How a lack of joined-up thinking could block EV charging expansion

A number of letters published in recent issues of E&T have discussed issues relating to electric vehicle charging and smart metering. As part of my work with IEC technical committees covering smart grids and smart metering, I have been looking at some of the challenges that will arise as EVs become more popular.

The UK government, in common with other countries, is keen to introduce real-time pricing that incentivises the use of energy during off-peak times. This is seen as a way to reduce carbon emissions by influencing consumer behaviour through pricing. The plan is to introduce consumers to half-hour pricing (or 15 minutes in continental Europe) through tariffs downloaded to smart meters. Prices are generally set the day before, so the consumer can plan activities the day after.

This becomes rather difficult and time-consuming, so it is envisaged that there will be an opportunity for demand-side response service providers (DSRSPs) to enter the market who will provide load-management services to the generators and grid operators, and offer attractive tariffs to consumers. It isn’t clear how this will work at scale given that consumption may exceed supply when prices are low and there may be excess supply when prices are high, but the DSRSPs will have to deal with that.

There is a view that smart appliances will be able to shift load and act as virtual power stations at times of high demand; however, most modern appliances are very energy-efficient and individually are insignificant. A test on a wash cycle on my own washing machine, for example, showed it took 1.2kWh for a cycle, costing less than 25p, so it is difficult as a consumer to see what the incentive would be to do my washing at an inconvenient time.

The picture changes when EV charging is considered. A 7kW charger would take up to 10 hours to charge a battery from flat, costing around £14 at a rate of 20p/kWh. It’s this business case – encouraging consumers to charge their cars overnight by making energy cheaper – that receives the most attention, as it manages load on the grid.

Generation and load on the grid together comprise only one aspect of the issue. The other –  which seems to have been overlooked but is where the major challenges will be – is the low-voltage network.

Readers of E&T have reported that despite having a 100A fuse on their incoming supply they have not been allowed to run their 7kW (30A) charger at full capacity, but have been limited to 3kW (13A). This is because there is another limiting factor in the supply – the fuse in the LV substation will be rated typically between 200A and 300A.

There are typically 100 houses fed by each three-phase LV distributor cable. On the estate where I live there are 99 houses, so for easy reckoning that is 33 houses per phase. If we assume that the fuses are 300A per phase and the average load per home is 1kW (which is the figure used by network planners) then at 33kW load there is 143A flowing through the 300A fuse on average and there is plenty of headroom for short peak loads such as kettles and ovens.

If we now introduce EV charging, it only takes five vehicles charging at 30A for the fuse on the phase to be running at maximum load with no headroom. And if the fuse is 200A then the maximum is only two EVs.)

During the night, there could be more capacity and the energy price will be low, so assuming EV owners can now take full advantage of the fuse capacity with minimal headroom it may be possible to charge nine cars per phase in a 10-hour period. Assuming one car per house, the other 24 cars would have to wait for a different night, but with planning all cars could be charged over 4 nights. Average car ownership where I live is two cars per house so the issue becomes more complicated and EVs may have to be charged elsewhere or during the day when energy is more expensive.

Clearly there will have to be some form of booking system for use of the LV network. EV owners may want to take advantage of a low-cost period for energy to charge their car, but those network slots could already be booked by someone else. The half-hour pricing for energy is only known the day before, so having to reserve a slot a few days in advance will mean the energy price will be unknown at the time of booking.

The situation becomes even more complicated when phase imbalance is considered. In many cases, the LV distribution cable relies on the sheath for carrying the neutral current. It is assumed that the phases are balanced and hence the sheath (or neutral conductor if there is one) is rated at a lower capacity than the phase conductors, therefore there is a danger of destroying the cable if the loads are severely out of balance. Similarly, the transformer will suffer during unbalanced conditions.

The booking system now becomes more complex; the EV owner can try and get a good price for charging their car, and reserve a slot with the distribution network operator (DNO), but that booking could depend on what the neighbours are doing on the other two phases and which slots they’ve booked. If they have decided not to book and the LV distributor cable is predicted to be out of balance then the deal is off.

The next issue is that the DNO doesn’t meter the transformer end of the LV distribution cable, so doesn’t actually know how much current is going down each phase until one of the fuses blows, by which time it’s too late. That problem can be addressed by installing metering at the transformer end, but unfortunately that doesn’t completely solve the problem either because the smart meter doesn’t report which phase it is on, and neither does the ‘smart’ charger. The DNO could detect there’s an issue with an overloaded phase at the substation, but it won’t know which houses are causing the problem so it can’t send any curtailment messages to the chargers or smart meters.

DSRSPs are expected to play a key part in controlling the demand-side response. It is envisaged there will be competition in this market, so several DSRSPs could be involved in the supply provided over a LV distributor cable, but the critical parameter that has to be controlled is the current flowing through the fuse in the substation. It is far from clear how this market will work, but presumably these companies will be expected to bid for capacity on that fuse. This then raises the question – does the highest bidder get the most capacity? What happens to the people who signed up with the lowest bidder, are they at the back of the queue for charging? What happens when the DSRSP didn’t take their contracted load and that leads to the infrastructure being under-utilised? Clearly there will be an issue to address here that the DNOs are going to have to deal with.

One way of overcoming the bottleneck at night would be for those who are at home during the day to charge when everyone else is out at work and the local LV network is lightly loaded. This would help balance the load on the LV network over the full 24 hours, but unfortunately the energy cost would be high during the day and hence a disincentive for anyone to charge their car.

Another issue would arise if there was an overload in the distribution network that meant the charger had to curtail the demand. This would either result in the car not being fully charged, or it running over its contracted cheap period and the owner being faced with a higher energy charge than was originally agreed.

Returning to the case where the DNO limited a charger to 13A (3kW), the scenario mentioned above shows that in the case of 7kW chargers only five could charge at peak times (or two on a 200A fuse) or nine through the night. Those numbers could double if the 13A limit were introduced, although it would take up to 19 hours to charge the car. This would produce a short-term fix, but not a long-term solution. What could be even worse for the DNO is if people install 13A chargers at scale that probably would not have to be registered or be ‘smart’. The DNO wouldn’t know who is doing what or on which phase and would not be able to control their network.

Vehicle-to-grid schemes, where EVs feed energy back into the network, have been proposed as a way of supporting the grid at times of high demand. This will cause even more issues for the DNOs because there could be a situation where some vehicles are supplying energy to others on the cable linking the houses, bypassing the fuse in the substation. A cable could be overloaded without any form of protection.

There is an expectation by some that the smart meter rollout will have a part to play in managing EV charging. As I’ve explained, the smart meter doesn’t provide data to indicate which phase it is connected to, and is therefore not much use to a DNO who needs to control the load in each phase. The meter can provide near real-time consumption data that could be used to control the charger, but the meter only measures total consumption and doesn’t know where the energy is going to: it could be the car or it could be the oven or shower.

It can however provide the tariff information to the smart charger if required. The smart charger could be fitted with a meter, but the Government’s Office for Zero Emission Vehicles has specified that it wouldn’t need to be accurate or comply with measuring instrument regulations, as the data will only be used for demand-response purposes, overlooking the possibility of readings being used by the EV-charging service provider for billing purposes. The ability of EV charger meters to be linked to smart meters in domestic ZigBee networks was removed from the smart meter implementation specification some time ago, and is a missed opportunity.

The challenges I’ve described highlight the lack of joined-up thinking surrounding EV charging, with numerous bodies and organisations looking solely at their own individual issues. The IEC System Committee for Smart Energy is working on a system-level approach in an attempt to coordinate the development of compatible and interoperable standards, but this will take some time and require the co-operation of numerous technical committees and stakeholders. There is an opportunity for the IET to play a role in the EV-charging infrastructure.

John Cowburn is a chartered engineer and IET Fellow who chairs the BSI and CENELEC technical standards committees for electrical energy metering and control (TC13).