Back in late 2020 we had a look at the impact that an electric vehicle (EV) charging investment would have on the network costs for a 24/7 retail business – think fast food restaurant or service station – across the different Australian distribution network providers.
Why did we do that? Well, we know that many businesses with a decent sized property portfolio and consumer brand are already looking to position themselves as the “charger of choice” when their customer’s EV is away from home, e.g., Starbucks
So why the focus on network costs? Because they make up a significant part of the bill-stack for electricity users both in Australia and elsewhere and are often sensitive to ‘peaky’ energy use thanks to demand charges (and EV chargers are the poster child for their potential to generate peaky load shapes). Commercial and industrial customers on un-bundled “pass through” arrangements have visibility of this portion of the cost stack, so can actively look to manage their impact on the network and hence the costs they have to incur.
As Gridcognition expands our support for customers in markets outside of Australia and New Zealand, we thought it would be interesting to undertake similar analysis for another progressive, de-regulated energy market – the UK. This isn’t a competitive thing – it’s not The Ashes of the distribution network businesses – but it is a chance for us to learn a little and maybe draw the odd parallel or two.
Same as for the Australian analysis, so:
1. We’ve taken a commercial load shape for a fast-food restaurant, just the kind of business that is looking to offer EV charging to customers, and then placed our virtual restaurant in 14 different cities, from Southampton in the south to Aberdeen in the north, with each city located in one of the 14 major distribution network operators (DNOs) that service the UK electricity system.
2. For each network we’ve selected the pithily named LV Site Specific Band 2 tariff from their Distribution Network Use of System (DNUoS) schedules. This tariff was chosen as it’s for low voltage, commercial customers with load between 80 and 150kVA, which matches our restaurant nicely. Important to note that we’re newcomers to the UK scene and so please correct us if this was the wrong tariff choice! Example from Scottish and Southern Electric below.
3. Whilst the rates and time-of-use periods vary between DNOs, each tariff includes volume (kWh) based pricing as well as some capacity (contract demand in kVA), excess demand (kVA) and reactive power (kVArh) elements. There’s also a fixed daily charge for each.
4. To establish a baseline, we’ve then calculated the DNUoS charges our restaurant would pay based on their business-as-usual energy consumption. See chart below.
5. We’ve then simulated the load shape with the addition of a 50kW EV charger and used a simple probabilistic model to determine when the charger is used (the analysis is highly sensitive to this usage model). Grab below of me building up the simulation in the Gridcognition software, in this case adding the EV charger to the site in Cambridge (UKPN Eastern)
6. Each charging event is assumed to be 30 minutes in duration and deliver 25kWh, so about enough time to enjoy a burger and use the facilities, and enough charge to drive about another 140 kms or 85 miles. This generates a new load profile, seen as the red trace in the graphic below.
7. We’ve then re-billed this new load shape against the respective network tariffs to work out the additional DNUoS charges payable purely from use of the EV charger over one year.
8. Finally, we’ve taken that additional cost and divided it by the total energy delivered through the charger (about 40MWh for the year) to come up with a simple cost-recovery figure in p/kWh. That is, the price they would need to charge users of the EV charger in order to cover the additional network costs.
In the interests of keeping this post to a manageable length I won’t talk to the results too much as a picture often does better than I could do anyway. Just a handful of observations and thought provokers, subject to all the normal disclaimers about this analysis being hugely sensitive to the particular inputs and assumptions used:
- There’s clearly a significant spread in terms of the way the different DNOs structure their tariffs. You see similar trade-offs between fixed charges, volume-based charges and demand charges that we see in Australia.
- Demand-based costs are more sensitive to the EV charger’s impact than volume-based. Again, this is something we see in Australia. It’s also something that would start to balance out as the capacity factor on the EV charger utilisation starts to increase.
- Less densely populated areas seem to suffer more than more densely populated areas, again something we see in Australia, although our definition of population density and distance is on a very different scale!
- At first glance, adding EV charging hardware behind existing metering points seems to make more sense than creating a new connection.
- Manchester continues to best Liverpool, and not just in the Premier League.
Electrification of the transport sector has huge ramifications for energy systems around the world; for the network owners, for the electricity retailers, the regulators grappling with so much and so rapid a change, and for energy consumers who face a whole new set of questions, risks and opportunities.
For businesses, the potential to leverage vehicle fleets, property portfolios, brands and customers in order to benefit from this shift is huge. I don’t think there’s much doubt that we’re at the beginning of a multi-decade feeding frenzy in the sector as folks jostle to own and control the best lucrative infrastructure and customer relationships. But there’s a lot to think about, with commercial success or failure in energy projects often decided at the margins, and with a new breed of highly steerable energy assets re-writing the rule book on how energy consumers engage with the grid.
It’s fantastic to see consumers and businesses driving the agenda and getting on with it, particularly in Australia where the political leadership at the Federal level has been near absent, and great to see the more agile and forward-looking utilities supporting and accelerating, rather than stalling and blustering.
The fine print
- The model to estimate EV charger utilisation was based on public data (Google) on visitation of some well know fast-food restaurants and then applying some random utilisation within that. This ended up with about 1,628 x 30-min charge events over the year (9% utilisation) and those will have been weighted to when the business was busier.
- Tariff details taken from the respective DNOs 2022 DNUoS price schedules
- Demand/capacity charges in the baseline scenario based on the max kVA for the year. This assumes the MIC is exactly ‘right-sized’ for the load, which might be an over ambitious assumption. Demand/capacity and excess demand charges for the EV charger scenario again based on the max kVA for the year with MIC set to the tariff band max of 150kVA
- All other energy supply costs, including retail, transmission and market costs have been excluded.
- For simplicity EV charger losses are assumed to be 0%
- DNO boundaries curtesy of a shape file provided by National Grid. Thanks you guys!