Could the Ontario Grid support an all-electric fleet?

[wayner]

One of the criticisms that you hear against EVs is that the grid couldn't support the power draw required. I am trying to do the math for Ontario - please give me feedback to see if I am correct.

There are 6.7 million light vehicles in Ontario so I will use that as the total fleet number. The distance that I have for average distance travelled is 16,000km/yr which is 44km/day. Going 44km in my P85 Model S will use about 11% of the battery or 9.3kWh. The average person could "fill up" in just under an hour with a 240V 40A connection. And this assumes that a Model S is a good proxy for the average vehicle in terms of fuel/power consumption.

Multiplying this by 6.7M gives you a daily draw of 62.3GWh. Spread out over 12 hours, assuming that most charging would be done from 7pm-7am gives an average draw of 5GW. The total capacity of the grid is around 20GW but, of course, that can vary as some generation types, like wind and solar are intermittent.

But it seems to me that there likely is enough capacity to supply that additional amount of power - as the draw at night is generally well below 15GW but that may not be the case on very hot summer evenings. Currently (10am on Saturday) the demand in Ontario is 12.8GW. One issue might be managing when EVs are charged - it might make sense to have some randomizer or some process of managing which vehicle is charged when as you don't want all of the vehicles charged at once.

Some other ways of looking at things. The total energy required in a year to power the fleet by my calculations is 22.7TWh. In the previous few years the total output in Ontario was 137TWh so this would increase overall demand by about 17%.

I may be slightly underestimating the power consumption since there would also be vampire losses of a kWh per day or so depending on the vehicle.

Any comments on whether my math is correct?

 

[skrenes]

I think you're in the ball park. I noticed you assumed a P85 only goes 400 km on a full charge and that a full charge is 85 kWh. I believe those numbers are more like 420 km and 77 kWh respectively but then the charging is only 90% efficient, so the draw is about 85 kWh. A Nissan Leaf would use quite a bit less energy in low-speed, stop-and-go scenarios. In any case, you've shown that the current grid could handle it. In reality, as the grid is pushed to its limits, the energy companies would raise the price of electricity (or incentivize things like solar PV). In either case that would make adding solar panels to residential roofs even more lucrative, thus increasing capacity.

Most detached houses can produce enough energy to power the entire house -- some with better south facing surfaces have enough for the entire house, a couple electric cars, and then enough to pump back into the grid. My house in Calgary, for instance, uses 4200 kWh/year plus another 3000 kWh/year for my Model S. Each kW of panels in Calgary produces 1290 kWh/year. So I only need 5.6 kW of panels to break even. That's easily doable on pretty much any detached house around here.

 

[wayner]

I have 10kW of panels on my roof and I produce roughly 13MWh per year. However, I use a lot of power as well - about 52MWh. I have a pool, hot tub, two ACs, etc but gas heating and gas water heating.

But using solar panel to directly charge your car will likely require batteries as most of us are not home during the day when you get peak solar generation. That increases the cost considerably.

 

[daniel]

The claims that the grid could not support a 100% electrified fleet ignore the basic fact that the fleet will not switch to electric overnight. The percentage of the fleet that is electric will grow very slowly as EV manufacturers slowly increase their production capacity, electric cars slowly gain popularity, and gas cars are slowly taken out of service and replaced by electric. As this happens, the grid will be slowly upgraded to handle the increased demand. Both the carrying capacity of the power lines, and the generating capacity of the power plants will grow as needed. Increased demand will tend to push the cost of electricity up, but people installing PV panels on their roofs will tend to push cost down, as well as shunting some of the charging from off-peak by grid to daytime directly from the roof solar panels.

Also, the majority of car charging will be done at off-peak times (which you take into account, but the nay-sayers don't), or directly from home-installed solar panels as noted above, so that the pressure on the grid will be far less than the overall increase in energy usage.

In conclusion, the nay-sayers are trying to compare present-day grid capacity with a distant-future demand and a presumed lack of any planning.

I have heard that the grid in North America is in bad shape and needs a lot of upgrading, irrespective of any pressure on it from electric cars.

 

[Phillip L]

I read recently in a rebuttal to an anti-ev newspaper rant (where?) that the real limiting factor as more ev's come online is less the everall capacity but the ability of often quite old neighbourhood transformers to handle the extra load, but even that consideration is a future worry. Since ev adoption will be gradual, but not uniform throughout cities (just in my small city of Owen Sound within a block of me; there are, including my own, 3 Model S's, 2 Volts, and a BMW i3; as far as I know there is one other Model S and one or two Leafs elsewhere in the city) we will need only a gradual replacement of those transformers.

 

[jbcarioca]

I read recently in a rebuttal to an anti-ev newspaper rant (where?) that the real limiting factor as more ev's come online is less the everall capacity but the ability of often quite old neighbourhood transformers to handle the extra load, but even that consideration is a future worry.

Accepting the OP demand premise is certainly a worst case in that much of the eventual demand will be less consistent than posited and far lower overall because of increased BEV efficiency.

However, widespread commercial vehicle BEV use will happen also so the overall demand might be much higher ( I have no data on Ontario commercial vehicle fleets so I cannot guess). Bus fleets will probably be the first large scale commercial penetration. A wild guess suggests that the timing of charge demand from those will also be night hours, so non-peak.

I saw an analysis done by the US EPA roughly five years ago that suggested ~50% of all road vehicles in the US could be electrified with no incremental electrical generating capacity, but with localized grid augmentation, principally transformers. The study suggested that advances in energy storage (batteries) and local generation from wind and solar could even displace much of the grid deficiencies while reducing need for 'peakers' to supplement continuous power generators. Sadly, I have no copy nor a way to get one, but I do know many public utilities have made similar evaluations around the world which have often produced similar results.

For Ontario, with majority of power from nuclear, off-peak generation capacity is mostly wasted.
https://www.cns-snc.ca/media/ontarioelectricity/ontarioelectricity.html

Nuclear cannot easily be turned on and off. Thus increasing off-peak demand in Ontario is almost certain to have a virtuous effect of increasing efficiency and reducing cost per unit. Even with incentive rates fir off-peak and BEV use the overall cost per kWh should reduce in Ontario were BEV's do dominate all road-going vehicles.

In areas with different dominant power sources the conclusions can be much different. Possibly the least virtuous option in these terms would be dominant gas-fired plants, which typically can easily adjust to intermittent use requiring only an hour or so to adjust. Coal- fired plants plants are more flexible than nuclear but cannot adapt to hourly demand changes.

Somewhere in this evaluation, therefore, should probably be the virtuous benefit of BEV in stabilizing energy demand by increasing demand in low demand hours. In many cases that effect might result in significant reductions in net electricity unit costs for all rate-payers. This aspect seems to have little discussion. Even BEV-haters should like reduced electricity unit costs.

 

[SageBrush]

Oddly, I cannot remember much teeth gnashing from the concerned public over power draw from AC.

Cynicism aside, the answer to intermittent renewables and heightened demand/draw is a heterogenous mix of generation and use, demand meters and an upgraded grid with regional power sharing. Energy storage has a place, but not that much in a well designed system. It really is that simple, and the naysayers are just FOS. Their typical moaning goes on like this:

NY: "Wind is terrible. It blows at night when we don't need it!"
NY: "PV is terrible. It does not shine at night when we need it!"
NY: "EVs are terrible, increasing the load during the day!"
NY: "PV is terrible. It drops demand during the day to nothing and leads to inefficient/expensive thermal plant use"

 

[ItsNotAboutTheMoney]

Supporting 100% should be easy.
Electricity is currently a just-in-time system where the grid responds to demand.
Peak demand is in the evening and then falls off creating a "bath tub" of spare capacity at night before the morning ramp.
In hot climates, AC use also creates a second, afternoon peak.
Because of the peakiness, grids have a _lot_ of spare capacity.

If everybody drove home, plugged in their EVs and then started charging it would add to the peak. But a simple demand-based or time-of-use pricing system that makes that expensive (which is how it should be anyway!) would shift charging to the night-time bath tub.

At _very_ high proportion demand would then become inverted with evening and night-time demand higher than day-time demand. But there would be two things to deal with that: the first would be a shift in pricing, and the second (and probably more important) would be smart charging.

With a highly plugged-in transportation fleet, transportation demand would become a very substantial part of, or the majority of, total demand. But plug-ins are plugged in for far longer than they need to be to charge fully, and modern vehicles have connectivity and computing power (and computing power generally is now cheap) so those plug-ins would quickly become a massive controllable power and demand sink that would allow for better use of transmission capacity, easier integration of variable renewables and higher capacity factors for conventional generation.

In addition, the implication of total electrification of transportation is that batteries would have to be cheap, which would imply that batteries could also be used affordably for grid storage.

Instead of asking whether grids can handle electric vehicles, we should instead be asking how do we make full use of them.

But even ignoring smart charging, it doesn't _really_ matter whether the grid can handle demand, because you can always build more grid. It is only an economic challenge related to the shape of demand and production.

 

[daniel]

... Even BEV-haters should like reduced electricity unit costs.

There are people who think that BEVs are impractical or uneconomical. There are people who doubt that charging will ever be as convenient as pumping gasoline. Neither of these necessarily hates BEVs. Both groups will be happy to see other folks adopt a system that lowers electric rates for all by buying off-peak power.

Then there are the haters. Hating someone or something that has done you no injury is irrational, and people who harbor an irrational hatred will never accept any rational argument. Such people are lost to civil discourse. However, I think they are a tiny minority when it comes to BEVs. I've been driving electric for eleven years, IIRC, and I very very rarely get negative comments from casual encounters. What I get almost exclusively is, "Boy, I'd love one of those but..." Then "I need to pull a trailer," or "I need to drive farther than that," or "I just can't afford one." These people welcome electric cars and will get one when their concerns are addressed to their satisfaction.

I don't worry about the haters. They're everywhere, in every field, and when it comes to BEVs, they're few. (In other areas, they are more numerous and pose significant problems.)

 

[Saghost]

One issue might be managing when EVs are charged - it might make sense to have some randomizer or some process of managing which vehicle is charged when as you don't want all of the vehicles charged at once.

Any comments on whether my math is correct?

If only the cars had an always on connection to the internet and a way to update the firmware over the air so we could add things...

As I've suggested before, there's a big opportunity here for Tesla, the Utility, and the owners once Tesla cars become a meaningful factor in grid usage. What I'm envisioning is a program where you authorize the power company to schedule your charging at their whim, as long as they get you to your chosen charging target by a time that you choose (presumably right before you leave on a normal morning,) in exchange for a reduced rate on your electricity (either for the car or the whole bill.)

The power company can then shape the demand curve - and depending on how fast the response is, possibly even use this to offset local demand changes - cycling your car down or off when your neighbor's oven or A/C turns on. If the regulators get on board, it might even allow them to safely reduce the amount of spinning reserve they have to keep alive - built in waste in the system that exists to keep the grid stable.

Tesla's gain in this scenario is just that the car is more desirable to consumers since they know they can get cheaper power for it (as if they need another competitive advantage right now.)

There are already deals out there where the electric company gives people a significant savings on the overall electric bill in exchange for the right to shut your A/C off for up to 15 minutes at a time in the middle of peak usage.

 

[SageBrush]

What I get almost exclusively is, "Boy, I'd love one of those but..." Then "I need to pull a trailer," or "I need to drive farther than that," or "I just can't afford one." These people welcome electric cars and will get one when their concerns are addressed to their satisfaction.

People rarely speak their mind -- they rationalize. Your point though is spot on; these people are not haters, they are conservative and/or reactionary.

 

[daniel]

... What I'm envisioning is a program where you authorize the power company to schedule your charging at their whim, as long as they get you to your chosen charging target by a time that you choose (presumably right before you leave on a normal morning,) in exchange for a reduced rate on your electricity (either for the car or the whole bill.)

The power company can then shape the demand curve - and depending on how fast the response is, possibly even use this to offset local demand changes - cycling your car down or off when your neighbor's oven or A/C turns on. ...

It's pretty much inevitable. It's just a question of how soon.

 

[daniel]

People rarely speak their mind -- they rationalize. Your point though is spot on; these people are not haters, they are conservative and/or reactionary.

Often they're not even conservative or reactionary, just not not yet ready to adopt an emerging technology that we all know still has some limitations around price tag and charging.

 

[SmartElectric]

I own two EV's. Charge them predominantly in the cheap overnight period.

A little perspective:

The local energy company was sending me a notice every month showing that I was in the top 40% of electricity users on the street. LOL. I was barely above the mid-point of the street! There are 3 solar installs on my street, which means they use practically zero electricity. This means there are a lot of heavy electricity consumers in my local street who don't give a thought to when and how much they use, and here we are talking about the need for care in the system. The current system has to deal with 1.5kW pool pumps, 2kW AC units, 2kW clothes dryers going on/off across hundreds of houses in the local service area. My little 1.5kW (12A 120V) draw for my Smart ED charging and the 6kW draw for the Tesla (24A 240V) during the overnight period is really not an issue. There are 5 EV's on the street out of 25 houses, with no local transformer upgrades.

Ontario can handle this, we have 6.5 GW of natural gas production capacity sitting idle 90% of the time. Most days the max gas output in 1.5GW, and recently many days when 500MW is the maximum.

 

[wayner]

There are 3 solar installs on my street, which means they use practically zero electricity.

I think you are making an incorrect assumption here. Assuming that these are microFIT installs, and you would have to be crazy not to be part of the microFIT program if you live in Ontario, then the solar panels do not affect consumption so these homes likely do consume a lot of power.

On my microFIT system, which is based on the 2015 price, I earn $0.384/kWh for the power I produce for the next 20 years. The price that I pay Toronto Hydro for power ranges from $0.087-$0.18/kWh based on time of day - plus certain fees for distribution, transmission, etc. So the all in cost that I pay for power is about $.14kWh for off-peak and $0.23/kWh for peak. But for every kwH I sell I get paid $0.384. With microFIT you have separate meters for generation and consumption and you can't feed your own house directly, so having panels doesn't help you if the grid is down.

 

[daniel]

FWIW, I don't have time-of-day metering. I set my car to charge at 1:00 a.m. just to be a good citizen. I think I pay around 6 or 7 cents per kWh. Our electricity here comes from hydro. So I like to say that my car is water powered.

 

[mknox]

I own two EV's. Charge them predominantly in the cheap overnight period.

The local energy company was sending me a notice every month showing that I was in the top 40% of electricity users on the street. LOL.

I only have 1 EV and I am regularly the #1 highest user in my area. Badge of Honor?

 

[mknox]

The one thing not mentioned here is the fact that the real choke point is at the local distribution level. Ontario generally has adequate generation and transmission capacity, and most LDCs have sufficient capacity at the primary feeder level, but it breaks down when you get to the local neighborhood level.

Distribution systems generally have 1 transformer (that green box on the lawn or tank on the pole) for every 6 to 10 homes. With diversification, each of these 6-10 homes may be drawing just a couple of kW a piece. Everything is sized for that type of load and the fact that most home loads cycle on and off giving this diversification (i.e. everyone's devices aren't all on at the same time).

In fact there is a phenomenon that occurs after a prolonged outage called "cold load pickup" whereby when the power is restored, everyone's heat or a/c all comes on at the same time overloading the local system. Utilities will sometimes go around and pull half of the meters (to cut power) on homes fed by a transformer, re-energize, let the first group of homes stabilize then plug the meters back in to the remaining homes. It's all based on larger loads like heat or a/c cycling on and off randomly.

Utilities also count on lower loads at night for the transformer's "cooling cycle". These devices are oil filled and heat up quite a bit with the day's loads, then are expected to cool down at night. Without the cooling cycle, the lifetime of the transformer is reduced.

I think you can see where I'm going with this. Add a couple of Teslas, with their 10 or 20 kW continuous load cycles to a neighborhood, and it all falls apart.

I think a combination of smart grid tools (like smart scheduled charging) combined with infrastructure upgrades ($$$) would be needed for a 100% EV penetration scenario.

 

[Blackout]

One of the criticisms that you hear against EVs is that the grid couldn't support the power draw required. I am trying to do the math for Ontario - please give me feedback to see if I am correct.

There are 6.7 million light vehicles in Ontario so I will use that as the total fleet number. The distance that I have for average distance travelled is 16,000km/yr which is 44km/day. Going 44km in my P85 Model S will use about 11% of the battery or 9.3kWh. The average person could "fill up" in just under an hour with a 240V 40A connection. And this assumes that a Model S is a good proxy for the average vehicle in terms of fuel/power consumption.

Multiplying this by 6.7M gives you a daily draw of 62.3GWh. Spread out over 12 hours, assuming that most charging would be done from 7pm-7am gives an average draw of 5GW. The total capacity of the grid is around 20GW but, of course, that can vary as some generation types, like wind and solar are intermittent.

But it seems to me that there likely is enough capacity to supply that additional amount of power - as the draw at night is generally well below 15GW but that may not be the case on very hot summer evenings. Currently (10am on Saturday) the demand in Ontario is 12.8GW. One issue might be managing when EVs are charged - it might make sense to have some randomizer or some process of managing which vehicle is charged when as you don't want all of the vehicles charged at once.

Some other ways of looking at things. The total energy required in a year to power the fleet by my calculations is 22.7TWh. In the previous few years the total output in Ontario was 137TWh so this would increase overall demand by about 17%.

I may be slightly underestimating the power consumption since there would also be vampire losses of a kWh per day or so depending on the vehicle.

Any comments on whether my math is correct?

I tried to start a conversation about this but it didn't get much traction...I'm happy to see someone is interested in it...I ask myself the same question since January...but I based my calculation differently...Basically, I estimate the capability of the current transformers installed in a neighborhood. And based on the load if every one would plug in their car at the same time, it would not be able to supply the load. Now I don't have my model 3 yet, but I do question your claim when you say people could fill up under an hour at 240V 40A. I do see where the math come from, but For some reason I have a feeling the reality is different, I would put several hours to that but maybe I'm wrong.
The question is in a neighborhood where 500 resident are currently living and let's say 250 of them needs to charge their cars for the night, how many of them can plug in to charge at the same time before the transformer blows up? Are they really going to stay plugged for only an hour?
Even if they don't fill up if they have a 60D and the charge half way, that's already 30KwHr which is equal to what an average Canadian household consume in the winter. So the grid would see half of the houses in this neighborhood pluggin in their cars, double in energy consumption, can the system handle that extra load? According to my research, as it is now, no it cannot.

 

[Saghost]

I tried to start a conversation about this but it didn't get much traction...I'm happy to see someone is interested in it...I ask myself the same question since January...but I based my calculation differently...Basically, I estimate the capability of the current transformers installed in a neighborhood. And based on the load if every one would plug in their car at the same time, it would not be able to supply the load. Now I don't have my model 3 yet, but I do question your claim when you say people could fill up under an hour at 240V 40A. I do see where the math come from, but For some reason I have a feeling the reality is different, I would put several hours to that but maybe I'm wrong.
The question is in a neighborhood where 500 resident are currently living and let's say 250 of them needs to charge their cars for the night, how many of them can plug in to charge at the same time before the transformer blows up? Are they really going to stay plugged for only an hour?
Even if they don't fill up if they have a 60D and the charge half way, that's already 30KwHr which is equal to what an average Canadian household consume in the winter. So the grid would see half of the houses in this neighborhood pluggin in their cars, double in energy consumption, can the system handle that extra load? According to my research, as it is now, no it cannot.

But if every owner is charging 30 kWh every night as your assumptions appear to require, they must be driving ~100 miles/160 km per day, 36,500 miles per year. Most people don't, even in a Tesla.

Since you're looking at neighborhood averages, average miles per day is a sensible way to go here.

Replacing every gas mile in the US in 2014 with a Model S mile would increase the total electric consumption by 26% according to the math I did a while back.

If everyone plugs in a 5 pm and charges for that first hour, yes the infrastructure would be in trouble. But there's no need for that, and as several folks have pointed out in various forms up above, there are a bunch of methods and tools which would permit better scheduling of the timing to avoid the infrastructure impact.