Interview: How Mathematical Optimization Can Help Power Today’s Energy Industry Transformation

Date: 7/26/2021

The energy industry is in the midst of a major transformation as it shifts away from fossil fuels towards renewable energy sources including wind, solar, and hydropower.

Although the industry landscape is rapidly changing with the introduction of new strategic initiatives and innovations and the rise of new challenges (including managing supply security, demand increase, and environmental sustainability), one thing is constant: Mathematical optimization – which has been used for decades by all the key players in the energy market – remains an absolutely critical technology for energy enterprises today.

Mathematical optimization empowers energy companies to make data-driven, optimal decisions on how to integrate their networks and utilize their resources – so that they can deliver energy to consumers in the most efficient, affordable, sustainable, and reliable manner possible.

Researchers from the academic arena are collaborating closely with their industry partners to spearhead greater innovation and utilization of mathematical optimization to address today’s energy market challenges.

One academic who is at the forefront of this initiative to develop and deploy new mathematical optimization technologies is Dr. Alireza Soroudi, an Assistant Professor in Power Systems at University College Dublin’s School of Electrical & Electronic Engineering as well as a Senior Member of the Institute of Electrical and Electronics Engineers (IEEE).

Recently we spoke with Dr. Soroudi to learn about his work and get his insights on how mathematical optimization can help power today’s energy industry transformation.

Gurobi (G): Please tell us about your background and your work.

Alireza Soroudi (AS): I did my PhD in Electrical Engineering at the Grenoble Institute of Technology (Grenoble INP) in France in 2012, and was the recipient of the INFORMS Young Researcher Prize in 2013 for my work on optimization under uncertainty (when modeling the impact of renewable energies).

I’m an Assistant Professor in the University College Dublin’s School of Electrical & Electronic Engineering (where I’m currently teaching a module called “Optimization Techniques for Engineers) and a Senior Member of the IEEE. Additionally, I serve as an associate editor for two journals: IET Smart Grid and IET Generation, Transmission, and Distribution.

My primary research interest is the optimal design and operation of power systems, especially given the presence and increasing penetration of renewable energy technologies.

G: How did you get interested in mathematical optimization (and specifically in how this AI technology is used in the energy industry)?

AS: The first time I really became interested in the concept of mathematical optimization was during my PhD thesis in 2008, when I was trying to assess the impact of small-scale renewable energy technologies on distribution.

After that, I started researching and discovering how mathematical optimization techniques and tools, like solvers, can be used to plan and operate the power systems.

Indeed, mathematical optimization techniques and tools can help us answer (and ask) some really important questions in the energy industry today, and address some of the industry’s most critical and challenging problems such as energy storage, power flow, unit commitment, transmission network planning and optimization, and energy system integration. In 2017, I wrote a book called Power System Optimization Modeling in GAMS, which provided an overview of these and other applications.

Beyond the energy industry, I’m very interested in exploring how mathematical optimization (and in particular Python-based tools for solving optimization problems) can be leveraged to tackle problems in other industries including supply chain, aviation, and healthcare (especially recently with COVID-19 vaccine allocation and capacity utilization problems).

I also love using mathematical optimization to solve problems related to games, like chess. And, of course, I have a deep passion for teaching others how to use mathematical optimization, and opening their eyes to the power of this technology.

G: Why do you think mathematical optimization is such a powerful AI problem-solving technology?

AS: Simply put, mathematical optimization can give you the capability to make more efficient use of your limited, critical resources. So, if you are facing a business problem where you have limited resources and you want to achieve certain business objectives (such as maximizing profitability or minimizing delays or risk) while taking into account your business constraints, then mathematical optimization is the right technology for you.

In the energy industry, there are numerous key business objectives: We want to deliver services to consumers at the lowest possible cost, while at the same time we have to ensure compliance with environmental regulations and fossil fuel emission targets and also ensure the security of energy supply – and mathematical optimization can help us simultaneously achieve all these (and other) business objectives and supply energy in a resilient and secure way.

G: What do you see as the most significant business challenges in the energy industry today?

AS: The energy industry is in the middle of a period of transformation. With the European Green Deal, the EU has put in place some really ambitious targets in place: Reducing net greenhouse gas emissions by at least 55% by 2030 compared to 1990 levels, and eliminating those emissions entirely by 2050. But this is not going to happen without the proper utilization and integration of the energy systems.

Indeed, as the penetration of renewable energy sources in the market increases, new challenges are appearing. One of these is definitely energy system integration.

For example, electric vehicles (EVs) need to be charged, but if too many EV owners are charging their vehicles at the same time, that can put strain on the system. Or if the number of solar panels increases dramatically, that can create voltage problems to the system.

We need to figure out how to address these challenges – and mathematical optimization can help us with that. Currently I’m involved in a project with the Energy Systems Integration Partnership Program (ESIPP), where we using mathematical optimization techniques and tools to help us figure out how to manage the interactions between different energy sectors (including electricity, gas, heat, hydrogen, and water), coordinate and exploit the flexibilities in different sectors, and improve the efficiency of our existing infrastructure so that it can host more renewable technologies (for example, by utilizing available wind capacity to supply green hydrogen for future transportation or by creating hydrogen and injecting it into the gas network).

Another major challenge that we are facing – with the introduction of all these renewable energy technologies – is attaining and maintaining control over our energy networks. In general, the controllability of these new technologies is lower than fossil fuel-based technologies (you can’t, for example, tell a wind turbine to turn when there’s no wind!).

These new technologies need new control and communication infrastructures. Mathematical optimization can help us here, by enabling distributed control and optimization across the network – so that energy providers can efficiently and effectively react to real-time signals and fluctuations in supply and demand.

Other important challenges that we are dealing with in the energy industry today include:

• Enhancing the security and resiliency of the energy system,
• Coping with increased demand volatility,
• Figuring out how to store electricity (as it’s expensive to use batteries for large-scale systems) and where to locate new facilities, and
• Controlling energy costs for consumers.

Mathematical optimization definitely has a pivotal role to play in helping us address these challenges and ensuring our energy systems operate in the most efficient, sustainable, affordable, and secure way.

G: It sounds like mathematical optimization is an essential technology in today’s renewable energy revolution.

AS: Indeed, it is. When policymakers formulate green energy plans and policies, they set emissions targets for the future, but they don’t specify exactly what should be done to reach those targets.

As engineers, we have to figure out the technical aspects of these plans and policies, and determine which strategies and tools we need to operate these new energy systems and use our energy resources as efficiently as possible.

Without mathematical optimization, this energy industry transformation is simply not going to happen. Mathematical optimization is an essential technology that helps us to make optimal decisions when it comes to facility location, energy storage, energy system integration, and many other challenges.

G: Thinking about your career as an academic and researcher, what do you see as your main accomplishment?

AS: I am striving to distribute the knowledge of mathematical optimization between academia, researchers, and industry. This will help them explore new, uncharted areas.

On the education side, I want to teach students (as well as professionals) how to use mathematical optimization technologies, and apply them in the real-world.

On the research side, my focus is on working with industry partners to discover new techniques and approaches to using mathematical optimization, and to build and deploy mathematical optimization solutions that enable energy providers to deliver power to consumers in the most cost-efficient, clean, and reliable manner.