Review of mitigation technologies for terrestrial power grids against space weather effects

Electrical Power and Energy Systems 82 (2016) 382–391

Contents lists available at ScienceDirect

Electrical Power and Energy Systems journal homepage: www.elsevier.com/locate/ijepes

Review of mitigation technologies for terrestrial power grids against space weather effects Michael Johnson a,⇑, George Gorospe b, Jonathan Landry c, Anja Schuster d a

University of Limerick, Ireland NASA (Ames Research Center), CA, USA c École de Technologie Supérieure (ÉTS), Montréal, Québec, Canada d Technische Universität Darmstadt, Germany b

a r t i c l e

i n f o

Article history: Received 19 August 2014 Received in revised form 18 February 2016 Accepted 25 February 2016

Keywords: Space weather Gemagnetically induced currents Power grids Transformers Mitigation strategies

a b s t r a c t This paper discusses the earth-based effects of solar weather and presents a review of mitigation and protection techniques for the terrestrial power grid infrastructure. Solar events such as Coronal Mass Ejections (CMEs), solar flares and associated recombination events are one of the driving factors in space weather and the solar wind intensity. Even though it is located at such a great distance from our nearest star, the Earth and its associated satellites are still directly affected by variances in these space weather phenomena. On the surface of the planet, nowhere is this more immediate and important than with the terrestrial power grid, which is responsible for delivering electrical power to much of the planets population. Large-scale variations in solar activity can result in potentially devastating effects on the terrestrial power grid and the associated infrastructure. A team project was undertaken at the International Space University (ISU) Space Studies Program (SSP) 2013 to categorize and mitigate the risks involved in such a solar event. As part of this research, which included risk assessment for satellite, spacecraft and terrestrial resources, this paper presents a review of the terrestrial power grid and its inherent susceptibility to such phenomena. Mitigation schemes, techniques and approaches ranging from adaption of the existing power grid to alternative systems are considered in this paper, which allow for continued electrical power delivery and transmission, even in the face of such detrimental space weather effects. Ó 2016 Elsevier Ltd. All rights reserved.

Introduction Opening ‘‘Space Weather Destroys Stuff” – [1]. A bold and undeniably truthful statement from Dr. Pete Worden and the T.P SolarMAX team of the 2013 Space Studies Program at the International Space University [2]. In the past 100 years, the Earth and associated (highly susceptible) technological advances of the last century have escaped the attention and associated destructive sideeffects of any major solar event. The cause and effect of such solar storms, space weather and solar activity are still an area of ongoing research within the scientific community, with new discoveries being made and mysteries solved even today [3]. In this paper, the term ‘‘Solar Events” is used to encompass any solar-related phenomenon with wide-reaching effects on other ⇑ Corresponding author. E-mail address: [email protected] (M. Johnson). http://dx.doi.org/10.1016/j.ijepes.2016.02.049 0142-0615/Ó 2016 Elsevier Ltd. All rights reserved.

bodies in the solar system. Coronal Mass Ejections (CMEs), solar flares, magnetic recombination on the surface of the sun and other associated phenomena are all included under this heading [4]. In the Earths recent history, the largest coronal mass ejection and associated phenomena recorded occurred in 1859, and is nowadays referred to within the scientific community as the ‘‘Carrington Event” [5]. Were it to happen today, an event of this magnitude would have a disastrous impact on humanity, both on and off our planet. In 1859, aside from the spectacularly visual aurora associated with the phenomenon, the Carrington Event, while it was of a magnitude and scale heretofore unseen, the solar activity went largely unnoticed. Telegraph operations were one obvious exception, with the space weather events sufficient to interfere with transmissions and, in extreme cases, power the telegraph lines unaided for hours at a time. Nowadays, a space weather event on this scale would have wide-reaching and, in many cases, disastrous effects on Earths technology, infrastructure and assets. With all of the technological and electrical improvements in the past 150 years, our modern

M. Johnson et al. / Electrical Power and Energy Systems 82 (2016) 382–391

planet and its people are largely dependant on technology, electronics and electrical power for many of the services, conveniences and comforts taken for granted in the modern world: light, heating, communications, transport, healthcare, safety and security foremost among these. Solar activity comparable to that witnessed in 1859 would see the Earth suffering major setbacks, losses and potential regression due to our singular state of unpreparedness with respect to these space weather incidents. Nowhere would this impact be more widely felt than in the modern-day terrestrial power grid, which supplies electrical power to the majority of cities, communities and homes across the globe. Terrestrial power grids During geomagnetic storms, electric currents in the atmosphere and the associated magnetic field undergo large variations due to fluctuations in the solar wind intensity. These variations cause geomagnetically induced currents (GICs) in terrestrial conductors, such as power lines and buried pipelines. The terrestrial power grid, or electric-power transmission grid, refers to the bulk transfer of electrical energy, from generating power plants to electrical substations located near distribution centers. The associated transmission network is the interconnected mesh of these transmission lines, used to distribute the electrical power within a country or geographical region. Given a sufficiently active solar event, the currents induced in terrestrial power grids and transformers may be large enough to cause temporary or indeed permanent damage to the constituent elements of the power grid, resulting in temporary loss of power for large areas of the grid. Historically, the transmission and distribution of electricity have been coordinated by the same company or institution. Starting in the 1990s, however, many countries have liberalised the regulation of the electricity market, which has meant that the electricity power generation business is now generally distinct from the power distribution and transmission lines sector (Fig. 1). Most modern-day transmission lines use high-voltage threephase alternating current (AC) for most transmission and industrial usage. One noteworthy exception to this is railway electrification systems, which still use single phase AC [6]. High-voltage directcurrent (HVDC) technology is used for greater efficiency in very long distances (usually hundreds of kilometers)), or in submarine power cables [7]. HVDC links are also used to stabilize against control problems in large power distribution networks where sudden new loads or blackouts in one part of a network can otherwise result in synchronization problems and cascading failures. Electricity is transmitted at such high voltages (110 kV or above) to reduce the energy lost in long-distance transmission by minimising the transmission current through an inversely proportional increase in transmission voltage. Power is typically

383

transmitted through overhead power lines. Underground power transmission has a significantly higher cost and greater operational limitations, and so is only used in specific conditions, such as inner city power grids. Transmission lines are therefore designed to transport electrical energy as efficiently as possible, while still taking into account economic factors, network safety and redundancy. The associated power grids use components such as circuit breakers, switches and transformers in addition to transmission lines and cables to realize their desired implementation. A power transmission grid is therefore a network of power stations used in conjunction with these transmission lines, power grids and substations to supply electricity to the end user. The associated cost of electric power storage is high, therefore most power grids utilize the generated electricity immediately rather than store it for later use. As the typical electricity demand (load) is variable over days, regions and seasons, it is often more economically feasible for power companies to import any excess portion of the high-demand power requirements than to store or generate it locally. Because of the associated economic benefits of this load sharing between countries or regions, wide area transmission grids now span countries and even continents. However, this reliance on concurrent power generation and wide-scale power distribution networks and transmission lines, with minimal provision for energy storage, proves dangerously shortsighted when geomagnetically induced currents and space weather effects are considered. Geomagnetically induced currents Geomagnetically induced currents (GICs) are a manifestation at ground level of space weather, and are known to affect the normal operation of long electrical conductor systems, such as the power grids or electrical transmission systems described in the preceding section. During geomagnetic storms, electric currents in the magnetosphere and ionosphere undergo large variations due to correspondingly large fluctuations in the solar wind intensity caused by these storms. This time-varying magnetic field, which is external to the Earth in the atmosphere, induces telluric currents (electric currents) in the conducting ground. These magnetospheric and ionospheric variations also manifest in the Earths magnetic field, creating a secondary (internal) magnetic field, inducing an electric field at the surface of the Earth associated with these time variations in the magnetic field (Fig. 2). This surface electric field causes electrical currents (GICs), to flow in any conducting structure, such as a power or pipeline grid grounded in the Earth. This electric field, measured in V/km, acts as a voltage source across networks. GICs are often described as being quasi direct current (DC) in nature, although the variation

Fig. 1. Electrical power grid – image courtesy Wikipedia.

384

M. Johnson et al. / Electrical Power and Energy Systems 82 (2016) 382–391

Fig. 2. Geomagnetically induced currents – image courtesy Wikipedia.

frequency of GICs is actually governed by the time variation of the electric field. The largest magnetic field variations are observed at higher magnetic latitudes on Earth. For this reason, even though GICs are regularly measured at many different locations, large-scale GICs (attributed to the intensity of the auroral electrojet current) have been regularly measured in Canadian, Finnish and Scandinavian power grids and pipelines [8]. GICs have also been recorded at middle latitudes during major storms. Recent work by Ngwira et al. [9] suggests that power networks along the equatorial electrojet zone could be more prone to GICs than those in middle latitudes for instance. GICs of the order of tens to hundreds of amperes are possible. In power grids, GICs are especially hazardous, as large GICs can cause failures and, if left unchecked, permanent failure of highvoltage power transformers. Power distribution networks and GICs Modern power grids typically consist of generating plants inter-connected by electrical circuits that operate at fixed transmission voltages controlled at substations. Transmission systems utilize higher voltages (and correspondingly lower currents) with the lower line resistances to reduce transmission losses over longer and longer path lengths. However, these low line resistances make transmission lines susceptible to the effects of GICs. In this regard, the power lines act like VLF antennas. The transmission grid picks up the induced currents created by the solar storm and fully absorbs them into the power grid. These excess unregulated currents have the power to melt the copper windings of the electrically stressed transformers in a typical power grid [10]. A transformer is a static electrical device that transfers energy and alters voltage/current levels by inductive coupling between its winding circuits. In essence, a varying current in the input, or primary, winding creates a varying magnetic flux in the transformers core and thus a varying magnetic flux through the output, or secondary, winding. This varying magnetic flux induces a varying electromotive force (EMF) or voltage in the secondary winding (Fig. 3). This magnetic circuit in modern-day terrestrial power transformers is disrupted by these quasi-DC GICs – the field produced by the GICs offset the operating point of the magnetic circuit and the transformer may go into half-cycle saturation. This produces harmonics to the AC waveform, localized heating and can lead to high reactive power demands, inefficient power transmission and

Fig. 3. Typical transformer – image courtesy Wikipedia.

potential failure of protective measures. As a result, during severe solar storms, these high-voltage AC (HVAC) transformers are subject to catastrophic failure. Because these transformers have very long commissioning and repair schedules, their failure poses a catastrophic risk to the continued transmission of electrical power. In the early hours of 13 March 1989, a severe geomagnetic storm struck Canada, with resulting GICs causing the collapse of the Hydro-Quebec power grid in a matter of seconds as equipment protective relays tripped in a cascading sequence of events. Six million people were left without power for nine hours, with significant associated economic loss [11]. The cost of material damages to Hydro Quebec equipment from over-voltages caused by line openings amounted to 6.5 million Canadian dollars. The net cost of the power outage to Hydro Quebec was tagged at 13.2 million Canadian dollars. During this same geomagnetic storm incident, Public Service Electric and Gas (PSE&G) in New Jersey also suffered serious damage to a bank of single-phase generator step-up transformers at the Salem Nuclear Generating Station. However, the damage was not discovered until 24 March when high levels of dissolved gases representative of internal damage was found in samples of the transformer oil. Subsequent removal and inspection of the transformer revealed severe transformer damage and conductor overheating [12]. The cost to PSE&G of replacing this transformer was on the order of several million U.S. dollars. The cost of replacement energy during this time was approximately $400,000 a day for 6 weeks, or $16.8 million. The net cost of the incident for PSE&G was therefore over $20 million [13]. Nowadays, according to a study commissioned by the American National Research Council and completed by the Committee on the Societal and Economic Impacts of Severe Space Weather Events, a geomagnetic storm with a strength comparative to that of 1859 would result in at least 130 million people without electrical power and 350 broken HVAC transformers [4]. The overall cost of restoring the grid to its original functionality and the economic damage caused by the disruption would be around 2 Trillion Dollars ($2,000,000,000,000). A number of studies in the literature have highlighted the effects of GICs on terrestrial power grids under different conditions and geographical locations. In [10], for example, the authors identify a close relationship between the theoretical calculation of GICs in a large network, practical measurements of same, the results of dissolved gas analysis (DGA) records and damage in recently failed

M. Johnson et al. / Electrical Power and Energy Systems 82 (2016) 382–391

transformers in Southern Africa. Together, the authors show how these indicate that GICs may contribute significantly to transformer failures on large transmission systems in mid-latitude regions, where GICs were here-to-fore thought not to be significant. In [14], the authors address the question of what would have happened if a powerful interplanetary event (such as the CME of 23 July 2012 considered in the paper) had been in an Earthward direction. Using validated geomagnetic storm forecast models, the paper finds that the 23–24 July event would certainly have produced a geomagnetic storm that was comparable to the largest events of the twentieth century, demonstrating that extreme space weather conditions can happen even during modest solar activity cycles. [15] reports on a study for GIC measurements in Brazil conducted under a cooperative project between FURNAS (the Brazilian electric power company) and the National Institute for Space Research in Brazil. First-hand measurements and results are presented for GIC measurements during large geomagnetic storms at these lower latitudes and comparisons drawn with equivalent measurements at higher latitudes on the globe. Much work has been carried out to-date in studying, investigating and quantifying the space weather phenomenon and associated GICs which affect our planet. In [16] the authors highlight the fact that extreme geomagnetic event characterization is of fundamental importance for quantifying the technological impacts and societal consequences of extreme space weather, GICs, etc. In the article, they report on the global behavior of the horizontal geomagnetic field and the induced geoelectric field fluctuations during severe or extreme geomagnetic events. This includes an investigation of the latitude threshold boundary, the local time dependency of the maximum induced geoelectric field and the influence of the equatorial electrojet (EEJ) current on the occurrence of enhanced induced geoelectric fields over ground stations located near the dip equator. In [9], the study introduces a specially adapted model for simulating extreme space weather comparable to the Carrington superstorm of September 1859. In recent years, significant progress has been made toward the first-principles modeling of space weather events using three-dimensional (3D) global magnetohydrodynamics (MHD) models. 3D MHD models are playing a critical role in advancing the understanding of space weather. However, the modeling of extreme space weather events is still a major challenge even for modern global MHD models. The paper describes a specially adapted University of Michigan 3D global MHD model for simulating extreme space weather events with a Dst footprint comparable to the Carrington superstorm. In [17], GICs are modeled using a series of 100-year extreme geoelectric field and GIC scenarios, taking into account the key geophysical factors associated with the geomagnetic induction process. In particular, the paper derives explicit geoelectric field temporal profiles as a function of ground conductivity structures and geomagnetic latitudes. The computed extreme GIC scenarios may be used in further engineering analysis that will be needed in order to quantify the geomagnetic storm impact on conductor systems such as high-voltage power transmission systems considered in this paper.

385

Traditional GIC mitigation Traditional solutions to GIC in transformers include maintenance schedule changes, additional on-demand generating capacity, and ultimately, load shedding. These options are expensive and sometimes impractical for the current network configuration, and so are rarely adapted in their entirety. As recently as 1989, the specified procedure for dealing with induced disturbances in power transmission system, as outlined in [18], was through conservative operating procedures such as:    

Reduce the output of generating units to around 80% of full load Enable transformer tripping Remove problematic equipment Adjust loading on HVDC circuits to be within the 40–90% range of the nominal rating  Reduce loading on other critical transmission lines to 90%, or less, of their nominal rating. These guidelines are reactive at best, and are not specific enough to protect against disturbance such as GICs. All this changed in March 1989, when the province of Quebec in Canada suffered an almost complete electrical (power grid) blackout during one severe geomagnetic storm. Millions of Hydro-Québec customers were left without electricity for several hours [19]. After this incident, most electrical power grids were ‘‘strengthened”, or made more robust to induced currents. For example, HydroQuébec itself implemented the following measures, as described in [20]:  Recalibration of protection systems and raising of tripping levels.  Establishment of a real-time alert system.  Modification of power system operating procedures.  Installation of series compensation on power lines to enhance grid stability. After these steps had been implemented across the grid, according to the Canadian organization ‘‘there have been very intense magnetic storms since 1989 but they have not caused any problems”. In addition to improvements in overall power network topology and procedures, other steps are proposed to reduce the severity of GICs: including the measurement and monitoring of voltage phase angle [21] and devices similar to the neutral blocking devices described later in this section, as described in [22]. A complimentary approach traditionally adopted to reduce the damage caused by GICs is the inclusion of sufficient ‘‘grounding” points to allow the induced currents non-destructive paths to ground from the transmission lines. While this strategy lessens the chances of GICs being carried to critical equipment, complete protection is not guaranteed. Typical grounding requirements are in the order of:  50 km intervals in the latitude [North–South] domain,  75 km intervals in the longitude [East–West] domain,  longer intervals than 60 km closer to the tropics.

Mitigation strategies With the ever-increasing importance of electricity and electrical power to the modern age, the question of transmission line protection from the effects of GICs is becoming ever-more important and topical. This section outlines some of the potential solutions, new and existing proposals, to the problem of GICs in power grids. These solutions vary from operating procedures to forwardlooking solutions considering space-based power grids and other options.

Even though the risks to power grids from solar events are reduced using these methods, power companies are still developing and implementing mitigation strategies. In fact, in a highly competitive energy field with a marginally profitable market, power companies must increase the stability and reliability of their grids to guarantee supply and distribution of electricity under all conditions. For example, the National Grid Company plc (NGC) in England has, since January 2000, developed an operating application to lower the impacts of GICs on transformers. This software

386

M. Johnson et al. / Electrical Power and Energy Systems 82 (2016) 382–391

program combine forecasting from the Advanced Composition Explorer (ACE) satellite and measurements on the ground to visualize GICs in transformers [23]. To illustrate the complexity of the resulting predictions, Fig. 4 shows the operator display from the program for the solar storm of April 2000, where the arrows are the computed electrojets over England and the circles are the intensity of GICs in transformers. This display allow a qualified operator to predict and alleviate potential damage to the power grid by GICs. Since the implementation of this system, no false positives have been registered and all geomagnetic storms had been weathered without incident. On the other side of the Atlantic ocean, Hydro-Québec and Alstom, electricity producer and power transmitter, are constantly updating and improving their Smart Grid technology, which was adopted in 2012 [24]. Capacitor blocking schemes GIC prevention and avoidance in power grids may broadly be categorized as either operational or infrastructure hardening in nature. Of these 2 approaches, the latter, infrastructure hardening, is more significant and effective. Such modifications to the power grid are more reliable than operational mitigation strategies, as it provides a guaranteed solution, compared with the former, which relies on the situational awareness of an operator to mitigate the problem. Capacitor blocking schemes are one of the most commonly used methods of infrastructure hardening in terrestrial transmission lines. These schemes use passive devices, such as series capacitors or (in some cases) low-ohmic resistors inserted in the neutral-to-ground connection of the power system transformer. The addition of these devices at the point where most GICs enter the power grid allows the resulting GIC to be blocked or reduced before it can damage the associated equipment. While these passive shunts offer a relatively simple means of mitigating the unwanted effects of solar weather in power girds, their installation and maintenance is usually a costly undertaking, both in monetary and operating terms. A large portion of this

Fig. 4. The visualization of the April 2000 storm on the NGC control room operator display.

expense results from the fact that series capacitors need to be applied in all three phases of the electric transmission lines in order to completely block the GIC flow in the circuit (see Fig. 5b), which is an extremely costly undertaking. Furthermore, it should be noted that these devices only mitigate the magnitude of the GICs in the transmission lines; they do not remove them. Consider, for example, the situation described in [25] – here, even with a number of series capacitors installed into an existing power network, the overall GIC levels could only be reduced by between 12% and 22%, meaning it was still vulnerable to geomagnetic storms. An alternative approach sometimes adopted is to install neutral capacitive devices in the neutral-to-ground connection, which block all DC current flow. Again, this is a costly and timeconsuming installation, with so far only a small number of neutral capacitors having been installed to protect highly susceptible power transformers. As with all of these schemes, this approach also must consider the consequence of ferro-resonance issues in the power grid and the potential risk the associated impedance changes can introduce elsewhere in the network. It is becoming clear that for transformer protection against severe geomagnetic storms, a broader solution is necessary. As opposed to using capacitor blocking schemes, the global implementation of low-ohmic resistors in the neutral-to-ground connection (see Fig. 5c) in all transformers would lower the overall GIC level by 60–70%. In comparison to the in-phase series capacitors, the neutral blocking resistors are cheaper in production, as

Fig. 5. (Image courtesy Metatech) (a) GIC flow path on single transmission line, with a neutral-grounded transformer at each end, (b) application of series capacitors in each phase blocks all GIC flow in the circuit and the transformer, and (c) low-ohmic resistors in each neutral-to-ground connection notably reduce the GIC flow in the transmission line and the transformers.

M. Johnson et al. / Electrical Power and Energy Systems 82 (2016) 382–391

they need to withstand only lower voltages and current ratings in the neutral. Moreover, the resistors do not add resistance to the AC phase line in contrast to the in-line application option. Investigations showed that the appropriate usage of low-ohmic resistors do not have disturbing effects on the grounding performance of the transformers, that means they would not infringe the IEEE guideline recommendations. However, the scope of implementation of such a scheme is truly staggering – in the U.S alone, the estimated number of transformers which would need to be equipped with resistors ranges from 3000 to 5000 with installed costs per unit expected to range between $40,000 and $100,000 [26]. There are a number of commercial power and automation companies currently offering passive blocking solutions for GIC prevention to the power industry. The Swiss company, ABB, for example, holds a leading position in power and automation technologies and is one good example of technologies using infrastructure hardening for GIC mitigation. The company has developed the ‘‘SolidGround” GIC stability system (see [27] for more information), which provides constant grounding of the transformer using a metallic path to the ground during GIC events and is effectively a transparent transmission path under normal operating conditions. The system overall adds only a very small AC impedance path during GIC protective mode. Within the SolidGround system, the network is composed of a capacitor bank, an AC grounding switch, sensing and control electronics, a DC disconnecter and a resistor. After the detection of GIC flow in the system the AC and DC switches are automatically opened, enabling the DC currents to leave the transformer ground through the resistor and capacitor bank, blocking the quasi-DC currents and maintaining the continuous AC grounding path (see Fig. 6). The sensing and control electronics allow site-specific customization through user-configured thresholds and offer the choice between automatic or manual mode. The DC disconnecter will break the GIC flow as soon as the chosen threshold is crossed. In conclusion, infrastructure hardening is one of the most common and widely used strategies for GIC mitigation, allowing the additional option of combination with operational hardening. With so many available options and configurations within this solution alone, it is a challenging task for the power grid companies to find the most reliable and least expensive option for their particular scenario in this large pool of mitigation options. Space weather forecasting Space weather forecasting of major solar and geomagnetic storms allow for mitigation strategies to be implemented a priori.

387

Such forecasting services are usually composed of 2 distinct components: the space-based satellites which detect changes in the solar wind signalling an event and the forecasting programmes or software which analyse the satellite data for these indications. Typically, it takes approximately 2–3 days after a CME on the Sun for the associated solar wind to reach Earth and to affect the Earths geomagnetic field. An example of the former is ACE, or the Advanced Composition Explorer, described in [28]. The detection of the bow/shock that precedes the CME-induced solar wind by spacecraft such as ACE, located at the L1 Lagrangian point in space, gives a definite 20–60 min warning of a geomagnetic storm (again depending on local solar wind speed). Examples of space weather forecasting programmes currently being used for GIC prevention include Solar Shield, Solar Terrestrial Dispatch (STD) and European Risk from Geomagnetically Induced Currents (EURISGIC). With a timely warning power companies are able to protect the electric grid by shutting it down or reducing the power load on the system. The Solar Shield project ([29]) is a two-level experimental forecasting tool used in the Community Coordinated Modeling Center (CCMC) at NASAs Goddard Space Flight Centre since February 2008. From NASAs SOHO and STEREO satellites’ data a 3D model is generated, cone model parameters are derived and arrival predictions are made. In this first level a probabilistic estimate for endangered individual power grid nodes is provided. As soon as the CME is passing ACE, the data from the in-situ measurements of the CMEs magnetic field, such as its density and speed, is used as boundary conditions for the magnetohydrodynamic (MHD) model. Finally, with this second-level model the GIC forecast for individual nodes of the North American power grid can be completed and alerts are sent out. Further information is provided on the website at http://ccmc.gsfc.nasa.gov/Solar_Shield/Solar_ Shield.html. In [30], the Solar Terrestrial Dispatch (STD) space forecast center is presented. The STD center is the primary provider for the Northeast Power Coordinating Council (NPCC), responsible for a sensitive region of power supply in North America. The provision of fast updated real-time space weather predictions and timecritical warnings of possibly dangerous geomagnetic storms 24/7 are their primary jurisdiction. The software used covers all control areas in a coordinated way to ensure reliable and timely delivery of alerts and predictions of potentially hazardous space weather to the power industry. The system is inherently effective, since it is based on a five layered redundancy set-up, undergoes quality controls and is resistant to external stimuli, i.e. for instance that power blackouts do not affect their service to the customers. Oler analyzed the performance of five space weather forecasting services

Fig. 6. Concept of SolidGround protection system against GICs.

388

M. Johnson et al. / Electrical Power and Energy Systems 82 (2016) 382–391

in terms of how they handled the five strongest CME events in October/November 2003. STD performed best among those five space weather forecasting centers. It showed notably lower errors in their predictions and the shortest ‘maximum alert lead time’, which is defined as the time period between the CME passing ACE and the warning being send out. This achievement can be attributed to a prudent selection of model inputs from the available data and not only to an extremely precise modeling method. Concerning public outreach, STD spreads its forecasting information via email lists and websites such as spacew.com. The European Risk from Geomagnetically Induced Currents (EURISGIC) project, funded by the Seventh Framework Programme of the European Union has been working on the mitigation of the hazards from GICs from March 2011 until February 2014. The consortium consists of the Finnish Meteorological Institute holding the coordinating function, further European Union partners from Hungary, Sweden and the United Kingdom as well as international collaboration with Russia and the United States. The goal of the project is to produce a first Europe-wide real-time prototype forecast service of GICs in high-voltage power systems, based on in situ solar wind observations by ACE, empirical modeling, and thorough simulations of Earths magnetosphere. Moreover, the first map of statistical risk of large GICs throughout Europe will be derived from long-term recorded geomagnetic data. Currently, there is only one European magnetohydrodynamic code globally simulating the solar wind-magnetosphere interaction called Grand Unified Magnetosphere-Ionosphere Coupling Simulation (GUMICS-4), which will be upgraded within the EURISGIC project in order to provide real-time forecasts. Furthermore, the project accepted the challenge of compiling a simplified conductivity map of Europe facilitating the key problem of GIC computation. Models of Earths conductivity are essential for calculating the geoelectric field driving GICs through the grid system. EURISGICs contribution to the mitigation of GICs is composed of early warning forecasts, risk maps and worst-case-scenario assessments. Further information is provided on their website at see http://www.eurisgic.eu [31]. Another essential element in space weather forecasting is the comparison between model predictions and observed values in the shortest time possible, which is supported by close collaboration between researchers and forecasters. One example here is the recently developed forecasting tool of the U.S. National Oceanic and Atmospheric Administrations (NOAA) Space Weather Prediction Center (SWPC). It consists of the combination of a new observation platform (STEREO – Solar Terrestrial Relations Observatory), a visualization and specification tool (Geometric Localization and the CME Analysis Tool CAT) and a numerical, magnetohydrodynamic model (Wang–Sheely–Arge-Enlil Model). Of especial note here is the knowledge-based education and training provided as part of the system. Going forward, this could prove to be just as crucial as the hardware that is used in the system, described in [32]. Even with all of these advanced systems, there are still a number of gaps in predicting geomagnetic storms. Future work currently being undertaken will deal with the ability to not only forecast the arrival time, but also the magnitude of the ensuing geomagnetic storm, the creation of geospace respectively regional geospace models and the nowcast/forecast of the radiation environment in interplanetary space (including Low, Medium and High Earth Orbit). Redundant and back-up transformers The catastrophic failure of the megatransformers in traditional power girds can be avoided by using simple voltage division and current division principals implemented in the form of redundant arrays of smaller HVAC transformers. Distributed generation

allows collection of energy from many sources and may give lower environmental impacts and improved security of supply. Generally the transformer redundancy rule is recognised as being:  Small-odd-number-under-eleven = Number of Redundant Transformers.  Generally the transformer redundancy rule should be [smallodd-number-under-eleven 1] = Number of Redundant Transformers in Use (at any one time to backup the Mega-HVAC transformer).  Potentially up to 13 HVAC transformers could be used in parallel – but the buffering and cophasing networks as well as safety considerations might make such a system too complex.  There are personnel safety system issues involved with using an array of more than 11 transformers.  The excess redundancy of an array of 13 HVAC transformers is only recommended for the most northerly locations or extreme engineering conditions.  Each HVAC transformer must have its own accompanying buffering and matching network to terminate and match its output.  It is not fully clear exactly how this HVAC transformer network should be designed. There are at least 400 different Mega-HVAC transformer installations – and each may be locally unique. Let it be said that the required AC (matching/ cophasing) networks are not design impossibilities for this kind of application.  In principal as well as in practice: It is not recommended risking going below 5 redundant HVAC transformers, with 7 being the nominal recommendation.  The ultimate number of redundant transformers must ultimately depend on the static and dynamic load factors for the Mega-HVAC transformer.  Overall recommendations on redundancy: an array of 11 to 7 transformers is nominal. Specific Recommendations for selecting redundant and back-up transformers include:  Each Mega-HVAC transformer must have [as a backup system] an array of at least 7 HVAC transformers (6 in use at any one time) that can be switched on as a backup system at any time.  At least one redundant HVAC transformer array element must always be in ‘repair or maintenance mode’ or ‘storm buffering mode set aside’.  Each redundant HVAC transformer must be rated at 1/6th the combined Mega-HVAC (input/output) parameters, where an array of 7 transformers exist.  The individual HVAC redundant transformer array ratings should be 133–166% of the [(1/6th)  (Mega-HVAC-ratings)]. A solar storm may happen when a redundant transformer is set aside for repair, thus the 1/6th not 1/7th constant. Due to the high costs and associated installation downtime(s) for such a system, however, this option is not a popular choice from a business standpoint. Underground HVDC power grid The ever-increasing need for electricity in modern-day society creates a corresponding need for more electricity and more associated transmission lines within the power grid. Adding new transmission lines to the existing power grid network provides an opportunity to use new technology such as ‘‘Elpipes” [33] to address some of the potential problems inherent in traditional networks, such as the induced GICs described in this paper. Elpipes are

M. Johnson et al. / Electrical Power and Energy Systems 82 (2016) 382–391

389

insulated conductors installed in steel underground pipes, as shown in Fig. 7. Compared to standard or traditional overhead transmission lines, Elpipes have the following advantages:  Higher capacity: 12 GW compared to 0.2 GW.  Resistant to extreme weather events.  Can be made more resistant to fires, accidents and terrorism. Elpipes grid have less grounding points and are designed for high voltage direct current HVDC, thus making it easier to mitigate the effects of space weather. GIC disturbances only affect AC transmission systems – DC-based systems are inherently immune to GICs. With less grounding points, the resulting network would still need less GIC mitigation, even if it is transmitting AC electricity. The implementation of a redundant Elpipe grid (as shown in Fig. 8) would reinforce current AC grids by making them more stable and able to recover faster from any kind of system disturbance, GIC induced or otherwise (see Fig. 9). Finally, for similar voltage, 800 kVDC, Elpipe lines will cost 0.57 $/kW km for a capacity of 24 GW compared to 0.31 $/kW km for 6 GW. All the advantages listed above and especially stability and resistance to space weather events make Elpipes one of the preferred option for new and future transmission lines going forward.

Fig. 8. Redundant North American East Coast Elpipes grid – map courtesy of Power World Corporation.

Space power grids Space offers a clean, steady and reliable source of power. One square kilometer in space provides 500 MW of electrical power on Earth (conversion losses included), which is the equivalent power output of a nuclear reactor. Several ideas for developing a system providing mankind with space solar power (SSP) already exist, such as placing large collector/converter satellites in Geosynchronous Earth orbit (GEO), the ‘Lunar Space Power System’ [34], the ‘Sun Tower’ [35] or the ‘Solar Disc’ [35]. Regions with poor transmission infrastructure (islands, mountains) and disaster hit areas with mobile receivers are just 2 example areas which could benefit from these systems. A space power grid (SPG) would offer advantages over existing terrestrial power grids, including the inclusion of excess power from green generation plants and the easy distribution and delivery of electrical power to diverse regions at peak demand times. So far, the main (overwhelming) obstacle to the introduction of such a system has been the associated cost approximately $300B would be required in order to place the necessary satellites in their orbits and build the corresponding receivers/ground stations on Earth. However, a more affordable alternative exists in a proposal from Komerath [36], which proposes the viable establishment of a SPG in three distinct phases. The first phase is based on power exchange only and will be implemented over a 10 year period. 36

Fig. 7. HVDC Elpipes concept.

Fig. 9. Proposed SuperGrid proposal – image courtesy Wikipedia.

microwave reflector antenna satellites (SPG satellites) in low/mid Earth orbits (LEO/MEO) will redirect the power of 100 renewable-energy plants in form of microwave beams (baseline regime 200 MHz) taking advantage of time-zones, day-night and climate differences. The redirection of beams is not limited between satellite and ground station, but also allows transmission of power between the satellites. Estimates predict the break-even point for this investment of $4B to be reached after 10–15 years. In the second phase, after 20 years of operation, the aged SPG satellites are gradually replaced by augmented satellites (A-SPG) carrying solar collectors and sunlight-to-microwave power converters. This phase foresees 96 A-SPG satellites and 268 ground stations. The upgrade adds a small amount of power (8.2% of SPG phase) to the already functioning grid. This system, once established, is self-supporting. There are four technologies promising a higher power contribution from the extended satellite constellation: large area photovoltaic collectors, direct solar conversion to

390

M. Johnson et al. / Electrical Power and Energy Systems 82 (2016) 382–391

lasers, solar pumped masers and the usage of an optical rectenna (rectifying antenna) directly converting sunlight to beamed microwave (feasible by 2035 with an efficiency of 50%). The expenses of about $6B for the second phase are estimated to recover within 21 years. Phase three is composed of launching 96 very large yet ultralight collector/reflector satellites to high orbits. Their purpose is to focus sunlight on the phase two A-SPG satellites in LEO, which in turn convert the visible light to a microwave beam sent down through the atmosphere to ground stations. Since the high altitude satellites do not carry converters their mass and thus their launching costs are reduced. Those collectors have a diameter of 3 km increasing the factor of space power by 100. As soon as the collector/reflector satellites are in place power costs are expected to drop and profits to rise. Note, that the assumption was made that only 30% of the power sent to the A-SPG system gets to the customer as useful electric power. Within 40 years a Space Solar Power system with its corresponding ground infrastructure can be realized at reasonable and recoverable costs. One crucial requirement is international collaboration. Komeraths proposal is based on the model of the European Space Agency, namely a global public–private Consortium. International collaboration shall lower the risk and make low interest funding possible. The proposed SPG is an exciting departure in power grid technology, however note that it will still be vulnerable to geomagnetic storms and GICs. No one-hundred per cent effective satellite protection system against charged particles has been developed todate and from the ground stations (receivers) the power will still be distributed via some form of terrestrial power grid. If the ground stations were to switch to a DC-based distribution network, though, GIC issues would be a redundant worry, as such a network is inherently GIC-proof. Another possibility in this section could be the development of an underground distribution system consisting of collectors/reflec tors/converters using the same concepts as those outlined in the SPG. Power generated in space can be redirected on Earth using beams of a certain wavelength instead of transmission lines and transformers. Further analysis would be required in order to measure such a systems vulnerability to geomagnetic storms. Robust smart power grids Various planned and proposed systems to dramatically increase transmission capacity are known as super, or mega grids. A recent proposal by Transcanada priced a 1600-km, 3-GW HVDC line at $3 billion USD and would require a corridor 60 m wide. In India, a recent 6 GW, 1850-km proposal was priced at $790 million and would require a 69 m wide right of way. With 750 GW of new HVDC transmission capacity required for a European super grid, the land and money needed for new transmission lines would be considerable. As mentioned above, the electrical grid is expected to evolve to a new grid paradigm–the smart grid, an enhancement of the 20th century electrical grid. Traditional electrical grids are generally used to carry power from a few central generators to a large number of users or customers. In contrast, the new emerging smart grid uses two-way flows of electricity and information to create an automated and distributed advanced energy delivery network. Current research is mainly focused on three systems in smart grid- the infrastructure system, the management system, and the protection system. Smart grid infrastructure differs from a traditional power distribution grid in the addition of digital communication technology used to establish a two-way communication link between the power utility and residential or commercial consumers. A smart

grid is composed of transmission lines, substations, transformers and consumer power meters each generating data and reporting to a central command center. This technology enables more efficient transmission of power across the complex distribution networks. Additionally, the digital communication technology built into the smart grid will allow the utility to carefully manage the paths of power distribution. This can be used to predictively divert power from susceptible portions of the network before disturbances occur and thereafter be used to restore electricity quickly after such disturbances disrupt regular power distribution. This reduces management costs and affords the flexibility to integrate renewable energy systems and quickly adjust the network for peak power demand. Finally, smart grids which are dependent on domestically produced energy sources and integrate renewable energy sources are more environmentally friendly and robust than a traditional grid. At present, Transitioning to a smart grid from the more traditional grid will involve upgrading the aging power distribution infrastructure of the affected countries. In addition to the intermittent power outages associated with any such large-scale upgrade, during a smart power grid upgrade the affected grid(s) would also need to be protected from additional hazards such as threats posed by natural disasters or indeed foreign attack during the upgrade process. A robust smart grid uses the same technology as that employed in implementing smart power grids, namely interconnected nodes communicating their status back to a central command node. In addition to power management and load balancing, however, the robust power grid would use such communication channels to reduce the harmful effect of GICs on the network. Thus, a more resilient and balanced power network is achieved from the combination of two complimentary technologies, robust networks and smart grids. Such a network, however, is not without cost, most of which is in the initial large-scale investment in the network topology and infrastructure. Robust smart grids, by their very nature, would explicitly prioritize total system resilience, preservation of critical equipment or infrastructure, and modularity within the overall design. Implementation of such a robust smart grid would also directly incorporate space weather forecasting in automated processes to monitor space weather disturbances and reduce current in susceptible lines. Additional measures, based on historical data and regional GIC activity, can also be taken during the design of robust smart grids to create margins of safety or tolerance to severe space weather. A key feature of robust smart grid technology is the prediction and anticipation of GICs by using forecasting information coming from different sources like NOAAs Space Weather Prediction Center and real time information from resources such as the Advanced Composition Explorer (ACE), a satellite that monitors solar wind. Such predictive measures would allow the ability to prepare the grid for imminent GICs instead of reacting after the damage has occurred. Space weather prediction of such events would automatically initiate smart grid procedures specifically tailored to handle solar storm effects and GIC mitigation. Such preemptive action would obviously decrease the chances and extent of power loss and damage during such an event.

Conclusion This paper has presented a review of mitigation technologies and techniques for terrestrial electrical power grids when dealing with geomagnetically induced currents and their associated destructive effects on power grid technologies. GICs are primarily induced by magnetic activity in the upper atmosphere of our pla-

M. Johnson et al. / Electrical Power and Energy Systems 82 (2016) 382–391

net, which in turn is excited by increased solar wind activity and space weather effects resulting from increased activity on the surface of the Sun. These GICs can cause overloads and result in permanent damage to the high-voltage transformers which are inherent components of many modern terrestrial power grids. Mitigation techniques and strategies for terrestrial power grids range from low impact changes to the existing power grid to replacement grid topologies. Low impact alterations include schemes such as maintenance schedule changes, additional ondemand generating capacity, capacitor blocking schemes and load shedding for existing terrestrial grids. The importance of space weather forecasting and its usefulness as a tool for predicting potentially hazardous levels of GICs is also highlighted. More substantial alterations to the power grid to address these dangers would include the introduction of multiple grounding points within the network, and the use of backup/redundant transformers in the system. Longer term solutions include considering alternative power grid systems and topologies, from DC-based power delivery systems to space-based (satellite) power generation and transmission. Finally, the concept of robust smart grids as a future alternative and solution to the GIC problem is presented. Acknowledgements The authors would like to acknowledge the work of all team members in Team Project SolarMax at the 2013 International Space University Space Studies Program 2013 for their contributions towards the team project and this work. Team SolarMAX was composed of Ang Xu, Anja Schuster, Arnaud Sternchass, Ashley Dale, Bai Baocun, Beatrice Hainaut, Caroline Smoczarski, Chandrakanta Kumar, Charles Laing, Chunhui Wang, Eric Hall, Gabriele Librandi, George Gorospe, Gongyou Wu, Hester Vermeiden, Hongbin Shi, Jaime Babb, Jonathan Landry, Julio Ceasar Salazar Ospina, Kun Li, Leo Teeney, Mark Burke, Matt Palmer, Meifang Li, Melissa Battler, Michael Johnson, Morten Salvesen, Nicolas Thiry, Paul Tarantino, Remco Timmermans, Richard Passmore, Suquan Ding, Timo Nikkanen, Xianxu Yuan, Yevgeny Tsodikovich, Yuta Nakajima. Team SolarMAX would also like to thank David Haslam, Dr. Rogan Shimmin and Dr. Pete S. Worden for their leadership and guidance, and would like to acknowledge both the International Space University and Johns Hopkins University Applied Physics Laboratory for supporting this project. References [1] Pelton JN. Serious threats from outer space. In: Space debris and other threats from outer space. Springer; 2013. p. 1–15. [2] Farrow J, Peeters W. International Space University (ISU). Strasbourg, France. [3] Angelopoulos V, Runov A, Zhou X-Z, Turner DL, Kiehas SA, Li S-S, et al. Electromagnetic energy conversion at reconnection fronts. Science 2013;341 (6153):1478–82. http://dx.doi.org/10.1126/science.1236992. , . [4] C. on the Societal, N.R.C. Economic Impacts of Severe Space Weather Events: A Workshop, Severe Space Weather Events–Understanding Societal and Economic Impacts: A Workshop Report. The National Academies Press; 2008. [5] Cliver E. The 1859 space weather event: then and now. Adv Space Res 2006;38 (2):119–29. http://dx.doi.org/10.1016/j.asr.2005.07.077 [The Great Historical Geomagnetic Storm of 1859: A Modern Look] . [6] Bhargava B. Railway electrification systems and configurations. IEEE power engineering society summer meeting, 1999, vol. 1. IEEE; 1999. p. 445–50.

391

[7] Asplund G. Application of HVDC light to power system enhancement. IEEE power engineering society winter meeting, 2000, vol. 4. IEEE; 2000. p. 2498–503. [8] Pirjola R. Effects of space weather on high-latitude ground systems. Adv Space Res 2005;36(12):2231–40. http://dx.doi.org/10.1016/j.asr.2003.04.074 [Space Weather] . [9] Ngwira CM, Pulkkinen A, Kuznetsova MM, Glocer A. Modeling extreme ‘‘carrington-type” space weather events using three-dimensional global MHD simulations. J Geophys Res: Space Phys 2014;119(6):4456–74. [10] Gaunt C, Coetzee G. Transformer failures in regions incorrectly considered to have low gic-risk. In: 2007 IEEE Lausanne Power Tech, 2007. [11] Bolduc L. Gic observations and studies in the hydro-québec power system. J Atmos Solar Terr Phys 2002;64(16):1793–802. [12] Fallon DJ, Balma PM, McNutt W. The destructive effects of geomagnetic induced currents in power transformers. Doble Engineering Company; 1990. [13] Barnes P, Van Dyke J. Economic consequences of geomagnetic storms (a summary). IEEE Power Eng Rev 1990;10(11):3–4. [14] Baker D, Li X, Pulkkinen A, Ngwira C, Mays M, Galvin A, et al. A major solar eruptive event in july 2012: defining extreme space weather scenarios. Space Weather 2013;11(10):585–91. [15] Trivedi NB, Vitorello Í, Kabata W, Dutra SL, Padilha AL, Bologna MS, et al. Geomagnetically induced currents in an electric power transmission system at low latitudes in brazil: a case study. Space Weather 2007;5(4). [16] Ngwira CM, Pulkkinen A, Wilder FD, Crowley G. Extended study of extreme geoelectric field event scenarios for geomagnetically induced current applications. Space Weather 2013;11(3):121–31. [17] Pulkkinen A, Bernabeu E, Eichner J, Beggan C, Thomson A. Generation of 100year geomagnetically induced current scenarios. Space Weather 2012;10(4). [18] Albertson V, Bozoki B, Feero W, Kappenman J, Larsen E, Nordell D, et al. Geomagnetic disturbance effects on power systems. IEEE Trans Power Deliv 1993;8(3):1206–16. [19] Kappenman JG. Geomagnetic storms and their impact on power systems. IEEE Power Eng Rev 1996;16(5):5. [20] Hydro-Québec. Understanding electricity— march 1989 blackout— hydroquébec (Consulted in April 28th 2014). [21] Beland J, Small K. Space weather effects on power transmission systems: the cases of hydro-québec and transpower new zealandltd. In: Effects of space weather on technology infrastructure. Springer; 2005. p. 287–99. [22] Kappenman J, Norr S, Sweezy G, Carlson D, Albertson V, Harder J, et al. GIC mitigation: a neutral blocking/bypass device to prevent the flow of GIC in power systems. IEEE Trans Power Deliv 1991;6(3):1271–81. [23] Erinmez IA, Kappenman JG, Radasky WA. Management of the geomagnetically induced current risks on the national grid company’s electric power transmission system. J Atmos Solar Terr Phys 2002;64(5):743–56. [24] Alstom. Alstom grid and hydro-québec achieve new milestone in their smart grid technology innovation partnership (Consulted in April 28th 2014). [25] Samuelsson O. Geomagnetic disturbances and their impact on power systems. [26] Kappenman J. Low-frequency protection concepts for the electric power grid: geomagnetically induced current (GIC) and e3 hemp mitigation. Rep Meta-R 2010;322:94. [27] Rajapakse A, Perera N, Faxvog F, Jensen W, Nordling G, Fuchs G, et al. Power grid stability protection against GIC using a capacitive grounding circuit. In: 2012 IEEE PES transmission and distribution conference and exposition (T&D). IEEE; 2012. p. 1–6. [28] Murdin P. Ace (advanced composition explorer). Encyclopedia Astron Astrophys 2000;1:4518. [29] Pulkkinen A, Hesse M, Habib S, Van der Zel L, Damsky B, Policelli F, et al. Solar shield: forecasting and mitigating space weather effects on high-voltage power transmission systems. Natural Hazards 2010;53(2):333–45. [30] Oler C. Prediction performance of space weather forecast centers following the extreme events of october and november 2003. Space Weather 2004;2(8). [31] Viljanen A. European project to improve models of geomagnetically induced currents. Space Weather 2011;9(7):S07007. [32] Steenburgh R, Biesecker D, Millward G. From predicting solar activity to forecasting space weather: practical examples of research-to-operations and operations-to-research. Sol Phys 2013:1–16. [33] Faulkner RW. Underground HVDC transmission via elpipes for grid security. In: 2012 IEEE conference on technologies for homeland security (HST). IEEE; 2012. p. 359–64. [34] Criswell D. Lunar space power system for energy prosperity within the 21st century; 2005. [35] Mankins JC. A fresh look at space solar power: new architectures, concepts and technologies. Acta Astronaut 1997;41(4):347–59. [36] Komerath N, Boechler N. The space power grid. In: Proceedings of IAC.