting power from point A to point B. A few multiterminal dc projects are starting to appear at different locations. In the not-too-distant future, ac and dc systems will work in tandem. Trillions of dollars have been invested in the development of the ac grid for more than a hundred years. Thus, the ac power system cannot just be abandoned because a new technology is available. It is not possible to dispose of the grid and all its devices like a cell phone when a new “G” is available. However, I think that eventually dc systems will prevail. The dc systems are not only more efficient but dc lines on the same right-of-way can carry up to four times the power of ac lines. Eventually most (if not all) ac transmission lines will be converted to dc lines. When large transformation ratios are needed, for example, distribution at 25 kV to utilization at 240 V (a ratio of over 100), ac links in the form of magnetic transformers may still be needed. However, those ac links will certainly operate at much higher frequency.
Research on solid-state transformers (SSTs) is taking place. SSTs are still more expensive than traditional transformers. In the near future, the installation of SSTs can be justified only by the flexibility they offer because they provide access to several dc voltage levels, improved controllability, and the possibility to inject reactive power to all ac terminals. The single-switch solid-state distribution-level transformers may be the first ones to become a reality.
A challenging problem and motive of much research, with dc systems, is the design and operation of breakers. In 50/60-Hz ac systems, there are naturally 100/120 current zero crossings per second that are used opportunistically to extinguish the short-circuit currents. No naturally occurring current zero crossings exist in a dc system, and substantial effort is being made to produce dc breakers. An alternative to dc systems, still very much at the developing stages, is to use (very) low-frequency systems, which have most of the advantages of dc systems and regularly reoccurring current zero crossings.
The transformation of the power system is not only related to ac/dc, integration of renewables, and smart and microgrids. New and better diagnostics and condition monitoring tools for power T&D equipment are also currently developing at an accelerated speed. For example, the real-time thermal rating of overhead lines, cables, and transformers is receiving much attention. Research on unmanned aerial vehicles to inspect overhead transmission lines, insulators in particular, is getting traction. Many new sensors are constantly proposed to improve the available diagnostics information on T&D equipment. Online and offline transformer diagnostic tools are constantly improved. New monitoring methods to account for transformer aging and estimation of their remaining life appear frequently. Currently, there is an explosion in fault location techniques for overhead transmission lines and underground cables: from electromagnetic time reversal, to traveling-wave methods, to impedance measurement, to the location of high-impedance faults, with applications to both ac and dc (high- and low-voltage) systems. Additionally, several techniques have recently been developed for the location of faults in internal insulation in transformers and power electronics equipment.
Perhaps one of the technologies that has the potential of having important impacts on the future of T&D systems is synchrophasor measurements using phasor measuring units (PMUs). PMUs allow engineers to have a clear picture of what the operating conditions of the power system are. PMUs are making the power system be more observable. Placement, calibration, and interference are a few of the topics that researchers are working on today. I very much agree with a recently published visionary paper proposing synchronized measurement of waveshapes as the evolution of PMUs.
Less visible, but very important for the future of power systems, is the eternal search for the increase of cable ampacity and transformer efficiency. Cable ampacity calculations are constantly improved for existing and new installation techniques. More details are being increasingly considered, for example, unbalanced operation, cable crossing, soil dry out, and so on. Cables installed in tunnels (forced ventilated and nonforced ventilated) seem to be viable alternatives for system expansions in many instances. Design improvements of converter transformers are taking place. New liquid dielectrics are frequently tested. Much research is being done on dc cables because, for the same conductor cross-sectional area, they can carry more current than ac cables. Conductor losses of dc cables are less than the losses of ac cables because there are no induced eddy currents due to skin and proximity effects. There are neither insulation losses nor steady-state charging current. Therefore, the length at which dc transmission is more economical than ac transmission is substantially shorter for cables than for overhead lines. However, there are some particularities of HVdc cable technology that need to be better understood. For example, space charge accumulation poses electrical stress in the insulation, which limits the ampacity. Therefore, the current-carrying capacity is limited not only by the losses but also by dielectric stresses in the insulation.
Finally, I would like to briefly discuss topics related to the application of artificial intelligence (AI), data-driven control actions, and machine learning to T&D systems. Some of those techniques are certainly of great value for the diagnostics of T&D devices. A few important examples where AI is making an impact include diagnostics of power apparatus, fault location, and system reconfiguration. However, many papers are submitted to IEEE Transactions on Power Delivery that propose AI solutions to nonexistent problems. Some authors are trying to adapt a problem to an AI technique. The most common example is the application of AI to the classification of power quality problems. Power quality problems do not need to be classified—they need to be solved!
I would like to end on a positive note. Looking at the number of innovative and important contributions that have been made in the past few years and that for sure will continue at an accelerated pace, I foresee that the next generation’s power equipment and power system architecture will better serve future generations. Changes in power systems happen slowly because mistakes can cause serious disruptions to everyday life (blackouts) and are very costly. However, the power system will continue growing and continue providing an excellent service to humanity as it has already done for several generations.
Digital Object Identifier 10.1109/MPE.2022.3231000