David Anderson:
As we begin.
As we begin, I want to remind you all that this Webex call is being recorded and may be posted on DOE's website or used internally. If you do not wish to have your voice recorded, please do not speak during the call. If you do not wish to have your image recorded, please turn off your camera or participate by phone. If you speak during the call or use a video connection, you are presumed to consent to recording and use of your voice or image.
Good afternoon everyone. I'm happy to welcome you all to our seventh and final webinar focused on summarizing the results and insights from DOE's SMART Mobility Lab Consortium. My name is David Anderson. I'm the Program Manager for EEMS or Energy Efficient Mobility Systems. EEMS is a program of DOE's Vehicle Technologies Office; part of the office of Energy Efficiency and Renewable Energy. The SMART Mobility Consortium is a key research effort within the EEMS program and it's focused on understanding the system level impacts that emerging transportation technologies and services will have on mobility and on identifying solutions that improve Mobility Energy Productivity. That's our metric for the energy, the time, and the affordability associated with transportation. Now we kicked off this webinar series back in September and since then we've discussed our SMART Mobility Modeling Workflow, we've discussed our Mobility Energy Productivity Metric, we've talked about Connected and Automated Vehicles, and Moving People, and Moving Goods. Today, our focus is on Charging Infrastructure and specifically what that infrastructure might look like in a future or SMART Mobility System. You know DOE has been researching and developing electric drive vehicles and charging systems for a decade or multiple decades now. And in the past, the mindset has been that plug-in vehicles will either need to be recharged overnight at the owner's homes, or during the day at the owner's workplace, or that we need ubiquitous charging stations on every corner much like the current paradigm of gas stations. But all three of those use cases imply a model where vehicles are personally owned just as they are today. And so what we worked to understand through SMART Mobility was, "What about the other scenarios when the vehicles are shared, or when ride-hailing is the predominant paradigm?" or "What about when vehicles are connected and automated?" Our two speakers today are John Smart from Idaho National Lab and Eric Wood from the National Renewable Energy Lab. John led the Advanced Fueling Infrastructure pillar of the SMART Mobility Consortium and Eric was one of the key PIs leading our infrastructure research. Before I turn it over to John and Eric I will, one last time, give you the context for what we've been researching.
Now as most of you know by now, EEMS is one of the research programs within the Vehicle Technologies Office or VTO. VTO has a long and successful history of conducting research on vehicle powertrains and component technologies things like advanced combustion engines and fuels lightweight materials, energy storage, and electric drive systems. The EEMS program and the SMART Mobility Consortium extend VTO's reach up to the transportation systems level. So we look not only at technologies at the vehicle and powertrain and component level, but also at how vehicles interact with each other and with infrastructure at the small network or corridor scale and then we can evaluate the system in terms of overall traffic flow and energy consumption; finally understanding mobility and developing solutions that improve mobility at the entire urban or metropolitan area. To do this - to comprehensively research and model and develop solutions that reach across modes and domains - it takes a multi-disciplinary approach. We have to consider transportation holistically; as a highly interactive, very complex, highly dependent system of systems, in which changes in one system can propagate to and affect others. And so while we must understand each of these systems individually, it's just as important - or maybe even more so - to understand how they fit together, how they interact, and how they affect one another. To achieve this integration both vertically from the component level up to the metropolitan scale and horizontally across all the different domains relevant to transportation, the EEMS program took a multi-disciplinary consortium approach to transportation research. We created the SMART Mobility laboratory consortium by convening the five National Labs shown here on this slide. They brought to the table their most skilled transportation scientists and engineers to build a multi-year and multi-laboratory collaborative dedicated to further understanding the energy implications and opportunities of advanced mobility solutions. The first phase of our consortium was divided among five pillars of research shown here and they are: Connected and Automated Vehicles, Mobility Decision Science, Multi-modal Freight, Urban Science, and Advanced Fueling Infrastructure. The research results that John and Eric will describe today largely come from the Advanced Fueling Infrastructure focus area. But as I said before, the Consortium is multi-disciplinary and cross-cutting and these focus areas were very much interrelated.
Many of you are likely familiar with what has been called the Charging Pyramid. It applies to our current vehicle paradigm. The base of the pyramid says that most charging will be done at home whether single family... or multi-family dwellings. The middle of this pyramid says that the next highest need and use will be workplace charging and fleet charging while public charging - typically at higher power shown on the top of the pyramid - will fill in the gaps. But it may be that a future mobility scenario which, is automated and connected -and largely shared or fleet operated - could turn this paradigm on its head. So our research set out to understand this. To understand what the need is for different types of infrastructure whether it's centralized or distributed, "What operating modes make the most sense?," "Is it more chargers?," "Is it faster charge rates?" "Is it longer range of vehicles or more vehicles?," and "How do you manage the infrastructure to fit different scenarios?" We looked at charging infrastructure to serve both passenger movement and freight movement and even evaluated the opportunities for wireless charging. After all, if we have automated vehicles, it may make sense to have automated charging.
The research questions outlined in the Advanced Fueling Infrastructure Capstone Report - which is available to download on the VTO website - are summarized here:
What are the charging needs of future vehicle and mobility service segments?
What are the costs and benefits of different approaches to charging infrastructure specifically to serve light duty ride-hailing vehicles and class seven and eight freight vehicles?
What are the opportunities for automated vehicle charging?
And finally, if we plan and deploy the right set of charging infrastructure, what might the large-scale national benefits be?
Now before I turn it over to John and Eric I want to reiterate a point that I made at the beginning of this webinar series back in September. When we think about it, mobility is the foundation of how humanity interacts and connects. And we don't think that's an overstatement. It has driven the development of cities and of civilizations for centuries. Transportation gives us access to opportunity in the form of healthy food, good jobs, quality education, adequate health care, and plenty of leisure and recreational activities. Now we've seen with the current state of technology advancement, we have the potential to not just move more but to move smarter; and in this case, smarter means more affordably, more efficiently, and with more choice. This is important work that we're undertaking. Through the SMART Mobility Consortium, we've created what I believe is a trove of knowledge and understanding about the future of mobility. And our work continues. We've kicked off multiple new projects in the second phase of SMART Mobility to build on the foundational work we did over the last three to four years and we look forward to sharing the progress and results of that work with you in 2021. Thank you all for joining us today and throughout this webinar series.
Now I will turn it over to our two speakers for today: John Smart from the Idaho National Lab will speak first. John will introduce and tee up a lot of the work the consortium did on charging infrastructure. John is the mobility systems group leader at Idaho National Lab and he's been pioneering research related to electric drive vehicles and infrastructure for as long as I've known him; that's over 11 years now. John will tag team with Eric Wood in the presentation. Eric is a research engineer in NREL's Center for Integrated Mobility Sciences and he's been a key contributor to VTO research related to real-world travel data and analysis of advanced vehicle infrastructure and energy storage systems. He's also the lead developer of EVI Pro, which is one of the key pieces of software that we used in SMART Mobility to project EV charging infrastructure needs. After John and Eric wrap up their presentation, we will have some time for for Q&A and I encourage you all to type your questions into the Q&A box on your screen and we will get to as many of your questions as time will allow. Thank you and now I'll turn it over to John Smart.
John Smart:
Thanks David! It's been a pleasure working with you and the whole SMART Mobility team for the last several years. Hello everyone. Eric and I are honored to present to you today the results of a three-year multi-lab research program that evaluated the electric vehicle charging infrastructure needs of SMART Mobility.
When we proposed this program nearly four years ago, we recognized three trends: the advent and growth of ride-hailing, the development of automated vehicles, and the potential to electrify medium and heavy-duty trucks and buses. It was clear to us that electric transportation would become increasingly diverse and this prompted questions of how charging infrastructure should evolve to meet the needs of new electric transportation models.
We asked ourselves, "What is the right kind of charging infrastructure for each mode?," "How much is needed?," and "where should it be located?" After all, the benefits of transportation electrification can only be realized if adequate cost-effective charging infrastructure is in place to support it.
Since we began our research, market developments have validated the importance of answering these questions Waymo has launched the first public robo-taxi fleet using plug-in electric vehicles. Lyft and Uber have made significant electrification commitments. Driverless electric shuttle demos are taking place in numerous cities across the country and truck manufacturers are bringing electric trucks to market.
Our research found that there is no universally right amount of charging infrastructure. There is no golden ratio of charging stations to vehicles, because the charging needs of different modes and use cases vary. Additionally, even with individual use cases trade offs abound within those cases. Nevertheless, we found that it is possible to use modeling and simulation tools to manage cost-benefit trade-offs provided that we understand the behaviors and interests of stakeholders in each use case. In this presentation, we'll share principles we have learned - a guide charging infrastructure designed for multiple use cases.
This diagram is helpful for illustrating the use cases we studied. On the x-axis, we define vehicle use. Vehicles may be driven exclusively for personal use, exclusively to transport other people or goods - which we call shared use, or a mix of both personal and shared use.
Personal and mixed-use vehicles are privately owned; whereas shared use vehicles are commercially owned. On the y-axis, we separate human-driven vehicles from fully automated driverless vehicles.
The first use case we studied was human-driven ride-hailing. Following the model that Lyft and Uber popularized where private vehicle owners strive for both personal travel and to transport paying customers.
We also looked at charging infrastructure for free floating car sharing and for regional goods delivery.
Finally, we studied the charging infrastructure needs of automated ride-hailing and automated transit. We'll share the findings from each of these use cases. Eric will walk us through our first use case: Human-driven Electric Ride-Hailing.
Eric Wood:
Thanks John! Today I'll be presenting a series of SMART Mobility results on charging infrastructure for human-driven electric right-hailing. When considering the near-term electrification of ride-hailing, or TNC services - transportation network companies, it's important to acknowledge the variety of motivations and behaviors exhibited by drivers currently on these platforms. Working with Populus, a company that conducts recurring travel surveys in major metropolitan areas, survey responses from over a thousand Uber and Lyft drivers operating in 10 US cities were analyzed. While a majority of drivers reported primarily financial motivations, such as being in between jobs, approximately 30 percent of drivers reported discretionary motivations, such as wanting to keep busy or meet new people. Further about 10% of drivers reported driving for a TNC on a daily basis. The majority of drivers on the other hand were found to drive two days per week or less.
A similar distribution of driver activity patterns was found in a large GPS data set of ride-hailing activity in Austin, TX including one and a half million ride-hailing trips. This data made available by Ride Austin - a local TNC - found that half of the drivers in the dataset were operating less than 10 hours per week and only 10 percent of drivers were operating more than 35 hours per week. While full-time drivers are often the focus of TNC electrification conversations, our work finds that they currently make up a minority of drivers and rides on TNC platforms.
For the minority of drivers working for TNC's full-time, we wanted to investigate the economic incentive for electrification. This prompted running a five-year total cost of ownership analysis considering present-day EV prices, the amount of time required for mid-day charging, and excluding federal or state financial incentives. Results of this analysis found that 250 mile EVs provide similar total cost of ownership to representative gasoline vehicles and slightly higher costs than today's gasoline hybrids. These results were found to be sensitive to many variables including vehicle purchase price and fuel costs. Collectively, these first few slides suggest that there's a relatively weak financial incentive to electrify ride-hailing services under present-day economic conditions especially considering that not all drivers have purely financial motivations; a minority of drivers operate full-time and EVs currently come with an upfront premium. However, ride-hailing vehicles remain an attractive segment for electrification based on the potential energy and emission savings relative to vehicles operated only for personal use. In order to realize these savings costs effectively, progress must continue to be made on reducing vehicle purchase price through technological advances and economies of scale; potentially with purchase incentives used to bridge the gap. Importantly, these total cost of ownership results assume regular access to overnight charging. However, we know that is not always an option.
To help illustrate the role of overnight charging in TNC electrification, a simulation experiment was designed using several combinations of vehicle and charging technology. While we don't have time to review all of the results today I'd like to focus on a present day vehicle technology scenario where we highlight infrastructure design trade-offs between overnight level 2 charging is the primary charging location with public fast charging used as necessary versus a scenario where overnight charging is not available and public fast charging is the only option for ride-hailing drivers.
Simulations leveraged 5 million miles of real world taxi operation data from the Columbus Yellow Cab fleet in Central Ohio. Results showed that consistent access to overnight home or depot charging obviated much of the demand for public fast charging. In simulations where the fleet was provided access to overnight level 2 charging, only 9 of the fleet required midday public fast charging to maintain existing operations. However, in simulations where overnight charging access was restricted, we found 100% of the fleet relying on the public fast charging network one to two times per day. This dramatic increase in demand highlights the need to understand the socio-demographic characteristics of ride-hailing drivers.
Revisiting results from the Populus national survey of over 1,000 TNC drivers we find that 40 percent of drivers live in multi-unit dwellings with lower income households being more likely to rent and/or live in an apartment. For example, over 80 percent of TNC drivers from households with an annual income below thirty five thousand dollars either rent their home, live in a multi-unit dwelling, or both. This result emphasizes that efforts to electrify ride-hailing fleets must not only be accompanied by investment in public fast charging to support high mileage drivers, but also by investment in public and private level 2 infrastructure at or near where TNC drivers live in order to enable regular overnight charging.
Thinking further about investment in public infrastructure, it's no secret that the capital costs of fast charging infrastructure are significant. Does the site have sufficient hosting capacity? Are electrical upgrades required? How much trenching is necessary? The answers to these questions can result in a capital cost varying significantly from site to site. This realization may prompt charging networks to search the region in hopes of finding the least costly site for development of new fast charging stations. While tempting, our work suggests that failure to design networks around anticipated demand can make low capital cost locations financially problematic in the long term. A simulation study conducted in Columbus, Ohio in partnership with the local electric utility American Electric Power - or AEP - evaluated a dozen potential fast charging sites. Based on the simulated charging demand and existing tariffs, we found that the effective cost of electricity at sites with low demand was financially unsustainable in the long term. Sites with less than 250 sessions per month were forecasted to experience effective electricity costs up to 75 cents per kilowatt hour. These estimates exclude capital costs and are largely a result of inability to distribute demand charges in a sustainable manner. However, well-utilized sites were able to bring down the effective cost of electricity to below 20 cents per kilowatt hour; a much more reasonable expense. The takeaway is that while some minimum level of infrastructure coverage is necessary to enable TNC electrification, network expansion efforts should focus on anticipating future demand and planning investments accordingly. Ride-hailing data observed to date suggest that such locations likely include urban cores transit hubs and at airports.
While we've been focusing on charging infrastructure as it relates to ride-hailing electrification, investments in public charging do not necessarily happen in a vacuum. As part of our research, we wanted to explore what secondary benefits may be associated with investment in public fast charging infrastructure supporting TNC electrification. To do this, we couple the national model of consumer choice for forecasting personal vehicle sales with a charging infrastructure demand model in order to evaluate scenarios with increased demand for ride-hailing at various levels of electrification. Results suggest that investment in public charging infrastructure not only can benefit ride-hard drivers but also induce demand for sale of personal EVs by up to 10 percent in 2030. These benefits stem from increased public exposure to EVs as a TNC passenger, increased visibility of infrastructure alleviating range anxiety, and providing more away from home options to households without consistent residential access. Additional benefits are garnered by charging network companies who have the potential to experience increased utilization of charging infrastructure driven by demand from personal EVs and TNCEVs without residential access. With that, I'll turn it back over to John to present results from additional use cases.
John Smart:
Thanks Eric. Now I'll summarize findings from our study of Human Driven Electric Car Sharing.
ReachNow operated free-floating car sharing fleets in multiple cities until 2019. We partnered with them to model charging of their EV fleet in Seattle based on real-world vehicle use.
Originally, our objective was to determine the optimal location for 20 new 50 kilowatt DC fast charging stations with the goal of minimizing vehicle downtime for charging. We found, however, that increasing charging power above 50 kilowatts will have a much greater effect on minimizing downtime than adding additional charging stations. This is because in this fleet charging time was usually the dominant factor in the total downtime for charging.
In the scenario study, increasing the number of charging stations from 6 to 26 only reduce total downtime due to charging by 4%. Whereas doubling the charging power of the six existing charging stations and the vehicles could reduce downtime due to charging by 36 percent. This finding can be extended to other commercial EVs operating in a small geographic area where travel time to charging stations is relatively low.
Next we looked at electrification of trucking for Regional Goods Delivery. In a case study of a private regional hall hub and spoke motor carrier based in Dallas, Texas remodeled class 8 truck operation where the trucks have access to high power charging at loading docks we varied EV range from 300 to 500 miles and charging power between 150 and 350 kilowatts. And we determined what it would take for electric trucks to perform the same daily driving as the current conventional truck fleet.
Real world data describing the operation of this fleet indicated that drivers often chain several trips together into routes or circuits. Often trucks drive circuits of over 300 miles before returning to the originating regional distribution center. In our modeling, we made generous assumptions that trucks would plug in without queuing and charge continuously without any power curtailment at regional distribution centers and at every destination along the circuit where loading and unloading occurs.
We found that even when employing trucks with 500 miles of EV range and charging at 350 kilowatts without interruption during loading and unloading at every destination, some trucks still did not have sufficient range to complete their circuits. This is indicated in the figures on this slide as negative range.
This suggests that fleets of this type will need to either use trucks and charging equipment capable of very high power or adjust their operations to accommodate electrification. Fleets may need to lengthen dwell times to allow for sufficient charging, take time to charge up public charging stations, or limit electric trucks to specific routes. These complexities add real costs to fleet operations. To realize the potential benefits of electrification, fleets need new tools to help them not only define vehicle and charging infrastructure requirements, but also understand and manage required operational changes.
Now I'll describe our work with Automated Electric Ride-Hailing.
Using sophisticated modeling tools in the SMART Mobility Modeling Workflow, we simulated the operation of automated electric ride-hailing vehicle fleets in San Francisco to understand how charging infrastructure impacts the fleet's ability to serve passengers ___ we included human-driven electric and human-driven conventional ride-hailing vehicles in the simulation to represent a competitive market.
We examine trade-offs between fleet and charging infrastructure design, mobility, energy consumption, and cost by simulating fleet operation with either sparse or rich charging infrastructure networks.
The primary output metric from our simulation was passing passenger miles traveled or PMT; which is a measure of the fleet's ability to provide mobility to passengers. We found that PMT is directly proportional to fleet operating costs.
This figure plus these two metrics were two different charging infrastructure scenarios. Daily cost to operate the fleet is shown on the x-axis. This includes the amortized capital cost of both vehicles and charging infrastructure. The y-axis is daily passenger miles traveled per vehicle. The red circle at the lower end of the red line represents the cost and average per vehicle PMT of the entire fleet of both human driven and automated electric ride-hailing vehicles when they have access to the Sparse charging network. The red diamond on the upper end of the line represents the cost in PMT with the Rich charging network clearly the Rich charging network increases the fleet's capacity to serve passengers but at a greater cost.
Here we've added results for two intermediate size charging network designs so-called Rich 10% and Rich 20% networks ___. We see that an electric ride-hailing fleet of 100 mile EVs, 100 mile range EVs or BEV 100s ___ with access to widespread 50 kilowatt charging infrastructure ___ at one plug per seven vehicles can serve 150 passenger miles per vehicle in a day on average.
Growing the charging network to one plug per three vehicles reduces queuing time at charging stations allowing to flee the fleet to serve 180 passenger miles per vehicle per day.
However adding more charging infrastructure beyond this point cannot cost effectively increase passenger miles served by the fleet, because charging time not queuing time becomes the bottleneck. Easy range or charging power must be increased to alleviate this constraint. Increasing EV range to 200 miles represented by the yellow circle and 300 miles indicated by the green circle increases PMT with increasing cost at about the same rate of return as increasing the density of the charging network.
Now let's look at the effect of increasing charging power.
For BEV100s, increasing charge power to 100 kilowatts cost-effectively increases PMT capacity to 200 miles per day given one plug per seven vehicles.
If the fleet is composed of BEV200s with one 100 kilowatt plug per seven vehicles, it can cost effectively serve 230 passenger miles per vehicle.
Finally, upgrading to BEV300s with 100 kilowatt charging captures over 250 passenger miles per vehicle per day at the same cost per passenger mile. Adding more plugs does not increase passenger passenger miles per vehicle.
In summary, the cost to the fleet per passenger mile serve is about the same for many of these cases. To expand the capacity for human driven and automated electric ride-hailing fleets serve more passengers. Increasing the size of the DC fast charger network ___ increasing charging power or increasing vehicle EV range each are similarly cost effective to a point.
This means that in order to answer the question, "How much charging infrastructure is needed?," we first must answer the question, "How much capacity to serve passenger miles do we want to pay for?" Fleet managers should define a target for level of service in terms of PMT. This will allow them to choose the least expensive charging or vehicle configuration that meets that target. 
Now let's talk a little bit more in detail about how an automated electric vehicle fleet would work.
For automated electric vehicles - or AEVs - charging decisions should be part of an overall fleet management approach that dispatches vehicles to pick up passengers and reposition idle vehicles to better serve future passengers. We hypothesize that increasing the sophistication of automated electric ride-hailing fleet management will allow AEV fleets to serve more passengers at a reduced cost. To test this hypothesis, we simulated a hypothetical AEV ride-hailing fleet in New York City and we experimented with two approaches to managing the AEVs while they're idle between fares. First, we developed a simple approach where individual vehicles make repositioning and charging decisions independently - like human drivers do today. We also developed a more sophisticated approach where centralized dispatcher performs mathematical optimization to determine which of the fleet vehicle vehicles should be dispatched to satisfy rides requests. That is which should be repositioned in anticipation of future ride requests and which should be pulled out of service for charging. We found that for a fleet with about two thousand ride-hailing AEV's centralized fleet management would allow the fleet to satisfy fourteen percent more ride requests and drive forty three percent fewer deadhead miles compared to the case where individual AEVs make independent dispatching precision.
Our final use case was Automated Transit. We examined the potential for dynamic or in-road wireless power transfer to charge automated shuttles while they're operating on six routes.
A simulation of an automated electric shuttle bus network showed that 5 meter long 100 kilowatt wireless charging systems located at bus stops spaced about a mile apart along a fixed route could allow electric shuttle buses to operate continuously at low speeds of up to 15 miles an hour. Those are speeds ___ speed range is indicative of today's technology.
Increasing the wireless charging system length to 175 meters could allow automated electric shuttles to continuously operate at speeds of up to 50 miles per hour which is more representative of typical transit operation. This would allow driverless shuttle buses to operate with near zero downtime which would enable fewer buses to provide increased quality of service potentially at lower cost compared to equivalent human driven shuttle buses without dynamic wireless charging.
Eric now back to you.
Eric Wood:
Thanks John. For those of you that have tuned in to any of the other SMART Mobility webinars you'll notice that a common theme of the consortium is the ability to integrate data and modeling capabilities into a shared workflow.
Our research on charging infrastructure is no different this slide shows a high level overview of model interactions in the SMART Mobility End-to-end Modeling Workflow. This is the same workflow John mentioned during presentation of automated electric ride-hailing results. Centered around metropolitan scale, agent-based, transportation demand models, this workflow incorporates traffic simulation and vehicle connectivity at multiple levels in the simulation of passenger and goods movement considering an array of travel modes and vehicle technologies. Results can be analyzed with varying degrees of resolution and include key performance indicators such as vehicle and passenger miles traveled, energy consumption, and mode choice. All of which can be rolled into an aggregate Mobility Energy Productivity metric or MEP for short. As part of the consortium, this workflow has been applied to a variety of future scenarios with an emphasis on exploring impacts of various vehicle automation and mobility sharing pathways.
A particular interest to today's conversation is the ability to simulate deep electrification future scenarios this slide shows results for 1 million electric vehicles in Chicago and over 400 000 EVs in the San Francisco Bay Area with agent-based modeling conducted using Argonne National Labs POLARIS model and Lawrence Berkeley National Labs BEAM model respectively. These models consider multiple EV technologies including plug-in hybrids and capture demand from personal vehicles, human-driven ride-hailing, and fleets of shared autonomous EVs. These models have been integrated with NREL's Electric Vehicle Infrastructure Projection - or EVI-Pro model - to design hypothetical charging networks capable of meeting the simulated demand in each future scenario. Results provide an unprecedented glimpse into potential charging network designs and are one of the only tools for forecasting the spatial distribution of charging demand in deep electrification scenarios.
Everything that's been presented today has established the foundation of analytic capabilities within the SMART Mobility Consortium in the area of charging infrastructure modeling and simulation. And these capabilities are now being leveraged to inform planning efforts outside of the DOE system.
Two examples of such efforts are shown on this slide. An ongoing study being conducted in partnership with the New York State Energy Research and Development Authority  - or NYSERDA - is focused on estimating charging infrastructure investment needs for meeting the state's goals for zero emission vehicles. This work includes detailed agent-based modeling of electric ride-hailing in New York City with consideration for supply side constraints including limited access to parking and electrical infrastructure in New York City and Manhattan in particular. In a recently completed effort with the Ford SMART Mobility team, agent-based models of electric ride-hailing were developed in four US cities. In each city, five potential mixes of future grid generation were considered. With the desire to quantify the charging load flexibility of a fleet of shared autonomous electric vehicles, results suggest significant load flexibility from such a fleet with annual electricity cost savings up to 46% relative to an unconstrained baseline. With that I'll turn it back over to John to bring us home.
John Smart:
Based on our findings we make the following recommendations: 
First, design around demand. Site public charging stations in locations where ride-hailing drivers need to use them; even if those locations require expensive installations. 
Second, look beyond simply adding more charging stations. In many cases it may be prudent, if not necessary, to increase charging power and or EV range. And in some cases adding more charging stations may not have much benefit at all. Also, recognize that electrification may also require changes to business operations. Define mobility objectives to right-size charging infrastructure and avoid unnecessary costs.
Third, automated ride-hailing fleets should take advantage of opportunities to optimize their fleet operations through centralized intelligent dispatching.
And finally, dynamic wireless power transfer has the potential to substantially benefit automated electric transit and should be considered for fixed route operations.
We recognize the entire research team that made up the Advanced Fueling Infrastructure pillar; the SMART Mobility Lab Consortium from five different National Laboratories. we thank them for their contribution.
David Anderson:
All right. Thank you, John. Thank you Eric - for the very informative presentation. We will turn now to our Question and Answer period. As I mentioned at the beginning, please enter your questions into the Q&A box. We prefer you to use that method, not the chat box, the Q&A box that you see on your screen.
We'll start with one question related to the the Columbus analysis. So the question is: "Why did you only explore the use of 50 kilowatt chargers in the Columbus analysis example - based on taxi data - when higher power chargers are on their way or are here already?" Eric would you like to start with that one?
Eric Wood:
Yeah! I can take that question David. So as I tried to mention in the presentation, we did explore multiple vehicle and charging technologies within the study, but really focused results on today on the the existing network of stations which include mostly 50 kilowatt public fast charging. We did explore, in different scenarios, up to 400 kw fast charging in that analysis. But, you know, decided to focus today's results on the 50 kilowatt network since that represents a big chunk of the installed infrastructure today and is likely going to be a part of the potential transition of ride-hailing operations to electric vehicles.
David Anderson:
Okay, great! Thanks!
Here's another question related related to TNC. So transportation network companies like Uber and Lyft have made major electrification commitments and have announced major electrification commitments. How does that how do these announcements influence or impact the work the charging infrastructure work done through the SMART Mobility Consortium?
Go ahead Eric.
Eric Wood:
Sure! Yeah! I can shoot that one.
So the announcements actually came after a lot of the work that was presented today was conducted and so there wasn't necessarily a direct influence from those announcements on our research, but I do think it's fair to say that those announcements have highlighted the importance of this research and the relevance given the intention that these TNCs have kind of advertised for moving to more efficient electric systems. And I think that there, you know, there's still a number of challenges out there - right? So we tried to highlight some of the cost challenges related to, you know, drivers purchasing these vehicles and getting access to infrastructure. I think one of the big challenges that these TNCs are up against now is thinking about what their what their business model perhaps looks like in an electric future where today, you know, they don't really have control necessarily in a direct way over what kind of vehicles are being used on their platform. So developing their own fleets of vehicles perhaps and making those available to drivers is something that is beginning to happen and interested to see you know what direction these companies fed to try to fulfill their promises of deep electrification.
David Anderson:
Okay, great! Thank you.
There's a question here about about what drives your decision regarding where to install charging stations? I'll go ahead and stop there. There might be a second part to this question, but... So, yeah. In the remodeling analysis what drives the decisions about where to install charging stations?
Eric Wood:
Yeah! I can take that one. So a lot of the simulation work that we do is really what I would describe as demand side simulations. So typically we're looking at a kind of a fixed activity pattern based on real world data and looking for opportunities where charging could occur within the day with limited disruptions to the existing activity schedule. And so a lot of the locations that we end up identifying are a byproduct of the data that we work with. Another kind of dimension that we can take in thinking about locating, you know, these stations for TNC services is to look at the supply side of the equation. So thinking about, you know, where is their access to sufficient hosting capacity on the grid; what are potentially some of the the low installation cost sites that could coincide with sites of high demand? And so it's - I really, you know, I think it's helpful to think about the siding problem from a supply and a demand perspective. And I think we really tried to highlight today how important it is to try and anticipate demand and locate stations appropriately.
David Anderson:
Alright. Thanks Eric.
So related to station location and I'll steer this question maybe more towards John. So John, the conclusions that you walk through warned fleet managers to not simply focus on adding more plugs, but there's a number of factors considering vehicle range, and charge rate, and charge location and demand as Eric was just describing. So the question is, "Isn't this easier said than done? How do fleet managers adequately weigh those multiple competing criteria?"
John Smart:
Absolutely! And I pointed out that, frankly, new tools are needed. We have developed, as a lab complex, a number of modeling and simulation tools to to help study specific cases. And those tools ought to be applied to more cases - right? One thing that we know about really any kind of commercial fleet but particularly truck fleets is that each fleet is unique. And so we welcome collaboration with fleet partners, with manufacturers, and others to be able to apply the existing tools that we have to new case studies and work together to develop, not only the tools from a technical standpoint, but also to take into account the realities of business operations. And for that we need industrial partners.
David Anderson:
Are these tools the tools that you're talking about from the models and the data that Eric talked about are they publicly available or do they require partnering with the labs?
John Smart:
Some ___.
David Anderson:
Eric can you expand on that?
Eric Wood:
Yeah! I'd love to John.
I think that's that question is a really good segue from the the answer that John was just giving. So, yeah. I think that there is a need for more of these kind of analysis and forecasting tools to try and identify locations with high demand. And so, as John mentioned you know these tools are in a variety of states in terms of, you know, their level of sophistication and, you know, their their kind of production quality in terms of being ready for general use. And so, I think that there's a need to continue developing a lot of these tools and finding ways to make them externally available and intuitive and user-friendly interfaces. We mentioned the EVI-Pro model a couple times today and the approach we've taken with that model has been to maintain the model internally but to develop a light version of the tool that is hosted on the DOE alternative tools data center. And so, that tool allows you to conduct analysis around light duty passenger cars, looking at different scenarios for infrastructure requirements, and also charging load profiles for light duty EVs under different technology and flexibility scenarios. And I can envision ways that the tools that have been developed under smart could be migrated to similar platforms and made more accessible to a broad group of stakeholders. Certainly, I think in the labs we're always interested in collaboration but do acknowledge that, you know, our ability to collaborate is sometimes much smaller than the demand for these kinds of models or services. And so, thinking about how to try to make these insights accessible is something that I think is really valuable.
David Anderson:
Okay, great.
We have a question here. It wasn't directly discussed in the presentation but the question is, "Can you describe the trade-offs between level one and level two charging for light duty vehicles from a flexibility, cost, and grid integration standpoint?" Not necessarily the focus of SMART Mobility, but I know both of you have expertise in this area. So I'm not sure which one of you would like to respond.
John Smart:
I'll start. If we've learned anything it is that we need to start our thinking with the question, "What use case are we talking about?"; we walk through numerous use cases. I presume the question was oriented towards privately owned light duty passenger vehicles used for personal use. That's our conventional way of thinking about EVs and if that is true then level one charging as... might already be intuitive. It can be a superior solution, it can be more cost effective, and adequate for locations where vehicles are parked for a long period of time. In fact, I was just looking today and there are a number of level one EVSE installations at airports where vehicles are parked for multiple days - right. However as battery size increases, so that EV range can increase and drivers take advantage of that longer range, level one charge rates can be can be prohibitively slow in that batteries may not be fully recharged after an entire night of charging for residential. Also, it should not be a foregone conclusion that level one EVSE is cheaper than level two. There are a number of different products on the market, level one and level two, and depending on the features selected the the cost can be comparable.
David Anderson:
Alright. Thank you John.
There's a question here about consumer acceptance, really, the question is, "How receptive has consumer buy-in been since there aren't enough stations available on highways or near neighborhoods?" Reminds me of some of the chicken in the egg discussions we had early on in the EV and infrastructure discussion. So, I'm not sure which one of you would like to take that one. "How receptive has consumer buy-in been given not enough stations on highways and neighborhoods?"
John Smart:
I can share an anecdote while Eric thinks of a more analytically based answer. I spent some time in San Francisco during this research project and I took the opportunity to stop at a DC fast charger station where I know a number of ride-hailing drivers driving the EVs would charge. And I literally just hung out at the station and watched the drivers coming and going and it seemed appropriate to talk to one of the drivers and thankfully he was talkative and he shared with me that he lives 20 - 30 miles out of the city center, does not have a charging opportunity at home, and yet he was willing to purchase a battery electric vehicle and rely on the 50 kilowatt DC fast charging network that he had a discount for as part of his deal with Lyft. So there's a single data point where someone was willing to buy or interested in buying an eevee even though they didn't have a chance to charge at home and they were limited to a specific DC fast charger network.
Eric Wood:
Yeah! So I'm going to just piggyback on the anecdote there, John, and just you know confirm that you know uptake of EVs and TNC services has been slow so far and I think that, you know, it's logical to to believe that that's a function of the, you know, the premium that the vehicles come at as well as a number of other things. You know, our friends at the California Energy Commission, and Patricia Monahan in particular, like to talk about the three Cs of electrification which it really resonates with me in terms of cost, convenience, and consumer awareness. I think all of those three Cs have been challenges on the personal side and are similarly challenging on  the ride-hailing electrification side. We did some work as part of the consortium trying to identify and recruit drivers into a data collection program where we would identify someone that was currently using an EV for a TNC platform and try to understand, you know, what what they're doing to overcome some of the barriers related to infrastructure today and you know identifying and recruiting these people ended up being incredibly difficult, maybe even surprisingly so. You know, we know that some of the estimates from Uber and Lyft are that nationally there's maybe just a few thousand electric vehicles, you know, within their or on their platform, you know, relative to a million plus EVs on the road overall. And so, I think you know continuing to push on convenience in terms of infrastructure, the cost of the vehicles and also making consumers aware about the charging technology and the vehicle technology. You know, we've done work recently - thinking back about the level one question - that suggests there's a lot of - I don't know - misinformation or lack of education around level one infrastructure and how sufficient that may or may not be for easy charging. And so, yeah. I think that, you know, there's work on all of these areas and, you know, we certainly have a long ways to go to meet some of the goals that I think we collectively have and for transportational expectation.
David Anderson:
Great, thanks! Thanks to both of you.
Maybe time for one last question here. The question is, "Can you elaborate on how automation can improve the EV fleet?" We only have a couple of minutes. This could be a long discussion, but there are... You know, I've heard many times that automation and electrification kind of go hand in hand, but this question gets right to it. "Can you elaborate on how automation can improve the EV fleet?" 
John Smart:
So this is a question that's near and dear to my heart. One of my research interests is something that I talked about or shared one slide on today. With automated vehicles, if we're not expecting people to drive them then we shouldn't expect people to decide when and where to charge them - right? So one challeng or barrier for automated vehicles is developing the intelligence within the vehicle within the fleet to make wise charging decisions. But the benefit of automation there is that takes that question out of the hands of the passengers or even the owners. If, you know, privately owned fully automated vehicles come to market in that allow the computer allow the vehicles to communicate with the charging network to make decisions about range remaining to make decisions about ideal charging locations. Take that out of the hands of the drivers range anxiety truly becomes a thing of the past. And by drivers, I mean passengers.
David Anderson:
Great! Thank you. We'll go ahead and stop there. Thank you, everyone, for your questions. Thank you to John and Eric for your presentations. Thank you to everyone who attended today. We hope you found the information useful and informative. I want to give a special thank you to everyone who has attended all seven of our webinars. Like I said, we started back in September so thank you for sticking with us. Finally, I do want to give a special acknowledgement and thank you to both Connie Bezanson and Richard Bogacz. They are on our Comms team and did a whole heck of a lot of work behind the scenes to make this entire series possible. We literally could not have done any of this without Connie's and Richard's assistance and guidance. So thank you to both of you. Richard if we go ahead to the next slide...
I want to remind you that you can download all of our SMART Mobility capstone reports at the URL shown there. All six of them. Five for the five research pillars of the first phase of the consortium and then one that covers the development and results from the SMART Mobility Modeling Workflow. You can also download, from that URL all of the presentations, and transcripts, and videos from the entire webinar series that we're concluding today.
Finally, I would like to remind everyone that just last week the department announced a new funding opportunity - several new funding opportunities - across sustainable transportation totaling 128 million dollars. 60 million dollars of this is for the Vehicle Technologies Office-wide FOA - or Funding Opportunity Announcement. I encourage you all to visit our new EERE Program Information Center portal to find information about this Funding Opportunity. You can see the URL there. As does most of our VTO-wide FOAs, this opportunity includes multiple topic areas including Batteries and Electrification, Advanced Combustion Engines and Fuels, Materials, and New Mobility Systems. There's, also, a Transportation and Energy Analysis area of interest that includes several topics specifically related to EV charging - which was the topic of today's webinar.
Again, please visit the portal and read the announcement if you are interested in partnering with us through our Funding Opportunities. Thank you again everyone. Stay safe and best wishes as we get into the holidays. Thank you.