At Municon 2025, Jamie Feldman will present “Advancing the Use of Optimizer™ for Integrated System Planning in the City of Portland, Oregon.”
In that presentation and this conversation, Jamie discusses her work with the City and explains she used Optimizer software to streamline analysis and decision-making as part of Portland’s integrated system planning work.
About the Speaker
Jamie is a Project Engineer who focuses on water resources management. She specializes in stormwater management, hydrologic and hydraulic modeling tools, and data analysis and visualization.
Q&A
What is your background working with the City of Portland?
I’ve worked with the City of Portland and, specifically, the Bureau of Environmental Services (BES) for over six years. My early projects focused on pre-design analysis to enhance the outflows to the Columbia Slough. I performed GIS data screening and some modeling to estimate flows to design those BMPs.
I worked on the Sullivan Trunk, the main trunk line of Portland’s combined sewer system. The City has worked for a long time to find the best solutions for pipe capacity issues on that trunk line. We began with a traditional process by using hydrologic and hydraulic models to identify priority areas in the basin. To streamline this process, I helped BES develop a framework for using Optimizer.
What is the City’s integrated planning process, and how has Geosyntec partnered with the City in that process?
Integrated planning helps the City identify their priorities across different Bureaus, and then plan, optimize, deliver, and monitor solutions that meet those priorities. The first step is to understand and refine system risk, which is the chance of something happening multiplied by its consequences, and then develop priority areas to address those risks. This is followed by an integrated review process, then stages of characterizing the area, developing and evaluating solutions, and recommending the best approach.
These solutions are then passed on as a package to the four environmental service delivery streams: capital improvement program projects, program and regulatory policy, legal, and operations and maintenance.
Our biggest contribution has been creating innovative approaches to identify solutions in Portland’s combined sewer collection system. We started by developing preliminary alternatives using the City’s asset management system and hydraulic models. Then, we used Optimizer to identify improvements that can reduce risks related to capacity and pipe mortality.
This approach combines projects that address flooding risk (capacity) with opportunities to address the risk of structural failure (mortality). This allows the city to plan projects that will address the most total risk, while only having to open up the street once.
What is Optimizer, and how does it improve this process?
Optimizer is software that helps you walk through the steps of formulating optimization questions. It generates a range of potential improvements across multiple objectives and constraints, which would be challenging to produce and analyze manually.
First, you create an objective function, which is one or more equations with a goal. If you can minimize or maximize these equations, then you can say that you’ve met your goals for this project.
Then you load the input data, including the hydraulic model, asset data, and cost information. Optimizer helps you choose the options for the model to play with, such as which pipes do you want to toggle? And how big do you want to let them be? Next, Optimizer asks you to define criteria for measuring success.
Then, you run the optimization for however many model runs it needs to test different solutions (maybe 10,000 or 80,000) and it finds the ones that meet the objective functions. The program uses a genetic algorithm, where each generation of simulations builds on the successes of the previous ones, making it more efficient. Once it finds solutions, Optimizer can investigate the results and tease out the solutions that you’re most interested in.
Using this software is much faster and more comprehensive than traditional trial-and-error methods. Optimizer can evaluate thousands of assets in different combinations. We can adjust the model and quickly see which combination will perform best against multiple objectives, for example, not just hydraulic performance, but what’s going to minimize cost, or deliver additional benefits. With Optimizer, we can evaluate many variations and find the best solution to match the municipality’s goals.
What is one of the projects that you’ve worked on using Optimizer, and what was the outcome?
The project that stands out to me is optimization support for the Sullivan Trunk and Sullivan Basin. The trunkline runs along Interstate 84 on the east side of Portland, serving northeast neighborhoods along Sullivan’s Gulch from the Willamette River to the Rose City Park neighborhood. The basin covers around 1,825 acres with roughly 47% impervious surfaces. This trunk line is overloaded, and during large storms, especially with increased development, residents are at risk of basement flooding and flooding on roads and properties. This is an area where the City has a lot of capacity risk. On top of that, the infrastructure is aging.
So, the question is, how can Portland make improvements and repairs without having to come back to it again in 20 years and address a different issue? The trunk line is part of Portland’s combined sewer system and sometimes overflows into the Willamette River. The City has worked hard over the years to minimize the amount of volume that overflows.
We began the optimization with the objective of implementing different inflow controls and upsizing pipes to reduce capacity risk while minimizing cost. We focused on areas in the basin where the city could get the biggest “bang for their buck” in terms of risk reduction.
From there, we included additional functionality for pipe mortality risk and the benefits of reducing system volume, along with ways that the City can get credit for that as well.
How has working with Optimizer with the City of Portland prepared us to help other municipalities with their wastewater and stormwater projects?
Every city has different data available and questions that they’re trying to answer. Through the work we’ve done with Portland and Calgary, we’ve gained a lot of experience tailoring the software to meet specific municipal needs.
We’ve also worked closely with Optimatics (the developer of Optimizer) to add custom functionalities based on our experience. This allows us to offer municipalities a tailored experience, rather than just Optimizer’s standard capabilities. While our experience is with stormwater, it can also be applied to wastewater.
What do you think the next steps are for cities who are looking to improve their planning process?
A municipality can contact us to assess their needs and data and discuss whether Optimizer is the right tool for them. There are a lot of ways to start small with Optimizer; it doesn’t need a lot of data or a perfect model to be beneficial.
We are experienced in making sure the municipalities’ questions are well-suited for Optimizer. We come up with creative ways to use the tool with minimal effort, allowing the process to grow organically over time.
Whether a city is solving multiple objectives or addressing a simpler problem like isolated flooding, we can use the software to identify other potential trouble spots. There are no data prerequisites. It’s more about having questions and priorities without a clear and quick solution, and that would benefit from the trial-and-error capabilities Optimizer can offer.
