The Coincidence of Wind

The cyclical nature of customer demand on large scale interconnected networks is a well known phenomena.  Demand varies by time of day and responds to many factors that influence electric usage, including weather, seasonal activities and business cycles. Composite electric load generally behaves in a cyclical fashion for periods of a day and a year.  With the influx of larger amounts of  wind power, another cyclical characteristic is applied to the power system, that of the available wind generation.  Wind that drives turbines for wind farms varies continuously and generally behaves in a cyclical fashion for periods of a day and a year, just like demand.  The output capacity of a wind farm varies according to the prevailing wind.

Whereas demand tends to peak in either winter or summer, wind capacity tends to peak in spring and fall.  Furthermore, while demand tends to be highest at the hottest time of the day (for summer peaking areas), wind capacity tends to be lower during hot and sunny daylight hours.

Both demand and wind capacity have an impact on the thermal loading of transmission systems, and the non-coincidence of their cyclical behavior leads to interesting transmission usage patterns.

A Closer Look at the Cycles

Figure 1 shows the hourly variation in demand (blue) and wind capacity (red) for a sample system as measured over a one year period.  Demand is measured in percent of peak load while wind capacity is based on a 150 MW (installed capacity) wind farm.  (Note: click on the thumbnail to enlarge the picture.) The capacity factor for the demand (equal to the total energy divided by peak load and total hours) is 31%; while that for the wind capacity (equal to the energy delivered divided by installed capacity and total hours) is 40%.

Figure 1

Figure 2 shows the trend lines for the annual variation of demand and wind capacity.  While demand peaks in the summer months of July and August, wind capacity is lowest for these months.  In the spring month of April, wind capacity is highest while demand is lowest.

Figure 2

Figure 3 shows the average hourly demand and wind capacity in July for one day in the sample system.  For a summer peaking system, there may be direct correlation to solar heating for both demand and wind capacity.  As solar heating increases during the day, demand rises while wind capacity dips.  (Solar power offers the benefit of increasing output with demand and may represent a good complement to wind resources.)  In contrast, Figure 4 shows the daily averages for the month of April.  Here we see that wind capacity peaks in the twilight hours and dips in the afternoon.  The demand is relatively flat throughout the daytime.

Figure 3    Figure 4

Both the annual and summer daily cycles seem to indicate an inverse relationship between demand and wind capacity.  However, since both demand and wind patterns vary significantly from region to region, determining the appropriate relationship remains a case-to-case assessment.

The Use of Transmission

Now, both demand and wind capacity have a direct impact on thermal loading of the transmission system.  As load increases, the use of the grid increases, likewise with wind capacity, as it increases, flows on the grid also increase.  However, the thermal load of specific transmission lines and transformers is a function of the location of demand centers, wind resources as well as other resources on the transmission network.  Again, it is hard to generalize.

For discussion purposes, let us assume that flow of power on a specific line increases in direct proportion to demand, and that a wind farm would contribute 40% of its output to this line, then a graph of thermal loading for this specific line on a typical summer day is as shown in Figure 5.  The thermal load on the line (shown in purple) follows the same characteristic as demand shown in Figure 3.  The output of the wind farm increases the loading slightly as shown by the green plot in Figure 5..  However, given a line capacity of 250 MW, it requires very little additional power flow from the wind farm to take the flow on the line over the 250 MW rating.  The result is a potential for overload on the specific line around 6 PM.

Figure 6

Doing the same analysis for an April day, produces Figure 7.  Here, wind capacity is highest but system demand is not.  The net effect on the specific transmission line is only moderate thermal loading.

Figure 7

At higher levels of wind penetration, the component of transmission line loading that is attributable to wind also increases.  This may change the summer and spring characteristics shown in Figures 6 and 7 significantly.

So far, this has been a very simplistic approach to transmission loading.  In reality, many other factors impact line loading.  Some examples are:

  • Changes in generation dispatch as demand increases.  Peaking units may come online that are closer to the load centers.
  • System voltage and stability may require out-of-merit order generators to be dispatched at different demand levels.
  • In interconnected systems, the impact of flow-through power, inadvertent flows and parallel flows could influence local line loadings.
  • Different types of load may ramp up during the day at different rates, i.e., industrial and residential loads.
  • First contingency and reserve operating rules may impose dispatch adjustments that are a function of demand level.

Analyzing the Wind and Demand Coincidence Impact on Transmission

There are a number of tools to simulate the varying nature of demand and wind capacity in an interconnected electrical network.  Power flow based production simulators, such as GE MAPS, ABB’s Gridview and U-PLAN, may be adopted to this purpose.  However, these methods are both data and computation extensive.  Also, these methods may imply a determinism to the results for what is essential a probabilistic process.

A planning approach of testing for worst case conditions allows the use of power flows with acceptable results.  By testing conditions at the potential extremes of demand and wind capacity and some combination in-between, it is possible to identify the range of loading exposure for specific elements of the transmission system.  The “umbrella principle” would then apply – “if we plan for the worst conditions, we would weather all other conditions.”

One Final Coincidence …

One more idea before we close this Techblog.  Transmission line capacity is also a variable.  As wind and solar heating changes during a 24-hour period, the current-carrying capacity of the line also changes.  Though we are using to seeing fixed ratings for overhead transmission lines, there is in fact some variability to the rating.  The use of dynamic line ratings takes advantage if this phenomena.

The thermal capacity of a line increases as wind blows across it (to help in cooling).  Hence, there is a near direct correlation between transmission capacity and wind generation capacity.  As the wind picks up, more power is available from wind turbines and transmission lines are able to carry more load.  How about that for another coincidence, and perhaps a future Techblog article?