Camilla Braithwaite on RIDM: Taking the ‘Risk’ Out of Risk Management

As many federally regulated US Dam safety leaders begin to fully embrace a risk-informed decision (RIDM) approach to dam safety in response to updated guidelines, others in the wider dam industry know it’s just a matter of time before their regulator will require them to do the same.

Some dam operators and consultants getting up to speed with the new guidelines are reporting a significant increase in the level of effort and expertise required to develop a riskdriven approach, with a blank sheet approach to potential failure mode analysis and much greater detail on failure modes required.

With many in the race to prepare their next Comprehensive Assessment (CA) and looking for an Independent Consultant (IC), these dam operators are facing a huge lift in terms of the information and analysis burden.

Dam operators and their consultants are grappling with pulling together performance monitoring evaluations, field inspection information, a detailed review of the design basis and data to underpin potential failure mode (PFM) analysis and risk assessment, which includes analyses for flood and earthquake loading plus estimates of life loss, property damage and economic impacts from flooding as a result of potential dam breach. Not to mention the information and analysis required to prepare Inspection Plans and Pre-Inspection Preparation Reports in advance of the inspection.

The first big piece of advice coming from hydro dam operators working on a Comprehensive Assessment? Be able to find information easily.

Not just that, but with many estimating that the additional workshops, staffing and analyses required for just the Comprehensive Assessment costs between $250k and $1million, being able to maximize every single piece of information, knowledge and expertise is essential so that nothing gets missed.

Many forward-thinking dam safety leaders are looking at innovative technologies to help them make the transition to a more risk-driven approach. One such technology is Rezatec’s dam monitoring geospatial AI platform.

This solution digitizes and centralizes into one dam record all dam data and information. It boosts the dam operator’s dam record data with year-round satellite monitoring and analysis which tracks unusual changes in movement, seepage and vegetation across the entire structure and monitors downstream hazards in the flood zone. The platform’s advanced AI and visualization engine analyses this huge data set, tracks trends and changes, highlights potential issues and visualizes them on a map of the structure. It also features a risk management matrix, allowing the operator and their consultants to dynamically track risk and assign tasks.

The outputs are powerful. From an operational perspective, dam safety leaders are picking up errors they hadn’t noticed in instrumentation data, identifying movement over time that’s too fine to detect with the naked eye, spotting the development of potential wetlands that can threaten the ability to carry out maintenance work, and highlighting the requirement to reclassify dam hazard ratings. They are also able to demonstrate that their structures are stable so that they can prioritize investment on more pressing maintenance issues, where others are able to confidently question the extent of deformation to better align repair costs.

But it’s the advantages this type of technology brings to risk-driven methods that are really powerful. A big data approach, combined with proven AI and visualization, helps dam operators and their consultants to prepare the most comprehensive risk assessments and ensure they miss nothing.

The first step towards this mission is to ensure information is centrally stored and easily accessible. Many of the dam operators we work with tell us that their data are spread all over the place, often in filing cabinets or on personal drives. When they’re trying to get an understanding of a dam, especially when running their first Part 12 Inspection, they’ve just recently acquired the asset, they’re new to the job, or their regulator is asking questions they are struggling to answer, dam safety teams want everything in one place so that they can get a broader picture.

Being able to digitize all data and create one central dam record is useful, making sure it includes uploading visual inspections, photographs, checklists, emergency action plans (EAPs), instrumentation data including water levels, piezometer, seepage, geodetic, and other survey data. However, doing that in a platform where data can be compared and gaps identified easily is crucial to get that ‘big picture’ view.

Some dam operators monitor their assets for potential failure by sending engineering personnel on site to conduct an annual or biannual survey. Others also have staff checking the structure on a four-to-six-hourly rotation. These activities are vital; however many are finding the visual aspect of these inspections less able to quantify trends, or movement over time that’s too fine to detect with the naked eye, in a way that truly adds value to assessing risk. Added to that, the less frequent nature of some surveys results in data gaps.

Instrumentation is also a key part of monitoring but, with the investment needed to drill borings and instrument a high hazard dam sitting at the $300k mark, it’s simply not cost-effective to instrument the entire dam, sensors can only measure the points they are located, and occasionally they fail. As a result of this, data can be patchy.

Many dam safety leaders are relying on Rezatec’s dam monitoring platform to provide data across the whole dam and regularly at two week intervals. Additionally it uniquely provides a 3-year historical analysis of movement, vegetation and seepage. Using satellite data, combined with machine learning, it understands the normal behavior of the structure over time, alerts the owner to millimetric unusual changes and maps this onto a visual of the dam and any embankments, helping to highlight potential issues ahead of time.

Hydro dam operator, City of Spokane, has used this data to evidence non-movement on their rip rap, “Rezatec’s geospatial AI platform gives us more information about the dam than we would otherwise have access to, that helps us build the investment story for decision makers,” explains Jeanne Finger, Spokane’s Chief Dam Safety Officer. “Our recommendations are based on fact and objective data, which bolsters the case for them… The data really fits to solve the puzzles and gives us solid evidence that satisfies FERC’s questions.”

Can it be relied on? Well, yes. A recent ground-truthing exercise with another dam operator compared geodetic survey and other in-situ data with satellite AI analysis on a high hazard dam following a potential failure event. Not only did the satellite analysis correlate with the survey data within 80-90%; it highlighted some erroneous values in the geodetic vertical movement data and bridged that gap. More importantly, the satellite ground motion analysis demonstrated good stability prior to, during and after the event and that the event had not destabilised the dam, the case for which the engineering team put forward to their regulator.

This year-round satellite data, covering the entire structure, provides the baseline for any dam operator’s monitoring program and informs PFM evaluation, while surveys and instrumentation dig deeper and provide additional data to further evaluate potential issues.

Dam operators outside of the USA, who have already transitioned to a risk-driven approach, tell us that good quality data and the most comprehensive data sets underpin potential failure mode (PFM) analysis and risk assessment.

The real power for dam safety teams is being able to combine all their satellite, dam record and sensor data, which the geospatial AI platform does effortlessly. This information is further boosted by applying AI and visualizations to the data sets, so dam owners get a full view of the dam and its environment, review the analysis visually on a map of the area and compare time series graphs and charts.

Chief Dam Safety Officer, Daniel Turnbull, at Australia’s Hunter Water asserts, “It’s not physically practical to check every aspect of our dam, which has over 3 miles of embankment that is home to a major 50 mph speed limit road. In addition, the shoulders of the dam are constructed with sand, which moves during normal operations and the stable portion of the dam is solid clay core, which is directly under the road. Regular, physical monitoring under these circumstances is costly and inconvenient for road users, not to mention a health and safety concern for our engineering personnel.”

Hunter Water’s dam safety team finds satellite monitoring to be more effective at targeting engineers and investigation resources at the right place in a timely manner. They and other dam operators around the world are also using it to verify potential issues picked up in visual inspections, going back to the data to corroborate what they’ve seen on the ground and what’s being picked up quantitatively from all monitoring sources.

Many engineering teams are seeing value from working with the platform interactively, capturing notes, photographs and other information from visual inspections, which they’re feeding back into their digitized dam record. This further builds the data set and really helps them when it comes to preparing performance monitoring evaluations, PFM analysis and risk assessments.

Daniel Turnbull adds, “At Hunter Water we are now confident we are doing everything in our power to exceed regulatory requirements and safeguard our dams and the surrounding population.”

Assessing estimates of life loss, property damage and economic impacts from flooding as a result of potential dam breach is an essential part of the Comprehensive Assessment. The dam operators we work with typically have used aerial imagery to try to do this assessment. But it’s proved challenging and takes a lot of effort – after all, floodwaters follow the lay of the land and a wave runs in a lot of different directions. Depending on the volume of water and the topography, the flood zone can cover a wide area, and finding up-to-date imagery and carrying out manual reviews of aerial images at that scale takes weeks of activity and a fine eye for detail.

Geospatial AI helps with this by identifying at scale dam inundation zones – often across an entire dam portfolio or a whole state if it’s for a regulator – and it visualizes the buildings in that zone. The technology allows dam safety personnel to assess terrain and building heights in 3D, zoom in on exact locations, measure distances and points of interest, highlight significant buildings like hospitals and schools, and assess the number of buildings in the flood zone. It also pulls the address data for those buildings, creating a list of details for inclusion in EAPs and the CA.

These dam operators are already realizing they can strip out the sheer manual effort of preparing this type of information, not only saving time and resources, but getting to the important review and decisions more effectively.

Ensuring everyone involved with the dam can see its current state of risk, how the risk has changed over time, and what actions are being taken to mitigate or reduce the risk is essential when implementing a risk-driven approach. Dam safety leaders are advocating that this information is put front and center, so that everyone can see what’s happening.

Many of these dam safety teams are leveraging the dynamic risk matrix capabilities in the dam monitoring platform using the details of each PFM, its likelihood and consequences of failure. Engineers are logging site inspection, recording findings, adding notes to points of interest, assigning actions to team members and managing risk actions. It’s proving a popular change for those already putting in the ground work creating a process for preparing information for their regulatory submissions. Others are actively working with this dynamic risk matrix as part of their P12 workshops.

When it comes to adopting a risk-driven approach to dam safety, engineering teams need to be satisfied they haven’t missed anything. They also need to optimize the time, expertise and value of those involved in preparing evaluations, risk assessments and running P12 workshops to drive the most impactful outcomes.

Building a repository for all knowledge, data, information and actions is just the beginning, and there are proven, innovative technologies that will do this and more to reduce the manual overhead and ensure the risk is taken out of risk management.

Original article featured in the USSD Dams & Levees Bulletin Winter Edition