America’s Power Grid
& Space Weather
Arthur Cox
August 20th 2018
675.751 Space Weather Johns Hopkins
Table of Contents
Table of Contents 2
Abstract 3
Background 3
Objective 3
Space Weather Phenomena 4
Geoelectric Field α Scaling Factor (Figure 4-1)
Geomagnetic Map of North America (Figure 4-2)
Physiographic Regions of the Continental US (Figure 5 -1)
Geoelectric Field β Scaling Factor (Figure 5 -2)
Power Generation 9
Disaster Planning 10
TNMP Emergency Operations Plans (Figure 1)
Economic Disadvantage 12
Load and Demand (Figure 2)
Economic Loss (Figure 3)
Mitigation 13
Conclusion 14
References 15 Abstract
Background
PNM Resources is a Southwestern United States Power generation and supply company. PNM has a generation capacity of over 2,500 megawatts, 14,500 miles of transmission and distribution lines, and services near 750,000 homes and businesses across Texas and New Mexico. 35% of PNM’s generation is Coal fire, 35% natural Gas, 15% nuclear and 15% Renewables. Renewables consist of Solar, Wind, and Geothermal. A small percentage of power generation is customer-owned solar.
Modern power grids are very susceptible to the effects of space weather. The main concern for any power grid is solar wind, a constant stream of highly charged particles released from the upper atmosphere of the sun. When these highly charged particles strike Earth’s magnetic field we get a “Geomagnetic Storm”. This produces geomagnetically induced currents that have the potential to disrupt entire power grids.
Objective
This paper will study America’s Power Grid and its ability to handle various space weather phenomena. This will involve research on the American utility, Public Service Company of New Mexico’s (PNM Resources, or PNM for short) and its subsidiary, Texas-New Mexico Power Co. (TNMP). This paper will explain any procedures or contingencies the power company already has in place, and research possible further solutions for recover after a geomagnetic storm. There will be an extrapolation on these observations and conclusions drawn from PNM Resources to provide an analysis of how the United States utilities and its surrounding countries would be impacted by a space weather event.
In order to systematically understand how the power grid would fail and recover from space weather we will simulate 3 individual failure scenarios. These scenarios are an observation of PNM’s power grid and the supporting technology’s reaction to space weather. The scenarios would be set at three stages of downtime: Single Day, Multi-Week, and Multi-month disruption.
This paper will then detail current tactics to mitigate disruption and to repair damage given due to three scenarios. Speculation of possible improvements or additions to the power grid to diminish the impact of geomagnetic storms to the state of New Mexico will be suggested. These proposed improvements would be paired with quantitative analysis that suggests positive results. It is possible these improvements could come in the form of a repair technology that would shorten the downtime the electrical grid would experience service disruption.
Space Weather Phenomena
The main concern for any power grid is solar wind, a constant stream of highly charged particles released from the upper atmosphere of the sun. When these highly charged particles strike Earth’s magnetic field a “Geomagnetic Storm” forms. This produces geomagnetically induced currents that have the potential to disrupt entire power grids.
To understand the effects of space weather, we must understand the physics of transferring electricity. In the 19th century it was found that varying the magnetic field around a wire over time is able to produce an electrical current. Whereas the time rate of change of the magnetic flux passing through a current loop is proportional to the current generated around that loop [NOAA – Reference 3]. This movement of electrons over a wire to creates a current that produces a magnetic field, a magnetic field that can be easily disrupted by outside sources.
In many cases the Earth’s magnetic field, a dipole field, protects the Earth and its inhabitants from the ionized, electrically charged particles that are emitted from the sun. These are the types of particles that would produce intense magnetic fields in Earth’s upper atmosphere and Ionosphere; this is where you will see the famous Northern and Southern Lights. While the Earth’s magnetic field does protect us from the large amount of radiation produced by Corneal Mass Ejections (CMEs) and Solar Flares, the changing magnetic field does propagate to the surface and disrupt the electron flow of terrestrial electronics and power grids.
To define what kind of Geo Magnetic Disturbance (GMD) has the ability to impact the Continental United States, we will perform a statistical and mathematical analysis of past storms to model a potential GMD
The issue we encounter when understanding past geomagnetic storms is the lack of high fidelity data for event before the 1980’s. This is of note for the Carrington Event of 1859 as the tools, and therefor data, did not exist to make any models of the geoelectric fields. Meaning we cannot draw any conclusions directly from these events but must build a statistical model that can account for storms of the 18th and 19th century.
According to the storm-time distribution index DST, which does allow extrapolation to events before 1980, the probability of the Carrington and March 1989 events is between 1.6% to 13.7% recent over a 10-year period. While the range is large, it is reasonable to conclude that we will get a storm of this nature at least once in a hundred years. Meaning, according to North American Electric Reliability Corporation (NERC), the Carrington Storm is the type of storm that the Federal government should account for when regulating its power distribution infrastructure. This then results in a benchmark peak geoelectric field amplitude of 8 V/km.
An inconsistency in this assumption is that it does not consider any local properties that might affect a geomagnetic field produced by a storm. These properties include features such as crust thickness and particle density variabilities in the ionosphere. This could be remedied by taking advantage of models from more modern storms over a wider area, specifically taking measurements of the same storm’s geomagnetic affects over a wider area. As the “Spatial structure of high-latitude geoelectric fields can be very complex during strong geomagnetic storm events [NERC – Reference #]”.
The above calculated probabilities are calculated on the Mid-Canadian latitudes of 60 degrees. To scale the Geomagnetic disturbance for lower latitudes a scaling factor alpha (α), to account for local geomagnetic latitude, and factor beta (β), to account for local Earth conductivity structure would be applied where:
When L is defined as geomagnetic late in degrees
and where 0.1 ≤ α ≤ 1.0
With the above models not having historical data on parameters such as the wave shape or time series of the geoelectrric fields, like the previously mentioned Carrington storm, the Peak Geoelectric field (E Peak) can be optioned utilizing the scaling factors noted above.
Where 8 V/km is the reference peak amplitude value.
E-Peak is indeed an approximation with faults. As in the Geoelectric scaling factor β is sampled at a 10 second resolution over varying quantities of storms across the different physiographic regions.
Below are the factors and information sourced to calculate an E-Peak estimate for the North American Region.
Geoelectric Field α Scaling Factor (Figure 4-1)
Geomagnetic Map of North America (Figure 4-2)
Physiographic Regions of the Continental US (Figure 5 -1)
Geoelectric Field β Scaling Factor (Figure 5 -2)
For the service area of PNM resources, we will calculate the peak geoelectric field. The center of the service region is 35.08° North and 106.65° West, which gives us a geomagnetic latitude of 43.45° North. Looking at Figures 4 and 5 above we see that the geoelectric field calling factor alpha (α) is equal to 0.20. As the service area of PNM in New Mexico spans two individual physiographic regions the geoelectric field calling factor beta (β) is ether Earth Model CL1 of 0.76 or Earth Model IP4 of 0.41. In this case we will use largest β Scaling factor. As our goal is to always account for the worst-case scenario.
E-Peak for PNM’s service area is 1.216 (V/km)
A peak geoelectric field of 1.216 V/km is lower than the continental United States average of 5 V/km. This means that, typically, New Mexico is less suitable to the strong geomagnetic effects of a storm than most regions of the United States. This is because of the low latitudes of the State of New Mexico and Texas, and not necessarily because of the Earth’s Crust’s electrical conductivity. As a β value of 0.76 is higher than the mean value for the country. Power Generation
The power generation interstice, as stated above, is very susceptible to the affects of a geoelectric event. The federal government and the NERC have produced assessments of the potential impact of the 8 V/km benchmark geoelectric storm. The testing as brought to light two systems that are the most susceptible to the geomagnetically-induced currents of a storm, Transformers (Both step-up and step-down) and transmission lines. Whereas the transformers would experiences catastrophic thermal levels when trying to operate during a storm and the transmission lines would .
In order to make a determination what equipment should be deployed when constructing a sub station, NERC ran thermal models using the benchmark geoelectric storm field waveshape. These models were based of the functions below:
where
for any value EE (t) and EN(t) in Amps/phase
and GICN and GICE is the Northward and Eastward geoelectric field in A/phase/V/km.
With the assumptions of the benchmark geoelectric storm of 8/Km, These assessments have produced a requirement of a 75 amp per phase per transformer. that would allow this transformer to operate below the 200°c upper operating threshold during its life expectancy of 30 years. The requirement was produced with the expectation that the transformer will only will experience the high temperatures for a 30 minute periods three times during its lifetime.
Disaster Planning
PNM resources defines two levels of damage assessment that applies to all forms of natural disastrous. Their damage assessment guidelines identify if the disaster is minor or major by looking at the impact and potential downtime. Typically, the natural disaster would fall into the major category if the National Weather Service (NWS) or the Federal Emergency Management Agency (FEMA) has declared the storm to have a big enough to need of federal assistance. Whereas PNM does not have the ability to rebuild on their own and may require the resources of other power generation companies to compensate for lost generation. There is however a proclamation that can be made by FEMA and the federal government that suspends trading on the power market. This can be done during historical-level storms when grid stability is a worrisome concern and must be a top priority.
The electric power grid consists of phases of getting power to the customer. This includes, but not limited to: generation, step-up transformers, transmission lines, step-down transformer stations, distribution substations, and the local service drops and house meters. In the event of a disaster TMNP and PNM Resources have prioritized work as follows:
Transmission Lines
Distribution feeders by priority level as exampled above
Local service drops and step-down transformers for Critical Care Facilities, Schools, Large Commercial Etc.
Fused Laterals on transmission poles
Transformers
Services and company owned Generation
In most city and states disaster plans, bringing the power distribution network online is the top priority. As the grid usually needs to be energized in steps, the power company has identified distribution feeders to prioritize. These distribution yards would be ones that power waste treatment, community shelters, government buildings and water pumps/wells. Below is a snap shot from the “TNMP Emergency Operations Plans”. This shows a list of circuits near Texas City, Texas that have been identified to power important infrastructure. TMNP has thee priority levels: High, Medium, and Low.
TNMP Emergency Operations Plans (Figure 1)
Economic Disadvantage
According to U.S Energy Information Administration (EIA), the average wholesale price in 2017 is $34.63 per kWh. Assume an average of one working person per household.
Load and Demand (Figure 2)
Economic Disadvantage ($/Wh) = ƒ (duration, season, time of day, notice)
Economic Disadvantage is a function of duration, the season, time of day and how much notice.
To account for a worst-case scenario lets assume:
1) Mid-summer
2) Middle of the afternoon
3) 1 hour
4) No notice
5) Central New Mexico (PNM’s control)
With no notice FEMA, the power generation and local governments would have no time to organize any efforts to remedy the situation.
Per PNM Resources budget sheets, the cost to repair the greater Houston power grid from Hurricane Harvey was 350 million dollars. The damage to a power grid from water would be similar to that from an electromagnetic storm, as both would disable all phases of power generation and distribution. Repair costs would be a fixed cost between all scenarios. This figure would be modified though to take in account the size of Infrastructure in central New Mexico.
Determine the cost of social support during a power outage we will look at FEMA’s expenditures per person. For a loss of electric power, potable water and waste treatment, FEMA estimates the cost to be $255 per person per day. For the economic disadvantage we will assume the 2010 US census household size of 2.58 persons.
Economic Loss (Figure 3)
Mitigation
Recently, in 2013, the North American Electric Reliability Corporation (NERC) and Federal Energy Regulatory Commission (FERC) established a list of mandatory and enforceable standards to directly address the risks to the grid from Geomagnetic Disturbance Events. Canada, despite not being subject to NERC or FERC authority, voluntarily complies with the standards.
The Geomagnetic Disturbance Mitigation plan involves solutions on all fronts of the issue. Conclusion
References
Blotter, Jimmie. “Company Profile.” Investor Fact Sheet, 31 Dec. 2016, www.pnmresources.com/~/media/Files/P/PNM-Resources/Attachments/fact-sheet-dec-31-2015.pdf.
Cox, Steve. “Senior Manager BTS at PNM Resources” July 2018.
Electric Power Transmission.” Aurora | NOAA / NWS Space Weather Prediction Center, www.swpc.noaa.gov/impacts/electric-power-transmission.
Hathaway, David, and Doug Biesecker . “How 'Space Weather' Affects Planes And Power Grids.” NPR, 12 Jan. 2012.
“U.S. Energy Information Administration – EIA – Independent Statistics and Analysis.” Chinese Coal-Fired Electricity Generation Expected to Flatten as Mix Shifts to Renewables – Today in Energy – U.S. Energy Information Administration (EIA), 2017, www.eia.gov/tools/faqs/faq.php?id=97.
“New Mexico GDP.” Department of Numbers, 2016, www.deptofnumbers.com/gdp/new-mexico/
“Federal Emergency Management Agency BCA Reference Guide”. June 2009
U.S. Census Bureau. “Households and Families: 2010”. Apr. 2012.
Williams, Laurie. “NERC Reliability Governance Manager at PNM Resources” August 2018.
Lucas, Greg M., et al. “Calculation of Voltages in Electric Power Transmission Lines During Historic Geomagnetic Storms: An Investigation Using Realistic Earth Impedances.” Journal of Geophysical Research: Atmospheres, Wiley-Blackwell, 26 Feb. 2018, agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2017SW001779.