If you’re not ready for TEOTWAWKI, you’re probably asking, “Can’t we do something to keep the grid from going down?” The answer is “yes”. There are two approaches– early warning and hardening of equipment.
In theory, early warning relies on the ACE and DSCOVR satellites, located one million miles from the earth, to measure the intensity and polarity of a storm and then issue warnings, which utilities would use to take steps to protect their equipment. In reality, large storms are too fast, allowing maybe 15 minutes of warning. Nuclear plants are supposed to be in “cold shutdown” if outside power is expected to be lost. There is no way to do that in 15 minutes. If a utility wants to take pre-emptive action, it has to shut down before it knows the polarity of the storm. Polarity is key, because a huge storm with the wrong polarity may be no danger. The CEO who blacks out his system for a false alarm will be gone. So, no utility will do it. Nobody will say that early warning is not a practical defense, especially after we just spent $340 million to launch DSCOVR. However, GIC is generated by both geomagnetic storms and by the E3 component of a High-altitude ElectroMagnetic Pulse (HEMP) nuclear detonation. A HEMP detonation over New York City is predicted to take out 551 EHV transformers, which is 51% more than the 365 predicted from a 100-year geomagnetic storm, and there will be no warning for a HEMP! However, hardening provides protection from both sources of GIC.
So hardening is the other preventive action. The goal of hardening is to have your equipment be able to ride out a storm without depending on human operators to make all the right choices at just the right time. The definitive source is John Kappenman’s Meta R-322 report, “Low Frequency Protection Concepts for Electric Power Grids: Geomagnetically Induced Current (GIC) and E-3 HEMP Mitigation”, in which he describes three basic choices– series capacitors, neutral blocking capacitors, and neutral resistors.
Series capacitors are installed on the transmission lines. They completely block GICs on lines where they are installed, and their “reactive power contribution is instantaneous and self-regulatory.” [Gruenbaum & Rasmussen, Series Capacitors for Increased Power Transmission Capability of a 500kV Grid Interconnect, pg 2. undated] They are the preferred choice in the long EHV lines in the western U.S. and Quebec, because they have the everyday financial benefit of providing “a considerable increase of the power transmission capacity over the corridor, reducing or postponing the need for additional transmission lines” [Gruenbaum & Rasmussen, p 6]. The down side to this option is that they are very expensive and the control mechanisms are subject to being tricked by the harmonics from GICs, potentially resulting in loss of reactive power just when needed most to maintain voltages during a geomagnetic storm. Also, modeling of their use in the western U.S. indicates that they would only reduce total GICs by 13-22% and in Quebec by about 30% for the entire system [Kappenman, Meta R-322, pp 3-4].
Neutral blocking capacitors completely block GICs in their transformers. However, by completely blocking these currents, they force the current elsewhere in the system, like series capacitors. Since the grid needs to be grounded for fault conditions, bypasses need to be added. These two issues vastly complicate the engineering of these devices over a system and add considerable expense. A FEMA workshop in February of 2010 concluded, “Hardening EHV lines and transformers through the installation of neutral-blocking capacitors is possible. But doing it for all utilities supporting 345kV and above is economically prohibitive.” Still, for a very at-risk high value EHV transformer, they may be the only option. After the 1989 storm, Quebec Hydro spent C$1.2 billion (C$32/person served) on a combination of series capacitors and neutral blocking capacitors.
The third option, and the one clearly preferred by Kappenman for most locations, is the low-ohm neutral resistor. The neutral resistor only blocks about 60% of the GIC flow through it. In the 100-year storm model, modified to include 1388 (about half of the eastern U.S. “fleet”) lower voltage 230kV transformers, 551 of 3550 transformers are predicted to be damaged. With 5-ohm resistors on all transformers, only 37 are damaged, which is a 93% reduction. [Meta R-322, figure 7-19 and pg. 7-14] Neutral resistors do not interfere with normal fault protection and are simple devices, which makes them the low-cost alternative.
A fourth option is to mitigate impacts from the storm (not necessarily the damages) by stockpiling spare transformers and other equipment, adding more backup generators and greatly increasing fuel supplies for all backups. However, buying spare transformers is expensive and risky, because you don’t know exactly which ones will fail.
Cost estimates for hardening are all over the map. John Kappenman, the primary author of the Metatech 2008 report, has been quoted as estimating $1 billion for “hardening and stocking replacement parts” [personal communication with Matthew Stein, When Technology Fails, NEXUS magazine article 2008]. The latest estimate I found was from Congressional sub-committee testimony by Joseph McClelland, director of the Office of Electric Energy Reliability for the Federal Energy Regulatory Commission (FERC) on June 12, 2012. He estimated the cost of hardening (type not specified) electrical grids against geomagnetic disturbances at $500,000 per transformer. For low-ohm neutral resistors total estimated costs, including peripherals and installation is $40-100,000 for each resistor [Meta R-322, p. xi]. If the 3550 transformers in the expanded 100-year storm model were all protected the maximum cost would be $355 million (or $1.15 per person) for a hardened national grid. This is minuscule compared to the trillions it would cost to recover from a 100-year storm.
There have been three bills introduced in Congress to require protection of the national grid– the GRID Act, the SHIELD Act, and the Critical Infrastructure Protection Act (CIPA). The sponsors of the GRID Act surveyed 150 companies in the bulk power industry and found that only 27% of the 90 respondents had “taken specific measures to protect against or respond to geomagnetic storms” and that “most utilities do not own spare transformers“ [Electric Grid Vulnerability, staff report of Congressmen Markey and Waxman, May 2013]. This indicates how little the power industry is doing voluntarily to address the grid’s vulnerability. It motivated the sponsors to re-introduce the GRID Act. None of the three bills have made it through Congress.
Maine was the first state to pass its own requirement for grid protection. Other states may follow, but it’s hard to imagine states having more success than federal legislators.
In 2012, with the early bills bogged down in Congress, FERC took the unprecedented step of issuing FERC Order 779, requiring the North American Electric Reliability Corporation (NERC) to establish minimum reliability standards for protection from geomagnetic storms and GICs. NERC is a unique organization that is the agency appointed to establish reliability standards for the electric industry. Its membership consists of the companies it regulates, and it takes 75% membership approval to pass a new standard. The grid– this country’s most critical infrastructure– is self-regulated. The exception to this is nuclear power plants, which are under the jurisdiction of the Nuclear Regulatory Commission (NRC).
NERC’s membership approved reliability standards for GICs in December of 2014. Power industry watchdog groups have attacked the NERC “100-year benchmark storm” as “junk science” [Dr. Peter Pry, Ex. Dir. of the Task Force on National and Homeland Security, in comments to the “The Blaze” 10/24/2014]. A summary of the comments from reviewers of the NERC draft standards in October of 2014 identifies the following defects in the proposed standards:
- The benchmark 100-year storm is 1170 nT/min at 60 degrees of magnetic latitude while previous research has established the 100-year storm to be 4000-5000 nT/min at 50-55 degrees magnetic latitude.
- The NERC standard, when compared to actual measurements in previous storms, underestimates by 100-400%. When Kappenman’s 100-year storm model is subjected to the same scrutiny, it is generally within 20%.
- There have been three storms in just the past 40 years that “greatly exceed” the benchmark standard [comments of Kappenman and Birnbach on Draft Standard TPL-007-1, submitted to NERC October 10, 2014]. Actual measurements in Tillamok, Oregon for a storm on Oct. 30, 2003 illustrates that the benchmark standard extrapolated per the NERC formula is only 1/30 of what is expected in a real 100-year storm.
When the proposed reliability standard is forwarded to FERC, they may only approve (without modifications) or disapprove. In an interview with “The Blaze” in October of 2014, Dr. Peter Pry commented, “It is better to have no GMD (geomagnetic disturbance) standard than a fake GMD standard that will lull policymakers and the public into complacency about an existential threat to our civilization.” It is believed by some that NERC wants a minimal standard approved by FERC. When catastrophe happens they can then dodge liability by claiming that their members “met the federal standard.”
The vulnerability of EHV transformers to HEMP and geomagnetic GICs is real and resulting damages are a matter of when, not if. FEMA-style recovery is not feasible for long term nationwide impacts. Early warning is impractical for geomagnetic storms and non-existent for HEMP attacks. Hardening appears to be the only logical approach to preventing economic and societal collapse, and a program of primarily low-ohm resistors seems the clear affordable path to transformer protection. FERC rule-making is not getting the reliability standards that are needed, and most legislative solutions follow a similar NERC reliability standards approach.
No modern power system has ever experienced a 100-year geomagnetic storm, so investor-owned (for profit) utility company execs cannot get their heads around the dire consequences of Kappenman’s model. But how can they be so blind to the fact that it is far less expensive to prevent damages than to pay the consequences of outages, much less the cost of transformer replacements, especially when they can pass the costs on to their customers?
In “Risk Mitigation in the Electric Power Sector: Serious Attention Needed”, Daniel C. Hurley, et al state that “the private sector generally will not invest in activities which negatively impact the bottom line or for which a known steady return on investment does not exist. Thus it falls to the government to invest in activities measured not by return on investment but rather in terms of the “common good“. When neither the private sector nor the government see the benefit of spending $1.15 per person to prevent TEOTWAWKI, it is left to informed individuals to spend thousands to fend for themselves. A 100-year geomagnetic storm is an inevitable natural event. Add to it the other grid threats of HEMP, physical attack, and cyber attack and preparing for a grid down world makes more sense than ever.