forecasting
What is forecasting and why does it matter for power grid planning?
Forecasting plays a crucial role in the long-term planning of power grids, as the energy landscape continues to evolve rapidly.
Forecasting in terms of trying to predict events goes back to ancient times. Egyptians, Greeks, and Mayans, relied on oracles, astrologers, and religious figures to predict the future (maybe with questionable levels of success..).
Today, forecasting affects the everyday of most people around the world. And most of us are well aware of those times when the weather forecast said no rain, we left our home without an umbrella, yet the sky still opened up and reminded us of who the boss is.
But how often do we reflect on what a forecast is and the meaning of it?
There are many slightly different definitions of what forecasting is, but most commonly definitions can be summarized as: making predictions of future events or trends. We can use more or less complex models and methods to improve these predictions, reduce error margins and uncertainty. We can ask experts or build models. But the outcome is always about predicting an event, or a trend. However, this description of a forecast turns out to be limited in some cases.
Let’s demonstrate this with an example.
Imagine that you are tossing a magic coin. The coin has two sides: heads and tails. Yet since it’s magic it doesn’t adhere to any rules or logic. So it’s impossible to predict what a toss will be.
Now, assume you have two scenarios.
In scenario 1, you know nothing about the outcomes from the coin toss.
In scenario 2, you know that if it lands heads up, it will be windy and –20 degrees Celsius. If it lands tails up, it will be sunny and 20 degrees Celsius.
Because of our magic coin, we can’t make a prediction about the outcome. As such from a forecasting perspective, scenario 1 and 2 are considered equal. Yet, even if scenario 1 and 2 are equal from a forecasting perspective, most sensible humans would prefer scenario 2 rather than scenario 1. Why? In scenario 2 you know the outcomes, even if you don’t know the probability. We can summarize this as scenario 2 contains more information. This information is not about the likelihood of the future, but information about one possible future. Knowing the outcomes, we can plan accordingly (like packing some extra warm clothes).
Now, the world isn’t black and white. But, at Endre we believe that the goal of forecasting in planning should be to maximize the amount of information, not simply calculating the probability of events. This becomes especially important when we consider long-time horizons were the uncertainty of predicting specific events is very large.
Why does it matter?
In long-term power grid planning we can separate the probability that a future will happen, and information about what that future would look like. Methods, tools and datasets needed for answering “will there be many electric vehicles in my grid?” are very different from the methods, tools and datasets needed to answer, “how are these electric vehicles expected to behave?”
With more and more data available in today’s society, together with more advance technologies and models (such as AI), answering question two is still hard but is becoming more feasible. And in the future of long-term power grid planning, that is absolutely crucial! A recent report from EY, Grids for Speed, points out that making grid investments based on long-term forecasts of demand and production is the most cost-effective grid strategy.
Are you interested in reducing risks and uncertainties of your future power grid related to the rapidly changing energy landscape. Don’t hesitate to reach out to us and stay tuned for more content on this important topic!
Forecasting plays a crucial role in the long-term planning of power grids, as the energy landscape continues to evolve rapidly.
Forecasting in terms of trying to predict events goes back to ancient times. Egyptians, Greeks, and Mayans, relied on oracles, astrologers, and religious figures to predict the future (maybe with questionable levels of success..).
Today, forecasting affects the everyday of most people around the world. And most of us are well aware of those times when the weather forecast said no rain, we left our home without an umbrella, yet the sky still opened up and reminded us of who the boss is.
But how often do we reflect on what a forecast is and the meaning of it?
There are many slightly different definitions of what forecasting is, but most commonly definitions can be summarized as: making predictions of future events or trends. We can use more or less complex models and methods to improve these predictions, reduce error margins and uncertainty. We can ask experts or build models. But the outcome is always about predicting an event, or a trend. However, this description of a forecast turns out to be limited in some cases.
Let’s demonstrate this with an example.
Imagine that you are tossing a magic coin. The coin has two sides: heads and tails. Yet since it’s magic it doesn’t adhere to any rules or logic. So it’s impossible to predict what a toss will be.
Now, assume you have two scenarios.
In scenario 1, you know nothing about the outcomes from the coin toss.
In scenario 2, you know that if it lands heads up, it will be windy and –20 degrees Celsius. If it lands tails up, it will be sunny and 20 degrees Celsius.
Because of our magic coin, we can’t make a prediction about the outcome. As such from a forecasting perspective, scenario 1 and 2 are considered equal. Yet, even if scenario 1 and 2 are equal from a forecasting perspective, most sensible humans would prefer scenario 2 rather than scenario 1. Why? In scenario 2 you know the outcomes, even if you don’t know the probability. We can summarize this as scenario 2 contains more information. This information is not about the likelihood of the future, but information about one possible future. Knowing the outcomes, we can plan accordingly (like packing some extra warm clothes).
Now, the world isn’t black and white. But, at Endre we believe that the goal of forecasting in planning should be to maximize the amount of information, not simply calculating the probability of events. This becomes especially important when we consider long-time horizons were the uncertainty of predicting specific events is very large.
Why does it matter?
In long-term power grid planning we can separate the probability that a future will happen, and information about what that future would look like. Methods, tools and datasets needed for answering “will there be many electric vehicles in my grid?” are very different from the methods, tools and datasets needed to answer, “how are these electric vehicles expected to behave?”
With more and more data available in today’s society, together with more advance technologies and models (such as AI), answering question two is still hard but is becoming more feasible. And in the future of long-term power grid planning, that is absolutely crucial! A recent report from EY, Grids for Speed, points out that making grid investments based on long-term forecasts of demand and production is the most cost-effective grid strategy.
Are you interested in reducing risks and uncertainties of your future power grid related to the rapidly changing energy landscape. Don’t hesitate to reach out to us and stay tuned for more content on this important topic!
Forecasting plays a crucial role in the long-term planning of power grids, as the energy landscape continues to evolve rapidly.
Forecasting in terms of trying to predict events goes back to ancient times. Egyptians, Greeks, and Mayans, relied on oracles, astrologers, and religious figures to predict the future (maybe with questionable levels of success..).
Today, forecasting affects the everyday of most people around the world. And most of us are well aware of those times when the weather forecast said no rain, we left our home without an umbrella, yet the sky still opened up and reminded us of who the boss is.
But how often do we reflect on what a forecast is and the meaning of it?
There are many slightly different definitions of what forecasting is, but most commonly definitions can be summarized as: making predictions of future events or trends. We can use more or less complex models and methods to improve these predictions, reduce error margins and uncertainty. We can ask experts or build models. But the outcome is always about predicting an event, or a trend. However, this description of a forecast turns out to be limited in some cases.
Let’s demonstrate this with an example.
Imagine that you are tossing a magic coin. The coin has two sides: heads and tails. Yet since it’s magic it doesn’t adhere to any rules or logic. So it’s impossible to predict what a toss will be.
Now, assume you have two scenarios.
In scenario 1, you know nothing about the outcomes from the coin toss.
In scenario 2, you know that if it lands heads up, it will be windy and –20 degrees Celsius. If it lands tails up, it will be sunny and 20 degrees Celsius.
Because of our magic coin, we can’t make a prediction about the outcome. As such from a forecasting perspective, scenario 1 and 2 are considered equal. Yet, even if scenario 1 and 2 are equal from a forecasting perspective, most sensible humans would prefer scenario 2 rather than scenario 1. Why? In scenario 2 you know the outcomes, even if you don’t know the probability. We can summarize this as scenario 2 contains more information. This information is not about the likelihood of the future, but information about one possible future. Knowing the outcomes, we can plan accordingly (like packing some extra warm clothes).
Now, the world isn’t black and white. But, at Endre we believe that the goal of forecasting in planning should be to maximize the amount of information, not simply calculating the probability of events. This becomes especially important when we consider long-time horizons were the uncertainty of predicting specific events is very large.
Why does it matter?
In long-term power grid planning we can separate the probability that a future will happen, and information about what that future would look like. Methods, tools and datasets needed for answering “will there be many electric vehicles in my grid?” are very different from the methods, tools and datasets needed to answer, “how are these electric vehicles expected to behave?”
With more and more data available in today’s society, together with more advance technologies and models (such as AI), answering question two is still hard but is becoming more feasible. And in the future of long-term power grid planning, that is absolutely crucial! A recent report from EY, Grids for Speed, points out that making grid investments based on long-term forecasts of demand and production is the most cost-effective grid strategy.
Are you interested in reducing risks and uncertainties of your future power grid related to the rapidly changing energy landscape. Don’t hesitate to reach out to us and stay tuned for more content on this important topic!
Forecasting plays a crucial role in the long-term planning of power grids, as the energy landscape continues to evolve rapidly.
Forecasting in terms of trying to predict events goes back to ancient times. Egyptians, Greeks, and Mayans, relied on oracles, astrologers, and religious figures to predict the future (maybe with questionable levels of success..).
Today, forecasting affects the everyday of most people around the world. And most of us are well aware of those times when the weather forecast said no rain, we left our home without an umbrella, yet the sky still opened up and reminded us of who the boss is.
But how often do we reflect on what a forecast is and the meaning of it?
There are many slightly different definitions of what forecasting is, but most commonly definitions can be summarized as: making predictions of future events or trends. We can use more or less complex models and methods to improve these predictions, reduce error margins and uncertainty. We can ask experts or build models. But the outcome is always about predicting an event, or a trend. However, this description of a forecast turns out to be limited in some cases.
Let’s demonstrate this with an example.
Imagine that you are tossing a magic coin. The coin has two sides: heads and tails. Yet since it’s magic it doesn’t adhere to any rules or logic. So it’s impossible to predict what a toss will be.
Now, assume you have two scenarios.
In scenario 1, you know nothing about the outcomes from the coin toss.
In scenario 2, you know that if it lands heads up, it will be windy and –20 degrees Celsius. If it lands tails up, it will be sunny and 20 degrees Celsius.
Because of our magic coin, we can’t make a prediction about the outcome. As such from a forecasting perspective, scenario 1 and 2 are considered equal. Yet, even if scenario 1 and 2 are equal from a forecasting perspective, most sensible humans would prefer scenario 2 rather than scenario 1. Why? In scenario 2 you know the outcomes, even if you don’t know the probability. We can summarize this as scenario 2 contains more information. This information is not about the likelihood of the future, but information about one possible future. Knowing the outcomes, we can plan accordingly (like packing some extra warm clothes).
Now, the world isn’t black and white. But, at Endre we believe that the goal of forecasting in planning should be to maximize the amount of information, not simply calculating the probability of events. This becomes especially important when we consider long-time horizons were the uncertainty of predicting specific events is very large.
Why does it matter?
In long-term power grid planning we can separate the probability that a future will happen, and information about what that future would look like. Methods, tools and datasets needed for answering “will there be many electric vehicles in my grid?” are very different from the methods, tools and datasets needed to answer, “how are these electric vehicles expected to behave?”
With more and more data available in today’s society, together with more advance technologies and models (such as AI), answering question two is still hard but is becoming more feasible. And in the future of long-term power grid planning, that is absolutely crucial! A recent report from EY, Grids for Speed, points out that making grid investments based on long-term forecasts of demand and production is the most cost-effective grid strategy.
Are you interested in reducing risks and uncertainties of your future power grid related to the rapidly changing energy landscape. Don’t hesitate to reach out to us and stay tuned for more content on this important topic!
Smarter grid planning using societal insights
Email hello@endre.tech
Visit us at Läraregatan 3, 411 33 Gothenburg, Sweden
Post address Stena Center, 412 92 Gothenburg, Sweden
Copyright 2025 | Endre Technologies