A Cost-Effective Plan to Harden Existing Facilities

This article provides an extension of my article in the June 2021 issue of In Compliance Magazine, describing the different ways to protect power system electronics in high-voltage power control houses found in HV substations [1]. The intention here is to provide a specific plan to start to harden power grids against the fields produced by high-altitude electromagnetic pulse (HEMP) and intentional electromagnetic interference (IEMI). In addition, we will discuss the differences in protecting power company substation control houses and control centers and even power generation stations against these threats. Finally, there will be a discussion of the approach to protect the high voltage transformers (V ≥ 100 kV) against the late-time portion (E3) of the HEMP, which also will provide protection against an extreme geomagnetic storm if it were to occur.

While the worst-case levels of the early-time (E1) HEMP environment have not changed, this is not the case for the late-time (E3) HEMP environment, due to the work of the U.S. EMP Commission [2]. The worst-case level of E3 HEMP has doubled, and the IEC is in the process of increasing the worst-case level in IEC 61000-2-9 Ed.2 draft [3]. While this increase is significant, the same new draft version of IEC 61000-2-9 also discusses the fact that the worst-case E1 HEMP field occupies a very limited portion of the ground exposure. And, when considering that there are over 9000 high voltage substations in the U.S., they all cannot be illuminated at the worst-case E1 peak HEMP level with a single high-altitude burst. Also given the costs of hardening a large number of buildings, there have been discussions in the IEC and in other standards organizations considering resilience aspects to reduce the cost burden of protection [4].

Figure 1 presents the draft versions of the worst-case HEMP time waveforms, including the new version of E3 HEMP. In the standard, the actual “incident” E3 magnetic field is provided, along with the method to compute the electric field depending on the earth’s deep conductivity. This accounts for the substantial variation of ground conductivities in many places of the world including the U.S. In the new standard, it is not assumed that the E3 HEMP electric field is the same everywhere and, in many places, could be more than a factor of 10 lower.

As this article will also discuss the additional protection needed for IEMI, Figure 2 describes the most recent presentation of the relationship of the electromagnetic fields in the frequency domain that can cause IEMI relative to E1 HEMP, lightning electromagnetic pulse and also standard levels of EM fields associated with EMC [1].

This article first discusses (in Section 2) the basic problem of hardening a large number of critical buildings to protect their electronics and then looks at the various options for protection. The issue of replacing existing buildings is also discussed. The role for high-level EM protection, such as recommended in MIL‑STD-188-125-1, is also mentioned.

Section 3 of this article discusses the method to determine the level of hardening of buildings depending on the EMC requirements that are necessary to operate normally. Also, the variability of the incident environments is discussed along with the idea of considering resiliency.

Section 4 presents the best hardening approach for existing buildings for E1 HEMP and IEMI, while Section 5 discusses the best approach for protecting the large transformers that can be affected by E3 HEMP. Section 6 describes the rationale for developing a hardening program over time. Section 7 provides a summary and recommendations.

It should be noted that due to the extremely large amount of material to be covered here, this article will rely strongly on references to provide the details, as we cannot cover all of the hardening techniques in a single article. Most of the references are IEC standards or peer-reviewed publications from the IEEE EMC Society.

Dealing with Improving EM Hardness of Existing Buildings

As mentioned in the introduction, most large power companies in the U.S. and worldwide have several hundred (or more) high-voltage substations connected to a control center. They also have an even larger number of distribution substations, although each of them controls much less power than a single high-voltage substation. The problem in terms of protecting substation control houses is that the threats of HEMP and IEMI are high impact, low probability (HILP) threats (HEMP has not occurred anywhere in the world since the 1962 tests by the U.S. and the Soviet Union – although the capability to detonate a high altitude burst clearly exists today).

As is clear to consumers over the past 5 years, it seems that the rates one pays for electricity are increasing, and power companies are not in a position to spend even more money on their existing infrastructure, when they are planning for increases in their overall grids due to shifts toward electrical cars and renewable power sources.

The best solution is to improve the hardness of the existing buildings by upgrading the protection of the best existing buildings, and to use this design for new substation control buildings, if needed. In addition, due to the criticality of particular substations (depending on their location and their service area), some of the existing buildings can be upgraded over time. For substation control houses, what should be the approach to evaluate methods to upgrade the hardening to HEMP and IEMI? Let us examine Figure 3, which describes the basic substation control house and the ways that EM fields and conducted transients can penetrate the building.

Beginning with a metal substation building, one can see in Figure 3 that there are many ways that EM fields and currents can penetrate the building and then reach the electronics inside (not shown). The best approach is to evaluate the control houses by testing their shielding effectiveness with emphasis on those recently built. The reason for considering recently built buildings is that one would like to emphasize those using current construction techniques from local vendors. The best test method is to use the signals from radio stations in the AM, FM, Digital TV, and cellular bands to measure the fields outside and inside the building. This allows the electronics to continue operating, as there is no new field being transmitted. This method is fully described in IEC 61000-4-23 [5] and is very quick to apply.

Once one finds the best building for a power company, then the next step is to evaluate the many possible EM leaking points, as are clearly observed in Figure 3. Using normal EMC protection techniques, one can improve the grounding and shielding of cable entries, shield windows with wire grids, provide gaskets for the doors, provide filters for the power entry, etc. [6]. The goal is not to protect all penetrations, but rather to determine which penetrations should be improved on a cost-effective basis. Once the best set of protection is installed, then testing should be performed again to ensure that the building achieves its recommended level of protection. While this approach will consider different types of building designs in the U.S., as there are over 150 major power companies in the U.S., there may be fewer or even one company operating a national power grid in European or Asian countries, which will make this process more efficient. Also, in the U.S. there are companies that make control house buildings in a factory that are transported for installation. In this case, there could be efficiency in the building evaluation process.

Factors that Can Reduce the Required Shielding Effectiveness for Some Buildings

One of the special characteristics of a high-voltage control house with modern solid-state electronics inside is that the electronics must survive the daily electromagnetic disturbances typical from the switching transients in the high-voltage yard. Because of this, the IEC has published a special set of EMC immunity requirements for electronic equipment in high-voltage substations and power stations [7].

While there are requirements for radiated and conducted environments in this standard, those that are most severe are those of the conducted environments, which include the electric fast transient (EFT) test as described in IEC 61000-4-4 [8]. This voltage pulse has a 5 ns rise time and a 50 ns pulse width. The typical coupled E1 HEMP voltage for an above-ground conductor, such as a microwave cable, GPS, or camera cable, has a 10/100 ns pulse shape. The typical common mode requirement for the EFT is a peak of 4 kV for the electronics in a control house, while the expected transient for a buried yard cable is ~20 kV. So only a modest level of E1 HEMP protection is needed for the incoming yard cables. For a building shield of 30 dB, the worst-case internal E1 HEMP field would be ~1.7 kV/m. The coupled levels of conducted transients to internal cables will be lower than the 4 kV EMC immunity level. Unfortunately, some existing concrete substation buildings have been tested to shielding levels as low as 6 dB, which would allow fields that are too high into the building.

While the 30 dB level of shielding (along with POE protection) appears adequate for high-voltage power control houses, the situation is different for control centers. Each power company typically has 1 main control center for their high voltage substations, and a backup control center in case there is a failure at their main control center. The control center typically has communications and computer rooms, and digital displays to connect to all of their substation buildings to provide real-time information to ~4 operators.

While most of the power system operates with computer control, there are times when a particular substation loses communications, or there is a natural event such as a fire, lightning, or a fault that impacts the operation of the grid. These control centers are important to ensure that each grid operates efficiently and to prevent a blackout. The significant aspect of the control centers is that the electronics are not designed to tolerate high levels of EM noise as are those in a substation control house. Typical electronics are usually required only to have a “residential” level of immunity from EM disturbances, which could be as low as 0.5 kV for the EFT immunity test or up to a factor of 8 below the 4 kV requirement for substation electronics. This means that a control center needs approximately 50 dB of shielding effectiveness to protect its electronics.

In the recent past within the U.S., 3 separate new control centers have been built to protect against HEMP. Due to the relatively high level of shielding required, a decision was made in all 3 cases to use the military standard, MIL-STD-188-125-1 [9], with some modifications to correct for aspects of the standard that are not cost-effective [10]. During the construction of the first new HEMP control center in Houston in 2013, the A&E firm developing the construction plans evaluated the additional cost of an 80 dB HEMP shielded building vs. “normal” construction to be approximately 4%. This is consistent with cost studies performed in the past for the U.S. military for highly shielded buildings. It should be noted, however, that the cost of building a highly shielded building when the levels of required shielding are not high, is not cost-effective.

A third category to be considered are the power plants generating electricity. Of course, there are many different types of power plants from thermal (including nuclear), to solar panels, to wind turbines, to turbines at dams. In most cases, the large power plants need to convert turbine medium voltages to high voltages for transmission to population centers, and thus require a power substation; renewable plants also need a substation to coordinate the final AC power flow to the correct voltages and the proper phasing with the existing AC network. Therefore, the protection levels and approach required are the same as the high voltage substation control houses.

Clearly, those power plants that produce a significant amount of power for a particular company should be considered as a protection priority from the threats of HEMP and IEMI. It is also noted that power plants are often not owned by the power company operating the power network, introducing another difficulty in the hardening process.

One factor mentioned at the beginning of this article is the fact that the HEMP standards generally specify the worst-case HEMP environments for two reasons. This provides a reasonable upper bound of the fields that could be produced, but it also avoids the variability of the fields that could be produced based on the height of the burst, the location of the burst, the yield of the weapon, the weapon design, and for E3 HEMP the deep ground conductivity under the burst.

One presumes that if an attack is planned, the attacker would try to maximize the field levels. Of course, even if this is done, one cannot maximize the fields over the entire footprint of the exposure. For E1 HEMP the fields toward the edge of the exposure region can be lower than the worst case by factors of 2 to 10, and the maximum field exposure area is typically less than 10% of the total area exposed. For the E3 HEMP the fields typically fall to 10% or less at the edge of the exposure, and if the deep ground conductivity is high, all of the fields will be smaller than the worst case. This means that only a few substations will see the maximum fields.

There are other factors to consider, including the orientation of power lines, which affect the coupling of E1 HEMP. Based on the polarization of the E1 HEMP fields for the center of the U.S., E-W oriented cables will pick up more than 10 times the peak current and voltage than will N-S cables (in the air or buried) [11]. These are important aspects of the HEMP variability, and one should consider the advantage of using lower levels of fields based on these variations.

The last point of consideration is that all of the discussion thus far has been to evaluate the best way of adding protection to a “partially” shielded building. It is possible that in some cases, if an outage can be accepted for some limited time, then a plan to accept electronic upsets, and limited damage to electronics might be acceptable. This could be achieved by having replacement electronics available in the building that are not connected to power or data and which are placed in a modestly shield cabinet inside the building.

This approach could be used for buildings that are not as critical to the overall operation of the power grid, although a criticality study would need to be performed. In the U.S., power companies have been asked by the North American Electric Reliability Corporation (NERC) to determine their 9 most important assets, and to consider them to develop protection plans against different threats (but not necessarily HEMP and IEMI).

Recommended Approach for Protecting Buildings – HEMP and IEMI

As mentioned earlier in this article, the best approach for substation control houses is to evaluate the construction techniques of recently built houses with a preference for metal buildings. A shielding effectiveness measurement campaign should be developed to identify the best existing buildings in the network. As indicated earlier, the use of radio communications signals is a very efficient way of testing an operating control house, as the radio signals are already occurring, and they are usually far enough away to be considered to create a plane wave incident field. This method has been evaluated in peer-reviewed journal articles and is presented as a testing option in IEC 61000-4-23 [5].

Once this process is accomplished, then the best building (or two) should have between 20 and 30 dB of shielding effectiveness across the E1 HEMP spectrum (1 – 100 MHz). From past experience, the priority for improving the protection of the building is usually first determined by any above-ground penetrations of the shield without complete bonding and grounding. These are usually cable entries for GPS antennas, microwave cables, camera cables, A/C mounting, windows without EM mesh, and door gaskets. If the yard cables penetrate the building walls and not the floor, this is a major leakage path to be considered for improvement.

The best way to minimize the repairs and their cost is to perform the improvements while making measurements, usually with temporary copper tape, to determine the most important leakage points. In any event, after the repairs are made, it is important to remeasure the shielding effectiveness of the building with the EMC repairs completed. For buildings manufactured in a factory and then shipped, the measurements should be made before and after shipping to determine the impact of the shipping process.

As mentioned earlier, this process works well when the target protection level is 30 dB but does not work well (on a cost-benefit basis) for a control center building for the reasons mentioned earlier, which needs on the order of 50 dB. It is very difficult (and costly) to raise the shielding effectiveness of a 20 dB building to 50 dB by making repairs. Therefore, it is recommended that the MIL-STD-188-125-1 approach be used, which is also presented in IEC 61000-6-6 [12], with consideration of reducing some of the unnecessary costs and correcting the errors in the standard [10]. It is also recommended that the newest version of the MIL-STD-188-125-1A [13} not be used because it is not published for public use and has not been peer-reviewed by commercial technical organizations (IEEE, CIGRE). It is recommended only for certain military projects.

If particular power system buildings are to be considered for HEMP and IEMI protection, then the power substation at the power plant can apply the control house procedure mentioned above. If there is a local control center building for the plant, then it should also be considered for protection, but at the higher level of 50 dB. Typically, a control center room for a power plant is much smaller than a control center room for a power company’s entire grid, so it may be possible to build a shielded room for this purpose at a lower cost than for an entire building.

While the emphasis in this section has been on the E1 HEMP, the IEMI has some differences to consider, although they are not usually very costly. First, the IEMI threat in the frequency domain is typically found between 100 MHz and 10 GHz. It is noted that in IEC 61000-2-13 [14], there are narrowband threats that are defined but also wider bandwidth threats (even single fast pulses, like JOLT [15]. The main difference with IEMI is that the threat comes from a local antenna outside the fence. The fields fall off rapidly from the antenna, and a solid metallic fence can cause the attacker to move further away to “fire” their threat over the fence. While normally substation electronics are in a building that is not close to the outside fence, there have been cases where they are close to the fence. These cases are clearly those where a new building needs to be built away from the fence to prevent very high IEMI fields from exposing the equipment.

When IEMI is considered in addition to the E1 HEMP, one factor to consider immediately is that the window meshes must be designed for higher frequency fields. E1 HEMP requires about a 4-inch mesh, while IEMI requires a mesh of a few cm [16]. Fortunately, there are commercially made meshes for a frequency of 18 GHz, which can be used for the IEMI threat. Of course, if the windows are not needed, they should be replaced with metal, eliminating the need for meshes.

Another point, in general, is that the cable penetration grounding is not as critical for IEMI, as the IEMI fields do not couple or propagate as well on external metallic cables as from E1 HEMP fields due to their frequency range. On the other hand, significant cracks in the shield allow more penetration of fields at higher frequencies. If the IEMI is important to a particular building due to close public access, then it is important that the building be tested at higher frequencies using cellular radio signals to ensure that important apertures are well sealed.

Finally, there are IEMI field detectors that are being made today [4], and these could be used to determine if an attack is underway. The placement of these detectors is important to ensure that the main attack scenarios are covered and that any alerts for an attack are evaluated against the possibility of false alarms.

Recommended Approach for Protecting Large Transformers

While this article has dealt mainly with the high‑frequency threats of E1 HEMP and IEMI on electronics that control the power grids, the late-time E3 HEMP is a serious threat to the large transformers that are the key part of the power transmission network. While high voltage (HV) transformers are defined to operate at V > 100 kV, most modern transmission systems operate at 400 kV (Europe) or 500 kV (U.S.). In China and India, new HV transformers are being built to operate at 1 MV to efficiently move power.

The process of coupling E3 HEMP fields and also geomagnetic storm fields into the power network is complex; there is an IEEE paper [17] that explains the entire process and a recent CIGRE Technical Brochure that reviews the worldwide measured geomagnetic fields from 1989 to 2018 [18]. It is noted that the E3 HEMP threat and the typical CME geomagnetic storm are very similar disturbances and couple to power grids and cause transformer difficulties in similar ways. Fortunately, there are modeling techniques that can evaluate power grids, which are essentially very large antennas, to determine where (which transformers) the largest currents will occur given an E3 HEMP or a large geomagnetic storm. This modeling process is not difficult and will identify those transformers at the highest level of risk.

Of course, it is prudent to validate the modeling technique used, and even a small geomagnetic storm from the recent past can be used for that purpose. It is useful to add geomagnetically induced current (GIC) monitors on transformer neutral cables to perform the validation. It is noted that the CIGRE TB 780 does provide information on how to install GIC sensors on transformers [18].

If the modeling process indicates a significant number of important transformers are at risk, the next step is to add additional GIC monitors on these transformers to observe the response of these particular transformers relative to others in the network. Over time, one should be able to confirm that these transformers will carry a significant portion of the GIC current. It is noted that transformers at the edge of the grid and transformers in regions of the earth where the deep ground conductivity is low are most at risk.

Once the utility is concerned that a particular transformer is at risk, and it supplies a significant amount of power to the overall network, then protection needs to be considered. The main cost‑effective treatment is to add a neutral resistor [19]. One of our customers did this, and it reduced the induced current in the transformer by about a factor of 2, as indicated by a GIC measurement made during a significant geomagnetic storm in the early 2000s. The second treatment is a neutral capacitor, but it must be protected against power faults and lightning surges with a bypass arrester. Otherwise, the capacitor will be damaged. The problem with the capacitor is that, with bypass protection, they are expensive, so on a cost-effective basis, the neutral resistor seems to be the better approach.

In terms of resilience, another approach is to have backup transformers at the critical substations where high levels of GIC may occur. While it is typical for power companies to purchase a few large transformers in advance, the selected transformers are based normally on the age of the transformers. In this case, the placement of the transformers should be based on the probability of a high GIC and the importance of the substation to the overall operation of the grid. As noted by the EMP Commission in 2008 [20], if one waits for large transformers to be damaged during an E3 HEMP event, the delivery time could be many years, especially if a large number of transformers were damaged during one event.

Perform Protection Over Time, Not All At Once

One of the questions that always occurs when the subject of HEMP and the power grid is discussed is why do we not protect the grid immediately? It is true that, as indicated in this article, we do know how to do the job. The problem is the cost will be very high due to the large number of high voltage substation control houses in the U.S. (~9000) and many more worldwide, and the number of experts available to perform the work is not large.

This is why the idea of evaluating buildings, which already exist, and hardening them on a cost-effective basis to achieve a sufficient level of protection is the best way to develop a prototype approach that can be used in the future, as power grids expand. This can be done separately by each power company. If these projects, including cost information, could be openly published as the work is completed, this would be a significant help to smaller power companies. It is possible that some national prototypes could be developed.

In the same way, the protection of power control centers requires higher levels of shielding, but it would be beneficial if those adapting the MIL-STD-188-125 approach to commercial applications as described by the IEC could publish their results so cost savings could also be shared across the industry.

Finally, the development of a group of backup power transformers at substations where the transformers are at significant risk from E3 HEMP is something that can be done over time and would only modify the procedures that are already embraced by the power industry. The main feature would be to determine the transformers at significant risk, along with other factors already considered by power companies.

Summary and Recommendations

The main recommendation of this article is to start the process of upgrading high voltage substation control houses to E1 HEMP and IEMI to protect the electronics inside by evaluating their best metal buildings for their shielding effectiveness and using the typical EMC hardening techniques to improve the shielding levels to at least 30 dB. Testing is needed to ensure the work is done on a cost-effective basis, and rapid test methods are recommended.

In a similar fashion, control center buildings need to be protected to ~50 dB due to the susceptibility of the type of electronics found inside, and this level is not amenable to reaching from a starting point of ~20 dB. This means the basic high shielded building approach should be used, but the MIL-STD-188-125-1 needs to be adapted and those adaptations published so it can be used on a commercial basis. The IEC has started that process by indicating areas where the military standard is not cost-effective, but more work is needed.

The consideration of the IEMI threat in addition to E1 HEMP is important, and while the threat does not cover a large area at one time (unless there is a coordinated attack), the IEMI threat is much more probable than a HEMP attack. The features of an IEMI attack are well understood, and many of those features are discussed in this article. The main factors are to ensure that an IEMI attacker cannot get close to the electronics, and to consider upgrading the substation fences to reduce the fields incident on the electronics. In terms of EM protection, the most important add-on for IEMI is to ensure that a fine metal mesh is used to cover windows.

The final aspect of this article is that the method of protecting the large power transformers that are very expensive and take many years to replace is straightforward. Validated analysis methods exist and can be used to determine which transformers are most at risk. Adding GIC sensors to those transformers and evaluating their measurements during future geomagnetic storms can confirm the potential vulnerability of particular transformers. In terms of protection, the neutral resistor appears to be the most cost-effective in that it can substantially reduce the currents that will flow in a particular transformer. A resilience approach includes providing backup transformers at the substations where transformers that are at risk are located.

As one who has worked directly for more than 20 power companies worldwide on this problem for over 20 years, I am trying to develop an industry-wide approach to cost-effectively protect power grids throughout the world. In addition, I have worked directly with IEC SC77C as the Chair for 25 years in the past and as an expert in writing and updating existing standards to be more accurate and cost‑effective. This is too big of a job for a small group of experts to perform, and we need to develop techniques that can be used and replicated easily.

References

1. William A. Radasky, “Protection of High Voltage Power Substation Control Electronics from HEMP and IEMI,” In Compliance Magazine, 28 May 2021.
2. Edward Savage, William Radasky, “Recommended E3 HEMP Heave Electric Field Waveform for the Critical Infrastructures,” Report of the Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack, Volume II, July 2017. http://www.firstempcommission.org
3. IEC 61000-2-9 Ed. 2 CDV: Electromagnetic compatibility (EMC) – Part 2-9: Environment –  Description of HEMP environment – Radiated disturbance, January 2024.
4. IEC 61000-5-6 Ed. 1.0 (2024-04-05): Electromagnetic compatibility (EMC) – Part 5-6: Installation and mitigation guidelines – Mitigation of external EM influences.
5. IEC 61000-4-23 Ed. 2.0 (2016-10-20): Electromagnetic compatibility (EMC) –
   Part 4-23: Testing and measurement techniques –
   Test methods for protective devices for HEMP and other radiated disturbances.
6. Edward B. Savage, William A. Radasky, “Protection Issues for Power Substations from HEMP Adverse Effects,” IEEE EMC Symposium, Spokane, Washington, 1 August 2022, pp. 524-528.
7. IEC 61000-6-5 Ed 1.0 (2015-08-21), Electromagnetic compatibility (EMC) – Part 6-5: Generic standards – Immunity for equipment used in power station and substation environment.
8. IEC 61000-4-4 Ed 3.0 (2012-04-30), Electromagnetic compatibility (EMC) – Part 4-4: Testing and measurement techniques – Electrical fast transient/burst immunity test.
9. MIL-STD-188-125-1, “ High-altitude electromagnetic pulse (HEMP) protection for ground-based C4I facilities performing critical, time-urgent missions, Part 1 – fixed facilities,” 17 July 1998.
10. William A. Radasky, Sergio N. Longoria, “Recommended Improvements for MIL‑STD-188-125-1,” 32nd URSI GASS, Montreal, Canada, 19-26 August 2017.
11. Ianoz, M., B. I. C. Nicoara, and W. A. Radasky, “Modeling of an EMP Conducted Environment”, IEEE Transactions on EMC, Vol. 38, No. 3, August 1996, pp. 400-413.
12. IEC 61000-6-6 Ed. 1.0 (2003-04-09): Electromagnetic compatibility (EMC) – Part 6-6: Generic standards – HEMP immunity for indoor equipment.
13. MIL-STD-188-125-1A, “ High-altitude electromagnetic pulse (HEMP) protection for ground-based C4I facilities performing critical, time-urgent missions, Part 1A – fixed facilities,” 17 July 2021. Note:  Available only with Federal Government approval.
14. IEC 61000-2-13 Ed. 1.0 (2005-03-09): Electromagnetic compatibility (EMC) – Part 2-13: High-power electromagnetic (HPEM) environments – Radiated and conducted.
15. C.E. Baum et al, “JOLT: a highly directive, very intensive, impulse-like radiator,” Proceedings of the IEEE, Volume 92, Issue 7, July 2004, pp. 1096‑1109.
16. Edward B. Savage, William A. Radasky, “Wire Mesh Radiated EM Shielding Effectiveness: Time Domain Measurement and Theory Verification,” IEEE EMC Symposium, Grand Rapids, Michigan, 2023, pp. 474-479.
17. William A. Radasky, “Overview of the Impact of Intense Geomagnetic Storms on the U.S. High Voltage Power Grid,” IEEE Electromagnetic Compatibility Symposium, Long Beach, California, 15-19 August 2011, pp. 300-305.
18. “Understanding of geomagnetic storm environment for high voltage power grids,” CIGRE Technical Brochure 780, October 2019.
19. John Kappenman, “Geomagnetic Storms and Their Impacts on the U.S. Power Grid,” Oak Ridge National Laboratory, Meta-R-319, January 2010.
20. “Report of the Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack: Critical National Infrastructures,” U.S. EMP Commission, 2008.