Safely operating power lines during hot summer days

Safety is the number one concern for all transmission service providers. Unfortunately accidents do happen despite the best efforts to prevent them. One of the factors that contributes to accidents is the fact that transmission lines may dangerously overheat on hot, sunny, windless days.

Power lines may overheat when there's no wind

Every transmission line has a maximum operating temperature that it can tolerate. But line temperature is not measured in 99% of high-voltage transmission lines. Instead, grid operators rely on calculations to determine how much power a line can handle.

These calculations are done according to international standards. The ability of a power line to carry energy depends on the type of conductor as well as the weather conditions. Since weather is variable, the standard approach is to select certain agreed-upon conservative assumptions. The key assumption in these calculations is that wind blows at 0.61 m/s onto the wires.

Under most circumstances, a wind speed of 0.61 m/s is very conservative. But power lines cover entire countries and some parts of a line may be in forested areas or inside valleys. It is those parts of a power line which are sheltered from the wind which are most at risk of overheating.

Importantly, if the actual wind is in fact 0 m/s, then the capacity of a line may be one third less than the international standard specifies. So there is a risk that a power line is overloaded by one third.

As a power line heats up the wires expand and drop closer to the ground. During normal operation the wires may move up and down by several meters. Normally this is not a problem, since each power line is designed to be high enough from the ground in the first place, with proper safety margins.

However, if there has been an error during design, or if the power line is old and has deformed over time, then these safety margins might already be breached. These safety concerns are exacerbated by the calculation method that is used for determining line capacity.

A 132,000 volt power line can electrocute an object from a few meters away even if it's not in physical contact directly. If there happens to be something underneath the line when it overheats and sags too close to the ground then there is a risk of a flashover. For example if a tree is too close to the line then it might spark a wildfire.

FERC Order 881 may lead to power lines overheating

The assumption that there is a 0.61 m/s wind blowing onto the conductor and cooling it down is overly optimistic at least 3% of the time (see for example WATT). To illustrate this let’s look at Estonia. It’s a cold and windy place, almost as far up north as Alaska. Yet even here the overheating of power lines is a problem. Air temperature goes above 25°C / 77°F only about 2.5% of the time, but when it does it’s usually sunny. Half the country is forested, so power lines are often sheltered by trees. The assumption that wind blows at 0.61 m/s is overly optimistic in these cases and flashovers do happen, especially under older power lines inside forests.

In the US grid operators are mandated by FERC to implement so-called Ambient Adjusted Ratings (AAR) by July, 2025. This means that grid operators must use ambient conditions, such as actual air temperature, when calculating the ratings of their power lines. However, AAR solutions still typically make the assumption that wind is a constant 0.61 m/s. Therefore the risk of overheating on hot summer days remains.

Accurate wind predictions are needed to ensure safety

Grid operators need accurate wind information in order to remove the risk of lines overheating during windless hours. But it is not feasible to install physical sensors under every section of every power line in the country. Grid Raven is solving this problem by employing machine learning for modeling wind with meter-scale accuracy. This covers every section of all power lines and provides an accurate view of the network to grid operators.

Until Grid Raven’s technology has been implemented everywhere, when you’re walking in a forest on a hot, sunny day and you see a high-voltage line that is suspiciously close to the ground, then it might in fact be so. Don’t raise your hands while walking underneath it!

Grid Raven's AAR solution takes wind into account. Read more about it here: https://www.gridraven.com/ferc-881

‍

Accurate line ratings with the example of Texas

The cost of grid congestion in the US was $20 billion in 2022, $3 billion of which occurred in Texas (source). The grid is under increasing stress. It is already commonplace that zero marginal cost renewable energy is curtailed because the grid cannot transport the power. This raises prices for consumers next door, where fuel-burning power plants are kept in operation.

Yet in fact, much of this congestion is artificial and self-imposed due to the methods that grid operators use for calculating grid capacity. The existing grid could transport on average a third more power, if operators accounted for the power line cooling effect from actual air temperature and especially wind. A third more capacity would significantly alleviate congestion and reduce energy prices, but this must be done in a safe and secure manner.

Zooming in on a power line near San Antonio

Below you can see a map showing most of the high-voltage power lines in Texas (top left corner, image below). Each line is color coded according to its capacity over the next 48 hours, depending on actual weather conditions (from HRRR). Blue means that the line is experiencing little wind and high temperatures. Yellow means the line has in fact roughly twice the capacity compared to what the grid operators use today. The top right corner shows a selected power line near San Antonio, and the bottom graph shows the actual capacity of this line, calculated according to three different approaches.

Traditionally, power lines have a static rating throughout the year, which is calculated by assuming the conditions of a hot summer day with no wind. While this approach has served us very well in the past it's no longer sufficient, since the energy system is rapidly changing.

Ambient adjusted ratings use the actual forecasted air temperature as a key input. It's relatively straightforward to predict air temperature with an accuracy of a few degrees. This is a step in the right direction, but the additional benefit is marginal, as can be seen in the case of the power line near San Antonio. The green line (ambient ratings) is often not much above the orange line (static ratings).

The greatest benefit comes from the wind cooling effect. Dynamic line ratings use the wind forecast and often lead to twice the capacity (the dark green line). But it's because of the difficulty of accurately predicting wind that dynamic ratings are not yet widely used.

Reduce capacity when there’s no wind at all

Notably, dynamic line ratings are also safer. In 45 hours' time after I'm writing this blog post, the wind will be completely still near San Antonio. If this power line happens to be fully loaded at this time, then there would be a risk of overheating of the conductor. It might then sag too close to the ground and if something happens to be underneath then there's a risk of electrocution or sparking a wildfire. During such hours line capacity should be reduced to ensure safety.

The technology for implementing accurate line ratings is available today. This helps ensure the safety of our network, while at the same time increasing grid capacity, allowing more clean power onto the grid and reducing energy costs.

Read more about Grid Raven's AAR solution here: https://www.gridraven.com/ferc-881.

Texas map from Wikimedia

The Weakest Link

Just like the proverbial chain an entire transmission line can be limited by its weakest link. But identifying the weakest link is not trivial, since it depends on the weather among other factors.

The span with the least wind

The weakest link of a high-voltage line is the section which reaches its maximum operating temperature first. It is determined by the wind in the specific location, since wind at a ninety degree angle cools the wires most effectively. In contrast, when wind blows along the wire then the cooling effect is significantly reduced.

The illustrations in this post show a case that we stumbled upon where the Dynamic Line Rating of an entire line is limited by a single span. In the shown hour wind happens to blow in parallel to a short section of the line. But the rest of the line experiences a significant wind cooling effect and could transport at least 50% more power.

Under Static Line Ratings when a single span limits an entire line it would make sense to refurbish this one span by for example lifting the towers and thus increasing its maximum operating temperature. But in this example the same span is not the weakest link when wind blows from different directions. The weakest link moves around depending on weather conditions, so refurbishing would not be a solution here, at least not most of the time.

Dynamic Line Ratings need accurate wind forecasts

This example shows the level of detail that Dynamic Line Rating solutions must contain. If the line were carelessly loaded by an additional 50% then there would be a risk of overheating that one span and causing structural damage, sparking a wildfire or worse. Dynamic Line Rating solutions must account with the wind in each span individually.

The Risk of Losing Capacity with FERC 881

FERC's intention is to increase grid capacity

FERC Order 881 in the US obliges all transmission owners to implement Ambient Adjusted Ratings (AAR) by July 2025. Utilities must move away from using Static Line Ratings (SLR), in which grid capacity is based on assuming constant values, and must account for the grid cooling effect from air temperature.

The intention of the order is to reduce power prices for consumers by increasing grid capacity: Chairman Glick said: “If we are going to [keep] customer rates just and reasonable [...], we need to squeeze everything out of our existing grid,” (source ferc.gov)

In the winter when it’s cool outside, accounting with temperature does lead to an increase in capacity. The red line (AAR) is above the static ratings (SLR) most of the time in January, mainly because it’s colder outside. The grid can carry more power. The below image shows line ratings in one week in January in 2022, based on weather data from the Pioneer Airfield, Arizona.

Gain in capacity is not always guaranteed

Whether AAR leads to a gain or loss depends on what the assumptions were that the utility used in calculating ratings in the first place. Here we illustrate this based on a hypothetical example from Arizona.

In May, ambient ratings (AAR) are consistently below the static ratings (SLR) during the day, as outside air temperatures reach or exceed 100 degrees Fahrenheit.

Implementing AAR in this example would lead to a loss of capacity a whopping 40% of the time on an annual basis. It’s especially ironic that the loss occurs exactly when additional capacity would be needed - during the day when industry and air conditioning is turned on!

Utilities have been optimizing already for decades

The reason this happens is that many transmission owners have been pushing their grids to the maximum for decades already and might knowingly have been using optimistic assumptions for wind speed in calculating their static line ratings. 

In the example shown above, the static ratings were calculated using a wind speed of 4.4 ft/second perpendicular to the conductor (1.34 m/s). The implicit assumption by the grid operator has been that the excessive benefit from high wind speeds is compensated for by the margin from a conservative assumption for air temperature.

Now, however, FERC 881 requires the actual air temperature to be used, thus removing this buffer. Without this buffer utilities are now forced to revisit and reduce the wind speed assumption in their calculations. Instead of 4.4 ft/second they might use 2 ft/second (0.61 m/s) in AAR. This then leads to a loss of capacity when it’s hot outside.

An easy fix to this would be to take more risk and continue using a relatively high wind speed assumption. But this would run the risk of the wires actually overheating, sagging too close to the ground, causing a flashover and igniting a fire.

Dynamic Line Ratings help increase capacity while maintaining safety

A more customized solution is to implement Dynamic Line Ratings right away. Grid Raven’s wind forecast machine learning model, Boreas, covers the full network and helps utilities use accurate wind speeds in their calculations. The below graph shows the same week in May, with Dynamic Line Ratings (DLR) shown, in which actual wind speeds were taken into account (green line). Dynamic Line Ratings can help avoid safety problems while maximizing grid capacity.

‍

‍