In this SustMeme Guest Post, Marshall Moutenot, CEO of Upstream Tech, examines the core challenges faced by hydropower as a consequence of climate change and explores how tech and AI can help it adapt.
MM: “The era of procrastination, of half-measures, of soothing and baffling expedients, of delays, is coming to its close. In its place we are entering a period of consequences.” — Winston Churchill, Nov 12, 1936
Today, we face a new period of uncertainty, this time brought about by anthropogenic climate change.
Yet, it strikes me in considering Churchill’s words, which were uttered on the eve of a traumatic global conflict, that we similarly find ourselves grappling with a new era of consequences.
In hydropower, these consequences are already posing three distinct operational challenges: increasing drought; unprecedented storm-surge; and high-precipitation events.
Such challenges now present operational risks to historically stable power-generation assets.
Growing challenge of drought
It is unsurprising to learn that drought is causing water levels in reservoirs to drop, threatening hydropower supply. In British Columbia, Canada, increasingly severe droughts have forced energy utility, BC Hydro, to purchase power from external sources to fulfil contracts and ensure grid reliability, alongside adjusting operational practices to adjust reservoir levels to maximise water retention.
In China, a prolonged dry spell in summer 2024 has dried the Yangtze Basin on which much of the country’s hydropower depends, with the net result that power supply was down by a massive 44 terawatt-hours, resulting in a surge in coal-fired power generation.
Similarly in Zambia, one of the severest droughts in decades – attributed to the El Niño phenomenon – has led to the country’s Kariba dam only being able to produce 7% of its installed capacity, leading to blackouts in towns and cities lasting 3-5 days.
All of these events pose a common challenge to decision making by hydro operators on a seasonal scale. Increasing drought — particularly in the case of China, which was on course for more stable capacity until a drier summer — puts hydropower operators under financial pressure, as grid supply is constrained and revenues fall.
Storm surges and flood risk
While hydropower operators in many parts of the world are battling drought, there are also those that are facing an entirely different challenge: high-precipitation events leading to storm surges and flood risk, threatening the integrity of infrastructure.
In the US, the increasing incidence of named storms has seen a sharp rise in dam safety events, with, most recently, Tennessee Valley Authority (TVA) calling for evacuations in the wake of Hurricane Helene.
TVA warned residents of imminent breach of the Nolichucky Dam as water levels rose rapidly, putting infrastructure under extreme pressure. In the event, the dam held, but other water and wastewater treatment facilities reported damage owing to flooding.
Flood events threaten not only the physical integrity of a dam but also force operators into reactive (and expensive) modes of operation, where emergency measures must be taken to protect assets and minimise the risk of a total catastrophic failure.
In the Nordics, Statkraft, one of Europe’s largest renewable energy producers, has been forced to invest €700M in dam protection against storm surges.
Key role in grid stability
Were they occurring at any other time in recent industrial history, the effects of drought, storm surges and flood events on hydropower facilities would be problematic enough. But, as hydropower increasingly becomes a stabilising base-load and peaking power supply for energy grids with ever-more intermittent renewable technologies like wind and solar, its predictable reliability is paramount.
Hydropower’s ability to quickly adjust its output to balance short-term fluctuations in wind and solar makes it crucial for maintaining stability in grids with increasing renewable energy penetration.
Without the ability to readily support clean-energy grids with flexible on-demand power generation, hydropower assets miss out on a significant revenue stream.
Furthermore, additional renewable energy deployment is compromised, too.
Similarly, while lithium-ion batteries — which are increasingly deployed for grid energy management — are effective for storing energy for several hours, hydropower reservoirs hold water for months or even seasons, providing an enduring supply of potential energy.
This makes hydropower an invaluable complement to wind and solar, ensuring energy needs are continually met during extended periods of low wind or solar radiance, especially in seasonal climates, while offsetting the need for extensive battery storage. Ultimately, this complementarity reduces costs and infrastructure demands while maintaining reliability.
Finally, with proper management, a well-balanced mix of hydropower, wind, and solar energy can match electricity consumption patterns with high levels of accuracy.
This ability to align generation with demand reduces wastage and ensures a steady, dependable energy supply, making hydropower a cornerstone of efficient, renewable energy systems.
Forecasting in times of climate shift
So, it is clear that accuracy in hydropower forecasting is vital not only for the optimal efficiency of power generation, but increasingly operational safety management. However, the models calibrated on historical weather patterns and hydrological response require evolution as the climate shifts.
Where altered weather patterns have led to more intense localised rainfall, for example, a lack of newer models means that without proper forecasts and data, operators are essentially ‘flying blind’.
But, by better understanding what happens upstream of hydropower assets, operators have a longer ‘lead time’ on volumes, giving them more opportunity to make changes to infrastructure operations.
Harnessing AI for predictive management
As artificial intelligence (AI) technologies have proliferated in recent years, public and private institutions have recognised the opportunity they bring to better support hydro forecasting.
As models have advanced, impressive results have been observed for both the short-term horizon for high precipitation and the seasonal horizon for droughts and water supply.
Seeking to capture and understand the power of AI to help solve a climate challenge for US power generation, the Bureau of Reclamation (USBR) ran a streamflow forecasting competition to ascertain the benefits of implementing machine learning into water forecasting.
Knowledge secured in competition enabled Upstream Tech and others to understand where better forecasting impacts most on hydropower operators, energy grids, populations and the environment.
Under high-precipitation events, high-precision forecasts allow operators to predict not only when a storm will hit but also the volume and rate of inflow to reservoirs.
Armed with this information, they can adjust water levels in advance, releasing water strategically to ensure that reservoirs have enough capacity to handle the incoming rainfall.
Such proactive management prevents overflow, reduces pressure on spillways, and minimises the risk of structural damage or flooding downstream.
In cases of severe weather, like hurricanes or intense rainfall, these real-time decisions can be the difference between controlled water flow and dangerous overflows, thereby safeguarding local communities and ensuring the continued reliability of the energy grid.
In the case of drought, improved seasonal forecasting enables operators to make decisions that help conserve water during dry months, adjusting reservoir management to prioritise critical needs and maintain base-load power for the grid.
By having a reliable understanding of their ‘fuel’ for the coming season, operators can strategically balance water retention with controlled releases to meet power generation targets while preparing for prolonged low-water conditions. By optimising this careful balance, they can maintain hydropower’s contribution to the energy mix even when natural water sources are limited.
For grid balancing and management, effective water forecasts enable operators to predict the availability of their water reserves, helping them coordinate hydropower output to complement the intermittent nature of wind and solar generation.
With these forecasts, operators can optimise hydropower output to fill in the energy gaps, ensuring the grid remains stable when other renewable sources cannot meet demand.
This form of grid management not only maximises renewable integration but also minimises the need for additional storage solutions, providing a reliable and efficient pathway toward a cleaner energy grid.
Technology, regulation and planning, together
The importance of hydropower as a climate solution cannot be underestimated.
That said, the consequences of climatic change, with which we now live, threaten the long-term sustainability of hydropower in operation, and compromise its ability to support the energy transition.
How we choose to adapt our understanding and management of water, and by extension, hydropower, will therefore provide the basis for current and future success.
AI-based forecasting tools offer a promising way forward, among many options, as they enable operators to manage their assets more efficiently and protect infrastructure from drought and flooding.
However, these technologies must be part of a broader strategy that includes collaboration (between stakeholders and nations), regulatory adaptation, and proactive planning.
The path ahead may be uncertain, but with the right tools and strategies, hydropower can continue to play a crucial role in the clean energy transition. For operators, the time to act is now.
Marshall Moutenot is CEO of Upstream Tech and a co-founder, overseeing product development, growth and partnerships. Upstream Tech builds real, scalable software-as-a-service (SaaS) climate tech products, such as HydroForecast and Lens. Previously, Marshall worked in a number of early- and late-stage technical companies that have collectively raised over $100M in venture capital. He was awarded Forbes 30 Under 30 in Energy and has served on panels for the United Nations Economic and Social Council (ECOSOC). Marshall holds a BA in Computer Science from Tufts University.
Further Reading:
- More about Upstream Tech; including its SaaS products HydroForecast; and Lens;
- Also on SustMeme, LDES: ‘Future of Energy’ in The Times;
- Also on SustMeme, Ageing dams pose growing threat made worse by climate change;
- Also on SustMeme, Ultra-early wildfire detection raises pilot-site alarm;
- Also on SustMeme, Flood risk: ‘Future of Infrastructure’ in The Times;
- Also on SustMeme, Dam fine dram: Water catchment with your whisky?
You can check out the full archive of past Guest Blog posts here.
Would you like to Guest Blog for SustMeme? For more info, click here.
SUSTMEME: Get the Susty Story Straight!
