Can data improve dam safety?
May 09, 2022
Managing dam safety requires multiple instruments collecting data 24/7. What if we had a streamlined data organizing and analysis tool?
Share
In our modern world, there seems to be a constant push towards being digitally connected, streamlining processes, and maximizing efficiency.
Dam safety professionals are feeling this same push, to the point where something needs to change. There’s a catch-22 in the industry. Thanks to technological advancements, we’re able to collect more data about a dam than ever before. This should mean we are able to identify potential issues. But sometimes, massive data sets can be more overwhelming than helpful in understanding the actual condition of a dam.
We need a better way to organize and analyze hundreds of thousands of data points. This would help us to better identify potential issues and even save lives by preventing dam failures.
Dams are a critical component of public infrastructure and safety around the world.
Perspective on how much data we’re talking about
Dam monitoring tools take continuous measurements. Sensors gather data like the amount of sediment in the water, the level of the water within a dam, and ground movement. Other instruments collect information about precipitation, air temperature, and flow rate. Other data collection tools include: Interferometric Synthetic Aperture Radar (InSAR), satellite imagery, radar, and total stations. Each one of these measures for a particular defect or condition within or around the dam.
We need a better way to organize and analyze hundreds of thousands of data points.
Right now, our team executes work for a state-wide utility corporation that oversees more than 40 dams. We organize and interpret hundreds of thousands of data points across various instruments and sensors. Then, we notify the client of potential issues. But, due to traditional collection and processing routines, the time between data collection and interpretation can span weeks or even months. The time between observation and action is lost time which is critical in terms of responding to a dam safety issue.
Think about how much data you’d have if you’re watching dozens of sensors collecting data all day, every day. What about if you have multiple dams? How can you organize and analyze each dataset to understand the full condition at any given time?
Dam management applies to mining tailing dams, too. While they look different from a typical water dam, the safety of the dam relies on many similar components.
What if there was a tool to bridge the gap between raw data and engineering interpretation?
Part automation, part dashboard. The goal of this tool is to allow engineers and dam safety managers to know the actual condition of their dam in as close to real time as possible. Then, dam safety becomes more predictive and less reactive.
Artificial Intelligence (AI): The AI component wouldn’t replace human engineers, but it could be helpful to quickly analyze massive data sets to search for anomalies. It could even predict future behaviors based off past data and forecasted conditions. For example, AI could remember how a dam reacted the last time it rained 10 inches in a week’s span, and project how an upcoming rainstorm may affect the same dam.
Automation exists in our lives and in other industries. Think of a self-driving car. It can see the traffic around it in real time and detect what’s happening. If someone walked in front of the car, it would know to brake. It’s time to apply the AI interpretation to dam data analysis, too.
AI could be extra useful in a tropical area that’s known for getting rainy seasons every year.
Big-picture Dashboard: The industry already has excellent data processing and visualization tools. But each tool is focused on the niche technology, instrument, or sensor for which it was designed. A system that can organize large volumes of data into one platform and automate the engineering analyses would allow us to respond to data more quickly. Real-time actionable results coupled with predictive alerts will put valuable time back into the hands of dam owners. This would promote safer infrastructure and benefit the community.
Some initial platforms are being developed, but they’re not 100% ready. We’re talking about a fully-functioning digital twin—an automated dam monitoring tool that is available for the public. Automated data analysis is starting to become a preferred practice in the industry as owners face resource constraints and look for more efficient ways to make informed decisions. As more tools become available, and the volume of data continues to grow, it will be critical to have a more efficient way to organize and analyze massive data sets.
Dams for different purposes will require slightly varied measurements. For example, a hydroelectric dam has more components to it than strictly a reservoir dam.
Looking ahead
As dam safety professionals, we strive for a world where all dams are safe. We are collecting more data than ever before. While that presents a tremendous opportunity to improve the safety of our infrastructure, it also presents a major challenge with managing, processing, and acting on the incredible volume of information. We need to be responsible stewards of the data and use it in a way that can proactively identify issues. We believe a centralized monitoring platform enhanced with AI can help.
The big win comes when all dam operators have access to a single source of truth for monitoring the health of their dam assets. This includes all companies, from multinational corporations with big risks and large teams, to smaller organizations with limited expertise in dam monitoring and the associated risks.
The idea of having all dam performance metrics in one place—with the added benefit of automated engineering analyses and predictive capabilities—is already here. It’s time to put it to good use and prevent future dam failures. Knowing and acting on data quickly could save human lives and prevent environmental disasters.