A new NREL study finds that large volumes of EV users using high-power Level 2 charging equipment at home could pose a risk to grid stability
Predictions as to the impact of widespread adoption of battery and hybrid EVs on the grid have varied wildly, from the concerned to the ridiculous. But the fact that even EV enthusiasts must recognise is that the switch to electrified transport may require some structural changes in power infrastructure, and potentially some habitual changes from consumers.
A new study from the US’ National Renewable Energy Laboratory (NREL) suggests that unless these efforts are coordinated, plugging in more cars could cause problems for the US grid. In his paper “Impact of Uncoordinated Plug-in Electric Vehicle Charging on Residential Power Demand,” transportation and energy systems engineer Matteo Muratori aims to quantify some of those problems, and identify some solutions.
His first point is to note that EV charging will affect the grid in two distinct ways. First, it will stoke additional demand in residential households, largely at off-peak times when vehicles are charged overnight, which may add complexity to managing power supply. Second, that EVs themselves can also provide backup in the form of vehicle-to-grid (V2G) storage, which can offer additional flexibility.
In general, more EVs in the transport mix means that overall electricity demand will increase, but it will also mean the shape and timing of that demand will change too. Muratori’s study uses several scenarios of EV market share and adoption, from 3% of the fleet to 100%. In his reference cases, uncoordinated charging essentially refers to drivers arriving home and plugging in their cars straight away, but all at different times.
Modelling average power demand from 200 residential homes and 348 vehicles in the US Midwest, he found that a small percentage of EVs using an uncoordinated Level 1 charging regime would have a muted effect in terms of overall demand increase; a 10% EV mix, for example, would push total electricity demand up by only 5%. However, Level 2 charging (at 240V) puts additional strain on the system, as demand both increases and becomes much more variable.
Modelling each scenario, he notes that at a 50% market share the effects become significant: “Level 2 charging leads to higher power demand concentrated over shorter periods of time, creating higher and more abrupt peaks in the overall demand.”
Muratori also investigates the effects of charging on individual neighbourhoods – in this case how charging regimes might affect six individual households fed by one transformer. The results here are significant – peak transformer demand with 0% EVs sits at 29,855 kW, but rises by around half in a 50% EV scenario using Level 2 charging to 47,132 kW.
This is also something of a double-edged sword –Muratori adds that “increasing the charging power leads to reduced charging times, and thus fewer vehicles are simultaneously charged when a more powerful charging infrastructure is available” – but serves to illustrate far greater variability in transformer usage than would be experienced currently.
Transformers working overtime will also burn out quicker. The study adds that “the expected life of a transformer decreases by two orders of magnitude when the maximum load factor reaches values 50% above its nominal capacity,” meaning utilities may have to upgrade many neighbourhoods to more robust equipment, or risk replacing it much more often.
The effects of this may also be more pronounced in clustered areas, e.g. areas where affluent or environmentally conscious consumers have high EV adoption rates, leading to even more variability in demand between neighbourhoods, and potential “pinch points” in regional grids.
These problems are unlikely to occur in the short term. The simulation concluded that an EV market share of up to 3%, equal to about 7.5 million vehicles in US, does not significantly impact the aggregate residential power demand, and given that EVs made up just 1.1% of new sales in 2017, that level of uptake is still years away.
However, Muratori’s findings are important. His conclusion adds that: “Future research should focus more heavily on understanding consumer driving and charging behaviour (to better estimate charging requirements) and the nuances determining the choice of residential charging infrastructure.”
Where previous studies assumed that utilities would be able to control when consumers charge occurs, Muratori points out that this might not be the case. A lack of coordination, or a lack of other flexible supplies such as V2G, could therefore begin to affect overall grid reliability as the proportion of EVs rises.
It also presents a powerful argument for more managed “smart charging” in which energy management systems are able to control and vary the supply of power to and from the grid. Muratori argues that this “might provide significant opportunities for grid operators to both limit the impact of PEVs on the electric power system and facilitate system operations (for example, enhanced demand flexibility will facilitate the integration of non-dispatchable renewable energy sources) while also allowing for cheaper vehicle charging.”
“Realizing the full benefits of vehicle electrification will necessitate a systems-level approach that treats vehicles, buildings, and the grid as an integrated network,” added NREL’s associate lab director for Mechanical and Thermal Engineering Sciences Johney Green Jr in a statement accompanying the paper’s release.
The issues posed by high-power home charging are also likely to diminish with the provision of adequate public fast-charging infrastructure too, but studies like this show that in as the energy system becomes more connected, discussions must be had now between all the parties involved – including government, utilities, grid network operators, charging providers and motorists groups – if EV uptake is to be a success.