Learnings
Outcomes
Functionality testing report
During January and February 2021, CrowdCharge instructed four pilot sites to carry out some testing, utilising their Wallbox App to prove the app was functioning correctly and the data received to the CrowdCharge platform was consistent with what the participant was seeing at site. Some results in the first round of testing were unsuccessful, as two vehicles were unable to discharge their EV using the app, or the site experienced a communication fault. The former was important to discover as this led CrowdCharge to further understand the operational limits of the charger after discussion with Wallbox’s technical department. This operational constraint of the charger not being able to discharge if the State of Charge (SoC) is 97% or above is critical to understand in designing the one-year customer trial charging/discharging optimisation algorithms.
Generally speaking, testing has been successful in that, all sites are able to operate their app and instruct their Quasar to charge their EV without any issues. As some faults have occurred with discharging operation, further Quasars are required to be installed in domestic customer’s properties to increase the volume of charger operation and charge cycle data on CrowdCharge. Therefore, CrowdCharge recommend proceeding with main trial installation of c.10-15 units. As each charger installed, CrowdCharge monitored the operation and checked the Quasar and Apps operation/faults (if any) from customers to understand if there were further instances of discharging faults/auto pausing. A decision was then made whether we can proceed with main trial installs. CrowdCharge continued to monitor the pilot sites, and new installs, to understand if these tested faults were consistent.
The CrowdCharge Atom App
This is a project specific process which has been developed and allows the onsite installer to take photos of the completed install, and complete post installation safety checks. The app allows the installer to commission the charger to the CrowdCharge platform via the CrowdCharge commissioning team.
This software process was designed during November 2020 and implemented across the following two months by a subcontractor – Jumptech. During January 2021, the app was released for initial pilot testing and integration development work with CrowdCharge CRM. The pilot test with installers using this app was successful.
V2G Charging Optimisation Platform (COP)
During winter 2020, CrowdCharge re-developed and re-designed their V2G COP and structure. This has been tailored to be able to accommodate the charging and discharging, as well as the parameters and control required to operate the one-year customer trial.
A new operational portal has been developed and created in January 2021. The portal displays each asset installed as well as displaying the charger cycle data for that particular asset. This operational port will prove useful during the trial lifecycle as it can be used to help diagnose faults and issues with the Chargers and/or Controllers. This operational interface allows CrowdCharge to easily view charge cycle data for any of the chargers installed.
Progressive Web-App (PWA)
For the purpose of operating the customer trial, CrowdCharge have designed and developed a PWA. This will be shared with all participants during the first stage of the trial to allow them to select to either ‘Optimise’ their charging/discharging, to provide their EV with the cheapest and green electricity possible, or the ‘Charge Now’ option which immediately charges their EV to the battery maximum SoC, irrespective of the charge cost and carbon intensity. This will be implemented in the Fixed Schedule with local control stage of the trial (from June 2021 to August 2021).
Group Control Analysis Report
This report contains results and commentary on the historical charging data captured between June 2021 and December 2021 on the Electric Nation PoweredUp project, and a simulation-based analysis of what the impact of group limits would have been on these charging events. Key observations are:
V2G smart charging is largely driven by tariffs with charging typically being moved to overnight cheap periods and discharging to the traditional evening peak period.
Combining users on different types of tariffs leads to a diversity of behaviour that inherently reduces peaks in demand so much so that constraining maximum power demand to 30% has a limited impact on charging behaviour.
V2G behaviour results in an increase in both charging and discharge activity, as additional charging is required to compensate for the energy lost during discharge.
As export tariffs are rarer and lack diversity, it is more likely that those using them end up with co-ordinated discharge operation. This co-ordinated behaviour led to the greatest absolute power peak during discharge. However, this effect is largely driven by existing tariffs that may evolve over time as the concept of traditional evening peak erodes. In addition, when combined with larger numbers of users without export tariffs, this impact is diluted.
Example daily ToU (time of use) tariffs and import and export data
Energy supplier’s tariffs do influence customer’s behaviours as to when the majority charge their EV and discharge onto the grid, this is based on when it provides the best value to the customer. Import aligns directly with the cheapest demand rate and the export aligns directly with the best export tariffs, which is at peak time of day. This shift in consumer behaviour is encouraging as it shows that by working with an energy supplier it should be possible to influence consumer behaviour with the right incentive and this aligns perfectly with the policy direction of the UK Government. The key to a successful energy transition is collaboration with energy suppliers in order to leverage flexibility services and these results provide a first glance into what may be possible in the future, more work needs to be done with Suppliers but these early insights are an encouraging first step. As we delve deeper into these trials and types of services it will be essential to collaborate to gain access to consumers and gain the relevant insights into what constitutes the right incentive.
Lessons Learnt
The most relevant lessons learnt from the project have been detailed below along with a recommendation:
Multiple EVs reside at 1 address
Upon launching customer recruitment, approximately 17% of enquiries detailed they have more than 1 EV at their property, this drastically increased from the previous Electric Nation 1 trial with only a handful of multiple EV applications. This multiple EV scenario became problematic for a couple of reasons:
1) Only the project charger would be the live charger at the participants home and as V2G services are only compatible with an EV with CHAdeMO connection, the 2nd vehicle would be unable to charge (if it was anything but a Nissan EV).
2) There are technical software complications with the charger demand management software platform which require further technical thought, this meant the demand management software could not control two chargers at any one time. A solution to this would require more consideration and additional development time which are not in scope of this project. These reasons resulted in CrowdCharge unable to offer a space for households with more than 1 EV.
Recommendation: A future trial should specifically investigate V2G services strategies to offer a V2G solution for households with multiple EVs.
Long lead times for G99 grid application approval decisions
Out of all G99 grid applications which received a decision (109), on average this took 62 days to receive an approval decision. This is not a suitable review period timeframe for a BAU environment where V2G chargers are being installed in large numbers throughout the UK. Although the current lead times on average for EVs are many months, in a BAU world this is not acceptable to customers to understand if they are able to have a V2G charger installed.
Recommendation: NGED and other DNOs should review their G99 grid connection application process, with a focus on improving the lead time for customers. Perhaps an interactive website where customers can enter their postcode to gain an instant preliminary indication on the likelihood that their G99 application would be approved, with information provided if reinforcement works are required. From here the customer can decide whether to submit a full application with their installer. Finally, NGED should look to improve the decision time to a matter of weeks though process efficiencies in preparation for the inevitable uptake of V2G charger applications.
Upon this recommendation, NGED have plans in place to have an online connections portal for precisely the reason stated above within the ED2 price control period.
Different responses from different Planner Depots
There does not seem to be a standardised response from NGED depots upon submission of G99 application, which can cause confusion with installers. Each depot seems to provide a different approval letter too. Dissemination of NGED policy regarding export limitation at the inception of project which involve export, would help streamline process for project partners and customers in future projects.
Recommendation: Updating installer training documentation, review application process across all depots and review approvals letters to standardise.
Upon this recommendation, NGED have created two centralised planning teams and an online portal to standardise the connections process.
Auto fault detection process for non-network hardware on future innovation projects
A fault detection process is required to be able to proactively detect faults with the charging hardware or inaccuracies in charge cycle data within the software as opposed to customer only reporting faults to project partners. This helps to improve customer experience and ensure chargers are in an operational state for the customer trial to ensure maximum value and learning for project partners/NGED.
Furthermore, Fault fixing was difficult to begin with, as CrowdCharge did not have sight of the Wallbox Quasar charger’s configuration, and how various parameters in the Wallbox App can interfere with the CrowdCharge Smart Scheduling stage of the trial. In future projects, agreeing with charger manufacturers to provide information from their backend platform for various parameters daily or via an API, such as Firmware Version, App Version, Locked Status, and Schedule applied etc. Participants are able to change this information - either by mistake, or on purpose daily - which in turn could cause an issue. This extra information helps CrowdCharge in diagnoses and ruling our faults - without this information the fault fixing process would be more challenging.
Recommendation: Create an auto-detection fault process for future projects involving innovative non-network hardware. Furthermore, gather as much relatable technical information as possible to inform the fault diagnosis and fixing process, with the aim to improve the customer experience and quality of data on the project due to the reduction in hardware downtime.
Requirement for frequent high-level management meetings with hardware suppliers, if the technology is highly innovative and untested at scale in a BAU sense
Requirement for a high-level monthly management meeting with hardware suppliers throughout the trial (if the technology is highly innovative and untested at scale in a BAU sense), not just when a major issue occurs. This helps to ensure a constant line of communication is open to key decision makers to ensure they understand the issues and problems occurring in the trial (the meeting can of course be cancelled month by month if there are no major updates, but the repetition is essential.) This also ensures that any issues which are raised, can be escalated accordingly to ensure the resulting fix is applied as efficiently as possible.
Recommendation: Initiate regular performance monitoring to capture emerging issues early.
The size of the project partner has an effect on the timescales taken to complete some of their actions/objectives.
Working with large organisations took a considerably longer amount of time for progress and internal decision to be made/signed off. This caused a delay in agreeing and issuing an Energy Proposition to applicants on the trial. This is a key learning from a project management point of view, we must allow sufficient time when dealing with large organisations, especially as these organisations have a large amount of expertise and resource to offer a project.
Recommendation: Take this learning into consideration when designing project plans for future innovation projects, to allow ample time for larger organisations.
Planning limitations
There were some planning limitations for some customers and either reinforcement was required to enable the added generation or customers chose not to participate with the trial. This was especially prevalent in areas where there were large amounts of domestic PV (Solar) installed. As all generation is treated equally and the assumption is that generation is exporting to its limit all the time (for planning purposes), adding additional 6.8kW generation on top of the PV generation meant that the network would be outside of safe operating parameters and required larger conductors and assets to accommodate.
Recommendation: Planning needs to reflect the type of generation being connected, solar (PV) only exports during the day, whereas V2G or V2H will either import or export during the evening hours, so to assume that the V2G export is in addition to solar is misguided
