## DOE releases report from 6 projects evaluating EV charging impacts on grid and customer charging behaviors

##### 21 December 2014
 Charging patterns with (TOU) and without (RES) whole house time-of-use rate during summer weekdays at Progress Energy, one of the participating utilities. (Peak period is in gray.) Click to enlarge.

The US Department of Energy (DOE) has released a report detailing the findings from six utilities which evaluated operations and customer charging behaviors for in-home and public electric vehicle charging stations. The work was done under the DOE’s Office of Electricity Delivery & Energy Reliability’s (OE) Smart Grid Investment Grant (SGIG program).

This report provides the results of these SGIG projects to help individual utilities determine how long existing electric distribution infrastructure will remain sufficient to accommodate demand growth from electric vehicles, and when and what type of capacity upgrades or additions may be needed. The report also examines when consumers want to recharge vehicles, and to what extent pricing and incentives can encourage consumers to charge during off-peak periods.

The electric power industry expects 400% growth in annual sales of plug-in electric vehicles by 2023, which may substantially increase electricity usage and peak demand in high adoption areas. Understanding customer charging patterns can help utilities anticipate future infrastructure changes that will be needed to handle large vehicle charging loads.

The six SGIG projects evaluated more than 270 public charging stations in parking lots and garages and more than 700 residential charging units in customers’ homes. Participating utilities were:

• Burbank Water and Power (BWP)
• Duke Energy (Duke)
• Indianapolis Power & Light Company (IPL)
• Madison Gas and Electric (MGE)
• Progress Energy (now part of Duke Energy as a result of a merger in 2012)
• Sacramento Municipal Utility District (SMUD)

Because there are relatively few plug-in electric vehicles on the road today, the SGIG projects focused on establishing the charging infrastructure with a relatively low number of stations and evaluated a small number of participating vehicles.

Although project results showed negligible grid impacts from small-scale electric vehicle charging today, they also gave utilities important insights into the demand growth and peak-period charging habits they can anticipate if electric vehicle adoption rises as expected over the next decade.

Major findings were grouped in three categories: charging behavior; grid impacts; and technology issues.

Charging behaviors. The studies found that the vast majority of in-home charging participants charged their vehicles overnight during off-peak periods. Where offered, time-based rates were successful in encouraging greater off-peak charging.

Public charging station usage was low, but primarily took place during business hours and thus increased the overlap with typical peak periods. Plug-in hybrid owners frequently used the (often free) public stations for short charging sessions to “top off their tanks.”

Grid impacts. Length of charging sessions and the power required varied based on the vehicle model, charger type, and state of battery discharge. While the average power demand to charge most vehicles was 3-6 kW (roughly equivalent to powering a small, residential air conditioning unit), the load from one electric vehicle model can be as much as 19 kW—more than the load for most large, single-family homes.

Technology issues. Installing a 240-volt charging station, which typically charges 3-5 times as fast as a charger using a standard 120-volt outlet, requires a licensed electrician and occasionally service upgrades. Public charging station installation had high costs and required substantial coordination with equipment vendors, installers, and host organizations to address construction, safety, and code requirements.

In addition, low usage at public charging stations will require longer capital cost recovery without substantial growth in usage.

Some utilities also found residential interoperability problems in communication between smart meters and charging stations. SMUD found that the two devices only connected successfully about 50% of the time during load reduction events.

The "one vehicle model" can only be the Tesla Model S (and the few Roadsters running).  Either can be set to pull only a limited amount of power, even down to charging from a standard extension cord (albeit not at all quickly).

It's no big deal to pull 240 VAC 75 A over service capable of handling 150-200 A.  The main neighborhood issue would be making certain that the local pole transformer is not overloaded.  J1772 allows this to be managed; I'm not sure if Tesla has a protocol for it.

With renewable energy costs in some countries passing grid parity (Au is one), and talk of a death spiral where consumers leave the grid causing prices to rise and profits to fall, it should be possible to retain customers and absorb increasing R.E. via smart grid spot pricing where off peak pricing could be matched by augmented especially solar supply.

This is already substantially the case when air conditioning and refrigeration load demand follow supply.

It would be nice to see a way to incorporate an availability or cost signal incorporated into the supply.
Possibly via information injection over transmission line at a localised network.

That could be incorporated at (one of) the step down transformers.

I'm sure that people working in the supply feild are looking at the reality of inevitable change as it plays out.

With renewable energy costs in some countries passing grid parity (Au is one), and talk of a death spiral where consumers leave the grid causing prices to rise and profits to fall

That is only happening because of net metering and flat per-kWh pricing.  The actual cost structure of the grid is very different, with around half the delivered cost of electricity coming from grid services rather than energy per se.  It's costly just to have the connection ready to provide power on demand, whether or not power is requested.  Those are the costs that net metering unloads onto other grid users.

If all users paid for capacity separately, and energy at hourly market rates, things would be very different.  First, net metering users would receive no payments for capacity (they provide none), would pay for capacity, reactive power and such, and would only be paid for back-fed power at wholesale market rates.  They could drive those rates to zero and the utility would still get paid for the services it is providing:  death spiral averted.

Yes, this would destroy the economic case even for subsidized grid-tied PV systems unless subsidies were massively increased.  This demonstrates just how uneconomic grid-tied PV really is, and how misleading the LCOE metric is.  Calculating the Levelized Avoided Cost of Energy (LACE) is difficult but leads to much better decision-making.

It would be nice to see a way to incorporate an availability or cost signal incorporated into the supply.

There's already a partial mechanism for this, in the form of interruptible service at preferential rates.  I had such service going on 5 years ago now.  But more detailed information, and systems capable of using it for arbitrage, is better... so long as it can't be hacked to damage or crash the grid.  That, and related security issues, are going to be one of the bigger problems going forward.

EVs are a new market for utilities. They should more than make up for what is lost to end-user solar and efficiency.

Some utilities are starting to realize what they have to gain. SoCal Edison recently put up $350 million to assist with the installation of 300,000 charge points, mostly workplace and apartment locations. A death spiral is quite unlikely. Utilities can sell late night wind and hydro for less than customers can store power. And few people are going to want to deal with running their own utility companies. (Australia is going to be interesting to watch, but that's a special case.) In my case 402 KWH 27.5c& 29.5c,the electricity charge is au$ 213. And the service charge is $222 incGST. Giving an average ~60cKWH. I wont attempt to give current costs to install pv battery stand alone and understand that the payback time is still problematic esp owing to relatively high(er) cloud suboptimal insolation @~34o sth. If I was lucky enough to have a bev or phv as substantial storage battery component, I would be ahead by virtue of 1: Not needing a large battery bank. 2; Having a practical dump load ( pumping is another) to get utilise the entire panel output. A downsized storage to supply refrigeration etc standing loads would be necessary which of course brings me back to the ubiquitous need for still expensive battery storage. I agree that hacking is a difficult problem not made any easier by expanded I.T. penetration however it would seem that the existing internal I.T. infrastructure is not exempt from same. And yes I do receive an (overly?) generous subsidy but on the other hand the high subsidy driven uptake has seen a reduction of P.V. cost by >2/3. This is helping to drive solar uptake and benifits future consumers. My panel are likely ~ 16 -18% efficient and improvements in the past 5 years see a doubling of efficiency with a KW price reduction of 2/3. Bring on the batteries and vehicles and I will certainly be better off economically as well as the political security balance of payment as wella s ceaner air anda better future for the next generation. Some utilities are needs to realize what they should gain. SoCal Edison recently placed$350 million to assist with the installation of 300, 000 charge items, mostly workplace and condominium locations.A death spiral is fairly unlikely. Utilities can sell evening wind and hydro at under customers can store strength. And few people will want to deal with running their particular utility companies. http://www.carbuyer.co.uk

I read that California made it legal for renters to install chargers, I assume that meant home renters. There were fewer than 50,000 EVs sold in the U.S. during 2014, at this rate the grid should not suffer any overloads due to car charging at night.

EPRI studied night charging of millions of EVs more than a decade ago, this shows that EPRI was right, you CAN charge EVs at night, with so few we should not worry about it.

"..400% growth in annual sales of plug-in electric vehicles by 2023.."

That means maybe 250,000 EVs will be sold in the U.S. in 2013, considering there are 14 million vehicles sold each year, that is less than 1 in 50 cars. 400% increase sounds like a lot in 9 years, but it is not.

I may have added an extra zero to get 300,000. What I'm finding now is 30,000.

The comments to this entry are closed.