Published on July 12th, 2018 | by Carolyn Fortuna0
Which Rollout Strategy Is Best For Charging Points? (#CleanTechnicaExclusive)
July 12th, 2018 by Carolyn Fortuna
The simultaneous growth of electric vehicles (EVs) and public charging points (CPs) has inspired a new study about the dilemma of choosing a rollout strategy that works best for municipalities. Published in Energy Policy in June 2018, the report describes two rollout strategies for charging points and the long-term advantages and deficits of each. After examining more than 1 million CP interactions, the authors conclude that, as the EV market evolves, no rollout strategy has a clear advantage over the other. We were also able to reach the lead author directly and elicit his ideas for future CP infrastructure.
“Assessment of public charging infrastructure push and pull rollout strategies: The case of the Netherlands” asks, “What are the differences in performance in the 2 rollout approaches?” The first rollout approach is “demand-driven” (DD), in which an EV driver requests a CP near to home. The second rollout approach occurs when local or regional governments decide to place a CP near public facilities or on strategic locations of expected use, known as “strategic.”
In order to conduct the study, the authors (see footer identification) acknowledged that they needed to evaluate the performance of CPs in order to assess the merits of each rollout strategy. Ultimately, they settled on a “charging time ratio (sometimes referred to as ‘load flexibility’), which is the total charging time and total connection time on a weekly basis.
A study of a 2015 Netherlands CP map shows that most DD CPs are in highly-populated areas — which makes sense. Higher population density increases the probability of EV use. Strategic CPs were more evenly dispersed geographically across the country. Strategic CPs were found in areas of low and high population density. Noting a “clear higher growth in incidental and regular users than frequent users,” the authors allow that performance differences may be attributed to “supply demand dynamics” (p. 42).
Hypothesis 1: Would strategic charging points outperform demand-driven charging points?
As they prepared the study, the authors acknowledged that “a mismatch in supply and demand may occur in the data due to the rise of EV sales, and distributions underlying the performance metrics will vary over time and therefore need to be compared as part of the root causes for performance differences (p. 39). As years progressed, strategic CPs outperformed DD CPs based on the number of unique users, a trend which the authors expect to continue.
Hypothesis 2: Wouldn’t demand-driven charging points display longer weekly connection times?
While the relative growth of duration at strategic charging stations is higher than at the DD CPs, DD CPs did outperform strategic CPs on weekly connection times. It was noted that both types of charging points display seasonal dips, with DD CPs showing more prominent impacts.
Hypothesis 3: Would demand-driven charging points perform better on weekly energy transfer than strategic charging points?
Currently, DD CPs are outperforming strategic CPs on weekly energy transfer. However, the differences are decreasing between the two annually, so the authors believe strategic CPs may outperform their DD counterparts in the future.
Hypothesis 4: Would strategic charging points outperform demand-driven charging points in charging efficiency?
As expected, strategic CPs outperformed DD CPs in charging efficiency, measured in charging time ratio. Charging time is a fairly constant element of EV users’ behavior, the authors note.
A Rollout Strategy in Relation to Peak Charging Times
Data on peak charging times is interesting as we look ahead to an era where most vehicles on the roads are EVs. DD CPs have a much higher peak for overnight/home charging, while much strategic CP use comes in at a morning peak. Indeed, the authors “expect the weekly connection duration to increase to a theoretical maximum per CP for its given context” (p. 43) based on annual increases of EV users. That means, as the number of EVs reaches a high point where EVs are normal, we’ll need a lot more CPs.
Charging time is a result of transaction volume and charging speed, and changing time ratio is calculated by dividing charging time over connection time. Longer connection durations at DD CPs are the main factors in charging time ratio differences. Transaction volume of charging sessions is a confluence of connection time and potential charging volume.
Strategic CPs? Short sessions. DD CPs? Longer sessions. The authors suggest this is because of limitations on connection times, which could be improved at strategic CPs by increasing the ampere per CP. Potential transaction volume is also limited by the state of charging at the beginning of a charging session and the battery size of the EV.
Conclusions about Each Rollout Strategy
The authors conclude that, unlike in the past when transaction volume was a combination of volume and frequency, now its most important factor is total energy transfer. They suggest that battery size increase may lead to increase in difference in weekly transaction volume.
Conclusions point to no single rollout strategy having a clear advantage over the other. DD CPs outperform on energy transactions, but strategic CPs outperform on charging time ratio. Another — and really important conclusion — is that the “maturity of EV adoption is in relation to the scale of charging infrastructure” (p. 45). They argue that a portfolio of DD and strategic CPs depends of the maturity of market (EV adoption) and technology (battery size). As a result, localities “may want to choose their best strategy according to municipal objectives such as clean air and business case” (p. 46).
Because battery size in the total population is a factor affecting both the weekly transaction volume and charging time ratio, the authors do suggest that “policy makers should subsidize only EVs that have a substantial battery size” (p. 46).
Interview with Lead Researcher Jurjen Helmus
After reading the study, I contacted Mr. Helmus to see if he would comment for CleanTechnica. I asked him, What would you recommend to municipalities that are at the very start of rollout public charging infrastructure? Would you recommend strategic CPs, given their overall applicability?
Here are his thoughts.
The best strategy for charging infrastructure is contextually dependent, as are so many other difficult decisions. Dependencies are on maturity of EV fleet and local adoption of EV fleet. Also are there strategic locations in a city available for public charging infrastructure? For example, we found that locations that attract visitors are perfect for strategic CPs.
In general, I believe that policy makers should strive to develop a charging infrastructure as a network of alternatives within relevant walking distance. This means that for each station there should at least be an alternative location to use in case of occupancy, etc. This allows to scale up charging infrastructure. Yet, before you reach a density that allows this network structure, a lot of CPs need to be placed. And then we are at the core question of our current research — how to start. 😊
Now, from our research we have seen that policy makers best start with demand-driven rollout (of course at locations where home charging is impossible). This ascertains that at least a CP has one user and may attract other EV users in the vicinity (read: relevant walking distance) from the CP. As soon as the first CPs are installed, strategic CPs may be placed as well, particularly if the local EV uptake get started.
Lastly, monitoring of charging infrastructure is of high importance, as without proper data analysis policy makers would be unable to effectively rollout charging infrastructure. Our idolaad project (www.idolaad.nl) helps policy makers in the Netherlands by monitoring charging infrastructure performance. We are continuously working with the local policy makers to test their ideas/hypotheses on how to best deploy this new technology.
All in all, charging infrastructure is a complex system that may display transformations of behavior and best strategies.
J.R. Helmus, University of Amsterdam (UvA), Computational Science Lab, Amsterdam, The Netherlands
J.C. Spoelstra, Technolution B.V., Burg. Jamessingel 1, Gouda 2803 WV, The Netherlands
N. Refa, Elaad, NL, Utrechtseweg 310, Arnhem 6812 AR, The Netherlands
M. Lees, National Research University ITMO, St Petersburg 197101, Russia
R. van den Hoed, University of Applied Sciences Amsterdam (UASA), Amsterdam, The Netherland