Establishing the Economic Impacts of Kevin F. Forbes Research
Supported by the United States National Science Foundation Did
Geomagnetic Activity Challenge Electric Power Reliability During
Solar Cycle 23? Evidence from the Balancing Market in England and
Wales Kevin F. Forbes The Catholic University of America
Washington, DC 20064 USA [email protected] O.C. St. Cyr Department of
Physics The Catholic University of America and NASA-Goddard Space
Flight Center 32 nd USAEE/IAEE North American Conference Anchorage,
Alaska July 2013 Research Supported by the National Science
Foundation
Slide 2
Electricity is produced at relatively low voltages but is
generally transported at high voltages so as to reduce transmission
losses Source: U.S.-Canada Power System Outage Task Force
Slide 3
Retail expenditures on electricity were approximately $370
billion in 2011, the most recent year for which data are available.
The economic contribution of this industry is much higher than this
because electricity is critical to almost everything we do and thus
gives rise to a very large consumer surplus
Slide 4
Area A Represents consumer expenditures on electricity Area B
represents the consumer surplus that society receives from
electricity B A Hypothetical Demand for Electricity Quantity of
Electricity Price P1P1 Q1Q1 A Hypothetical Demand Function for
Electricity, Expenditures on Electricity, and the Consumer Surplus
from Electricity.
Slide 5
While the wholesale price of electricity may be about $40 per
megawatt-hour, the economic literature reports that the value of
lost load is about $5,000- $10,000 per megawatt-hour Indicative of
the high value of lost load, it is not unheard of for the real-time
price of electricity in todays restructured electricity industry to
be close to $1,000 per MWh.
Slide 6
Slide 7
The power system is almost exclusively an alternating current
system There is a target level of system frequency, i.e. a desired
level of voltage and current oscillations each second. The desired
level of system frequency is 50 times per second in most of the
world and 60 times per second in North America. Maintaining the
desired levels of frequency requires that electricity supply equal
demand at all times, not just on average.
Slide 8
System frequency falls when demand exceeds supply and rises
when demand is less than supply. In either case, reliability is
compromised. System operators offset these electricity imbalances
by dispatching balancing power. These deployments can be large in
magnitude
Slide 9
Slide 10
Slide 11
Geomagenetic storms are disturbances in the Earths magnetic
field that are largely caused by explosions in the Suns corona that
spew out solar particles.
Slide 12
Source: NASA
Slide 13
Power Grids are vulnerable to geomagnetic Storms because the
power transmission grid acts as an antenna of sorts, picking up
geomagnetically induced currents (GICs). These currents have the
potential to impair the performance of transformers
Slide 14
Slide 15
Source: Alan Thompson of the BGS
Slide 16
GICs have been found to be statistically related with various
measures of real-time operations in 12 power grids including PJM,
NYISO, New England, England and Wales, New Zealand, Australia,
Ireland, and the Netherlands. It may also be relevant to note that
the Hydro-Quebec system collapsed in 1989 during a geomagnetic
storm.
Slide 17
The Day-Ahead and Real-Time Reference Price in the New York
ISO, January 1-31 2005
Slide 18
Slide 19
Net Imbalance Volume (NIV) the quantity of electricity that the
system operator uses to balance the system. Positive NIV values for
a market period indicate that the system was short Negative NIV
values indicate that there was excess supply
Slide 20
NIV tends to be negative because market participants are
significantly penalized for being short
Slide 21
Slide 22
Ambient Temperature The GIC proxy Explanatory variables that
reflect expected operating conditions such as forecasted load, the
level of scheduled generation relative to forecasted load, and
scheduled imports and exports Binary variables for the hour of the
day and the day of the week
Slide 23
The model was first estimated using ordinary least squares. A
number of the coefficients are highly statistically significant.
Unfortunately, the OLS residuals are highly autocorrelated which
renders the results open to question.
Slide 24
Slide 25
The goal of this estimation is to achieve white noise in the
residuals. There is no reason to have any confidence in the
estimates in the absence of white noise,
Slide 26
Slide 27
The GIC/NIV relationship is highly statistically significant.
The relationship is highly contingent on terrestrial based system
conditions. In short, the GIC/NIV relationship is more robust the
smaller the level of available generation relative to load.
Slide 28
Slide 29
The out of sample evaluation period: 1 Jan 31 March 2005 The
control area was expanded to include Scotland on 1 April 2005.
Slide 30
Slide 31
The forecast is less accurate if the estimated effect of the
GICs is removed from the forecast equation
Slide 32
The research reported here strongly supports the view that
space weather had electricity market effects during solar cycle 23
even though there is no published evidence of a major space weather
induced blackout. The research also indicates that electricity
market imbalances have a degree of predictability. Thus, it may not
insurmountable for a system operator to forecast the terrestrial
based vulnerability of its power system. Such forecasts may have
the potential to enhance reliability even when the role of space
weather is minor.