The Washington State Department of Commerce is looking to fund more than 5,000 electric vehicle chargers with its current set of grant funding. With a project of that scale, understanding future utilization is critical in order to get the most return on what is a very significant investment. Thoughtful prioritization of charging infrastructure is also critical in “reducing greenhouse gas emissions and fossil fuels, improving air quality, and promoting equity in access to electric vehicle charging infrastructure,” each of which are explicit goals of the Department’s program.
To equip local applicants with the data they needed to prioritize their proposed projects, the Department created its Publicly Available EV Site Mapping Tool in collaboration with The Timmons Group. The GIS-based tool maps funding priorities based on a number of factors, including high demand for EV charging infrastructure, demographics, social equity, and environmental health disparities.
Replica powers the ability for users of the tool to select a parcel, and see data about the number of trips intersecting its vicinity as well as the average dwell time (time between a person's trips) for that area.
We spoke with Policy Lead Steven Hershkowitz from the State and Dave Knauer, Senior GIS Analyst from the Timmons Group about the project and its impact. Our conversation, edited for length and clarity, is below.
Let’s dive right in. What are the origins of this tool?
Steven Hershkowitz [SH]: I’ll jump in. When we first got funds from the legislature, we didn’t have a team together. I moved into a transportation policy leadership position and was reviewing a great deal of research that suggested the importance of charging station utilization as a key metric.
The way it previously played out in our state was very narrative-based. Something like, “there’s a grocery store in our community and people go to it so it seems like a great place for charging.” That all makes sense intuitively but the way these stations are going to pencil out financially over the long term is hitting a certain utilization rate. And if we hit the utilization rates on the first stations we built, we’ll need to pay for fewer later.
It became very clear that the prioritization process to determine which charging infrastructure to invest in needed to factor in utilization. We needed to try to predict that in any data-driven way that we could.
We decided we needed to create a mapping tool that was using data based on people’s real behavior but also protected individual privacy. That was really the origin of it. We needed all the right data points to come into a single mapping tool that could ease the process for applications.
Makes total sense. From that perspective, it sounds like you laid out the problem statement first? And then the second question was which data sources made sense to solve that problem?
SH: It started with: “This is what we wanted to score.” We wanted to drive applicants to prioritize their projects, whether it be to prefer communities that have not been served by the private sector previously or specific locations that we know would have high utilization.
Then we worked with Timmons to identify what data sources were available. We became familiar with Replica and other tools in the space. We understood that the Illinois Department of Transportation was working with Replica, which was a good proof point.
Dave Knauer [DK]: We [Timmons] came in and assessed what data sets were available for the tool and then evaluated which ones were publicly accessible and which sources we would need to work with third-party contractors such as Replica to get access.
Any advice for agencies looking to select a dataset to purchase?
SH: I talk to as many people that I trust about what data they have used reliably. In this case, it really helped that we were working with a consultant for a separate project where we were modeling how many charging ports we were going to need in each census block for light-duty vehicles, and they were using Replica to do that work. They had vetted Replica for that project, and they bring a lot of rigor to their work, so if they were trusting Replica, it not only created consistency between the two projects, but gave me confidence.
DK: I would second that statement. We looked at many organizations that had used Replica successfully, including case studies with the state of New Jersey and LA County. In addition, meeting with multiple vendors and seeing what sample data sets could be provided and tested was fairly important. So was verifying if their customer support would be available for additional training, and if there was strong supporting documentation.
Replica had pretty extensive metadata and documentation on the modeling process, as well as publicly listing all of the attribute details that are in the platform. That made it a much easier choice based on how transparent the company was about the data that they were taking in and the data that was being produced.
I'd love to hear if there are success stories or initial implementations from the tool both in terms of decisions about where to put infrastructure, and also if you've gotten engagement from the press for the public.
SH: We've gotten really positive feedback from a lot of people about the simplified nature of the scoring and the fact that we’re paying attention to this. I do think we have more work to do, especially with tribal communities, about making sure that if we're going to score a certain way, we're recognizing how all communities share data.
We are on the cutting edge right now. We have more progress to make in how we're using the data. I think there are more stories to tell in the next wave of using the data, and this was a baseline to see how it would work.
What will success look like for the program overall?
SH: So, for this round of funding, we're looking at probably more than 5,000 chargers. That’s just the start because we have a long way to go to ensure that we bring access to people that right now just have no real opportunity to own or lease or use electric vehicles without the necessary infrastructure.
Looking ahead, what’s on your wishlist for the next version of the tool? Top of your Replica wishlist?
DK: Well we were really excited to hear that EV fuel type has been added for Replica data starting with Spring 2023. We’d also love to have a pre-calculated dwell time attribute. We worked with Replica’s team to create it for us, which was so helpful, but it’s not readily available yet.
Any other parting thoughts?
SH: This project is all about making sure that data is being used to better inform the quality of the financial investments we are making. If we invest well early on in the right charging infrastructure, there will be less investment that is needed long term. And it accelerates the whole industry towards a place of economic viability, which we're starting to see now as utilization rates increase.
You know, I think often in the public sector we are looking for something that is without risk and is perfect. But as this project demonstrates, I think that we need to be more willing to, in the public space, play around with cutting-edge data and experiment and innovate. And that's the way that we're going to improve.
So I would just encourage folks to be willing to experiment a little bit. They can certainly reach out to us to talk to us about our experience.
Replica is not the only company out there in the trip data space, but you are a leader in this space. We’ve used Replica for our modeling work before in the Transportation Electrification Strategy, and now in a follow-up analysis with a different consultant who's also using Replica data. It’s great that we’re seeing improvements in the quality of the data. It just shows that it's only going to get better and it's going to become a more useful tool for us to do the best that we can for the people we work for and represent.
Learn more by downloading Replica's EV Guide.
