Ignore the electricity grid Chicken Littles

There has been a lot of talk lately about electric vehicles and their impact on the electricity grid. Some say we can’t transition to renewable energy because electric cars will swamp the grid with demand. GMO’s Grantham was the latest of a long line who are concerned:

Perhaps the last great irony of the fossil fuel era will be that going off fossil fuels in the long run will require using one more spurt of fossil fuels in the short run. The quicker we convert our power grid, the worse the energy squeeze will be, and we cannot risk moving slowly.

I can’t make the numbers add up to anything like a problem. I have never seen anyone who claims the grid can’t handle a transition to electric vehicles express the problem in numbers that make sense to me.

Let me show you my numbers roughly. Hit me up in the comments if you can show me where I have gone wrong. 

Misleading data

I’m going to use the US as my example. The annual distance travelled in the US per person is much higher than in other countries, so other countries will find the transition easier. 

I think where most go wrong is looking at this chart:

US energy consumption by sector

But this chart is in units that are not comparable. It measures energy input. It doesn’t account for efficiency.

Setting a high hurdle for electric vehicle sales 

I’m going to start with the heroic assumption that every new car sold will be an electric vehicle until the entire fleet is replaced. The US sells about 17m vehicles per year. There are approximately 290m cars registered in the US. So we are already talking about more than 17 years to replace every internal combustion engine.

Estimation method 1: Total electricity grid energy usage 

The key here is that internal combustion engines are really (really) inefficient compared to electric vehicles. Check these:

ice vehicle efficiency
electric vehicle efficiency

Looking at the US energy consumption chart above:

  • The US uses about 26 quads of oil energy for transport, 22 if you exclude planes. 
  • But, that only translates to about 4.4 quads (22 quads x ~20% energy to wheels efficiency) of mechanical energy because cars are so inefficient. 
  • So, you only need 4.9 quads (4.4 at ~89% energy to wheels efficiency) of energy from the grid to get the same mechanical energy from an electric vehicle.

4.9 quads vs total electricity = a 13% increase in the grid. Spaced over 17 years. i.e. the grid needs to grow less than 1% per annum. It is not a problem. 

Sound too good to be true? I thought it did. So I looked at it another way.

Estimation method 2: Bottom-up electricity grid energy usage

  • The US total vehicle distance travelled is around 5.85 trillion km per year. 
  • An electric vehicle uses about 15 kWh per 100km. 
  • This means we are going to need 0.9 trillion kWh of electricity.

0.9 trillion kWh vs current US net generation of 4.1 trillion = a 21% increase in the grid. Spaced over 17 years. i.e. the grid needs to grow slightly more than 1% per annum.

Plus, in both calculations, I haven’t accounted for electricity savings from oil refining that will no longer be needed.

Is household energy use the bottleneck?

What if we limit the increased usage just to households, in case that is where the problem is? While there is <2% extra annual power needed in both of the above calculations, there may be a problem if this increased capacity is concentrated in households that may need major power grid changes to cope. Thus:

  • An electric vehicle uses about 15 kWh per 100km
  • The average US driver drives 14,000mi or 25,000km per year
  • To be very conservative, assume 80% of charging is done at home
  • There are about 1.9 cars and drivers per household.
  • So, each household is going to need 5,685 kWh more power per year
  • The average household use at the moment is 10,715 kWh/year, which means it will consume 53% more electricity, or ~3% annual increase over 17 years.

So, it looks like considerably more change to household power. But still, we are talking about 3% per year change, not a tall order. And, rooftop solar is meaning that more energy is being created at households already. 

Wrap up

How fast would you need to grow the electricity grid if you banned the sale of any internal combustion engine vehicle? One method says a little less than 1% per annum, the other says a little more than 1%. Say the numbers above are out by 100%, then we are still only talking about a 2% per annum increase in the grid size needed. 

Sure there are issues with how the power is generated, fast charging requirements, whether production could keep up, and a host of other short-term frictions.

But the overall size of the problem is just not as significant as it’s made out to be.

————————————————-

Damien Klassen is Chief Investment Officer at the Macrobusiness Fund, which is powered by Nucleus Wealth.

Fill out my online form.

Follow @DamienKlassen on Twitter or Linked In

The information on this blog contains general information and does not take into account your personal objectives, financial situation or needs. Past performance is not an indication of future performance. Damien Klassen is an Authorised Representative of Nucleus Advice Pty Limited, Australian Financial Services Licensee 515796. And Nucleus Wealth is a Corporate Authorised Representative of Nucleus Advice Pty Ltd.

Follow me

Comments

  1. Grid investment isn’t shaped by total electricity use in a year. The grid is designed to cope with peak demand, and increases in peak (not total) electricity demand are what will drive future grid capex. To use your stats, households will consume 53% more electricity. But the real issue is that households will consume the vast majority of that 53% within a narrow window of hours (e.g. when they get home from their commute and plug in their EV) and all households will do it at the same time. So what’s the % change in peak demand? It will be much more than 53%. The fact that this will all be happening just when solar is ramping down also presents major challenges, because if you can’t precisely synch peak demand with a similarly large ramp up in generation / storage discharge, then frequency starts dropping and the grid shuts down. There are solutions to all of these problems (and grid capex is a last resort) but it’s more challenging than the top-down math would suggest.

    • arescarti42MEMBER

      Agree. If electric vehicles pose any problem for the grid, it is peak demand. If everyone with an electric vehicle plugs it in to charge when they get home from work, the grid is toast.

      If they can delay it a few hours and charge after midnight, then we’re fine. If they can charge it while parked at work, then the grid is fine, and we’ve also solved the problem of excess solar production during the day.

    • True, I think, guess easiest way to reduce peak load would be to have all companies install chargers & have lots in car parks too, so government incentives needed, which should be cheaper than building more power infrastructure.

    • Diogenes the CynicMEMBER

      I cannot agree with that. Most people with electric cars seem to charge them at their work or in their garage at home. If you have a PV roof and face it a lot with EVs will then you also want to charge your e-vehicle in the day as your cost of fuel is very low (possibly zero if you get nothing for your exports to the grid). There is no need to charge every night unless you are doing massive kms each day. Few people do this as currently they only refuel once a week / fortnight and often that is not a full tank. So this idea that everyone will drive home and recharge at night seems off, some will but with work from home, the economic incentive to take advantage of cheap energy (solar PV) and the lack of need to be charging every night I’d expect this impact on the grid to be minor. I could d even envisage the cars being used VtL as backup battery to keep the household running during the evening and again charging during the day when energy is cheap (whether at work or at home due to WFH or on weekends).

    • TheLambKingMEMBER

      But the real issue is that households will consume the vast majority of that 53% within a narrow window of hours (e.g. when they get home from their commute and plug in their EV) and all households will do it at the same time.

      1) The grid is set for ‘peak’ which happens on about 4 days a year. 99% of the time the grid has plenty of capacity to meet this ‘end of day’ demand.

      2) You engineer a solution. Currently we have a lot of ‘off peak’ tariffs that people use to power their hot water. So you have a ‘smart charger’ that works during this time. (https://www.cleanenergyreviews.info/blog/best-smart-ev-chargers)

      3) You use economics to drive a solution. Create a ‘peak’ tariff during that time AND create a ‘peak’ feed in tariff (pay more for power into the grid at that time. That way if the feed in rate is high enough people will buy home batteries or charge their car for free during the day and FEED INTO the grid when they get home (charge at work, or at the free service at the shopping malls)

      • 100%. If grid export tariffs for homes are suitable high and if battery life is extended to many more cycles, cars may be able to decrease peak grid loads and spread out high demand over a longer period.

      • PalimpsestMEMBER

        Exactly so. Modern EV’s allow you to set charging times in the vehicle or, for at-home chargers, in the charger to go off-peak. I’m normally doing less than 60k a day in a typical week, and I get to recharge in the weekend. I can keep between 20 and 80% battery charge without midweek charging. A couple of days a week working from the home office balances out the extra trips for meetings and Bunnings. If I’m doing a longer trip then I just use an extension cord and plug in the slow charger to run all night (using mostly off-peak) the night before. I don’t have a home charger, and don’t need one. I just plug into the wall and leave it, either all day (to preferentially consume solar) or all night. This is rare.

        The only person I know that does take a lot of charge is a colleague that has a 150km each way commute. That’s a minimum of 300km each day he comes into the office. He has a wall charger at home and needs it. He swaps to solar input and charging during the day on WFH days – minimal impact on the grid.

        Quick segue – If one can charge during the day, it helps the grid by giving an outlet for all the solar power we generate now. If one can charge from the domestic array, then it arguably helps improve the network balance with local point consumption of local generation.

        PS: for most people, charging every night is wasteful, and unhelpful for battery life unless they set their normal maximum charge to 70%. If you do a deeper cycle like I do then the 20-80% rule, with exceptions for long trips is the optimal choice.

  2. They say if every car in Texas was changed to an EV then electricity consumption would be 30% more, so if it takes 20yrs to replace all their cars that would pretty much align with your growth figures per year,

  3. Muttafukaburrasaurus.MEMBER

    If you’re not using Australia specific data, you can’t claim results are relevant.
    The size of land mass, population, infrastructure, energy usage / availability etc. On a per capita cost basis is completely different.

    • Damien KlassenMEMBER

      I have done the Aussie & European calcs as well, I’ve just shown the US for simplicity and because they are a worst-case scenario. Australia looks much better than the US as the average distance driven is so much less.

      • Muttafukaburrasaurus.MEMBER

        I doubt the US has anywhere near the level of reliance upon fossil fuel for its export income.
        Electrification needs to start with stationary energy consumption, the vast majority of the Australian land mass doesn’t have any grid infrastructure whatsoever.
        Remote Australia is still largely diesel powered as well as the source of the majority of exports.

        • Australia has the highest rate of solar roof uptake in the world. This will eventually going into a home battery and then the car at night or it will go straight into the car. Australia is one country that will thrive on electric cars. And if there’s a problem then Elon will fix it.

      • Damien, I’m sorry to say, I find your analysis overly simplistic.
        If you want to get a more accurate understanding of the topic, I suggest you start here:
        https://xenetwork.org/ets/episodes/episode-139-vehicle-grid-integration/
        The above episode is outside the paywall, so you can hear the whole thing.
        I would strongly recommend you subscribe and listen to or read the transcripts of the whole lot – the entire energy transition topic is worth the effort.

        • If you can’t explain (or rebut) something simply then you don’t really understand it.

          • When you simplify a complex topic, you don’t make a simple explanation, you just make your explanation wrong…

        • Damien KlassenMEMBER

          First – great resource. I’d seen some of their stuff at the Rocky Mountain Institute, but I hadn’t seen this site.
          Second – I don’t think it is that complicated. If you listen for the 2 minutes at 46:50 mark he basically agrees with my entire argument!.

      • cZ0mzqFILC8zoVHqMEMBER

        I believe one factor not properly addressed with your North American example may be extremes of ambient temperature. What you deride as waste heat inefficiencies have a use in maintaining operational parameters. A heuristic for lead/acid batteries in climates colder than Aus. places efficiency at -20C to be about 10 percent of the same unit at +20C. The ‘waste heat’ also has a role in maintaining operators at optimal efficiency.
        Reports over the local winter excoriated Tesla for endangering the lives of passengers when they simply shut down in rural areas with no service options and no ability to self-diagnose; a classic black box problem.
        Your modelling might do with a disclaimer: ‘Assuming standard temperature and pressure’.

  4. TheLambKingMEMBER

    Not sure why everyone keeps missing a big point – refining oil into gasoline/petrol consumes huge amounts of electricity. Stop refining oil and you have a massive decrease in electricity consumption. The articles I have seen say that it is anywhere between 90% to 10% of electricity to drive the same distances (making up numbers to explain what I mean – if driving 100km in a an EV takes 10kWh, to refine enough oil to drive the same distance in an ICE might ALSO takes between 9kWh to 1kWh – as well as burning the petrol)
    https://greentransportation.info/energy-transportation/gasoline-costs-6kwh.html

    • Damien KlassenMEMBER

      Yeah, this came from an Elon Musk quote. I think it is an urban myth. If you look at the chart above you can see that industrial uses 3.25 quads of electricity, and I think less than 10% of this is oil refining. So, net effect is that I think oil refining might get you 5% ish of the energy electric cars will need. Small enough that I ignored it to avoid the argument!

    • Yeah most energy use in the oil production cycle unsurprisingly comes from oil/gas not electricity (ie you burn your own gas for power)

  5. Forrest GumpMEMBER

    There is sufficient spinning reserve and line capacity on the network (amongst other supply availability) for charging cars, predominantly at night. The electrical load profile is significantly low at night.

    Plenty of other arguments I can offer here, but there’s no use. For those that are interested, this has already been modelled by utilities. This is just a red herring argument.

    Signed Forrest Gump:
    20 years experience in power Distribution & Transmission engineering ….(Vic/NSW/WA/NT/Tas) not that it really matters

    • Is that the lived experience? On hot days, with half the east coast running aircon for days on end the grid craps itself constantly. To then plug incremental 40-80kw/h of load per household at peak times will trip even more local infrastructure. In this scenario, we cannot build for the average. You either build for those edge case days or you provide incentives/disincentives to smooth that load surely?

      • Forrest GumpMEMBER

        On hot days, with half the east coast running aircon for days on end the grid craps itself constantly. To then plug incremental 40-80kw/h of load per household at peak times will trip even more local infrastructure. In this scenario, we cannot build for the average. You either build for those edge case days or you provide incentives/disincentives to smooth that load surely?

        Forrest sighs heavily…says to himself..”We have another one in the audience…”
        its rarely hot across the ENTIRE EAST COAST say from Hobart to Brisbane…AT THE SAME TIME. Timing matters, its called temperature diversification. So scratch that comment.
        Half the east coast does not have Air con, and if they did, half the east coast would not need to operate their AC in total synchronicity (AT THE SAME TIME) This is called load diversity.

        Most cars will be charging in the evenings when the load profile is at its lowest, that is, you are home from work, your business/factory/industry/trains/shops…bla bla bla are closed, lights out, generally speaking, but not always. So the load on the grid is lower. This is a fact. You can look up grid load profiles on the web.

        So charging cars in the evening with lower demand is not an issue. Like I said above, this argument is a red herring. Its a waste of time to debate this.

        • Charging EV’s at 5 or 6pm using home chargers drawing from 5kW to 20kW of power is very likely to play out, and 5- 6pm IS rush hour for most grids. EV’s could easily blow out the peak load scenarios given they have potential to draw such a high power demand (depending on the size of the charger) 20kW chargers are not uncommon… name another home appliance used regularly at 5pm that draws 20kW??

          EV owners like to keep the battery ‘topped up’ – fact. So for the convenience of the average owner not too worried about ‘off-peak tariffs’ then EV will be plugged in as soon as they get home from work (5 or 6pm)

          And in WA there is no off-peak Tarif to exploit! so in WA EV’s will all be plugged in at 5pm, same time as all them ducted AC units are switched on for cooling in summer ! A new peak demand profile is coming…sure its some years away but its coming

          • drsmithyMEMBER

            There’s (broadly speaking) no need for an EV to actually be charging at that time, though, that’s just when they get plugged in.

            Shifting that load to, say, 12:00-05:00 shouldn’t be difficult, surely ?

  6. 50kh/w average pack size in cars equates to approximately 300km driven distance. Assuming 80% use of pack gets you to 240km driven. Average distance per day is 30km (I believe) which equates to top ups every 8 days.

    On an energy use calc this equates to 0.167 kw/h per km driven (on average).

    Per 1 Million Cars per day = 0.167 x 1m x 30km = 5.01m Kw/h per day energy use
    Per Car per day = 0.167 x 1 x 30km = 5.01kw/h per day
    Average household annual energy use 10,715 kw/h = 29.356 kw/h per day

    5.01 kw/h used / 29.356 kw/h daily energy use = 17.067% Increase in daily energy usage

    You can assume bulk charging would take place Thursday to Monday, thus placing bulk charging load on 5 days of the week, not 7. This would take daily demand per household from 17.067% up to 23.89% per day on key charging days.

    If you assume 100% swap out of the ICE to EV occurs, at that point you will need 24% greater load capacity on peak days. The problem is further amplified by time of day charging which means you would likely need to build for demand between 5pm to 2am each day. AEMO and retailers would need to smooth demand out with incentives/disincentives to avoid tripping local grid infrastructure.

    • Damien KlassenMEMBER

      I think the numbers we have are pretty similar. I assumed 80% of charging would be done at home, the rest at charging stations or work which brings your 24% back to 20%. Its going to take about 20 years to switch out the entire fleet.

      So we need to fix about 1% of the grid each year… yes it will need to be fixed, it just doesn’t seem like a herculean task…

    • Forrest GumpMEMBER

      You haven’t resolved the other side of your leger.
      1 example is when you are charging EV, you are no longer pumping a bowser at a petrol station. (Note to self, Bowsers run on grid electricity)

      I’ll allow you to ponder over the plethora of other examples

      • A 1kw/h petrol pump + ancillaries being compared to a fast charged 50kw battery pack. Sorry, I dont understand your point.

        Its a substitution argument with electricity replacing hydrocarbons as a means of propulsion and the grid being able to meet 100% of that demand. Everything I have read states that the grid COULD cope but requires a number of innovations to occur prior to large scale EV adoption. Researchers are working on this now but to make the wholesale statement that “we are all good” and anything counter is a red herring makes me wonder the value behind your experience. You noted above that there are reports which state we are good to go, please share them so we can review. On the flipside, here are a couple of places that disagree with your position.

        Lots of caveats here –
        https://www.lek.com/sites/default/files/insights/pdf-attachments/2057-preparing-grid-uptake-EV_2.pdf
        https://www.gses.com.au/what-are-the-vehicle-to-grid-implications-in-australia-electric-cars-and-the-grid/

        Here is another – repeating the same caveats;

        “EV charging can provide benefits to the grid if appropriately managed. If they are charged when there is plentiful cheap solar and wind power they can increase the use of renewable energy, with less need for electricity storage.

        Conversely, if EV charging is uncoordinated, additional generation and network investment may be required, increasing total electricity system costs.

        Working out how to manage EV charging to best complement an electricity system, increasingly powered by renewables, will require new technologies and business models, as well as coordination between the EV industry, electricity sector – including retailers, networks and market bodies – and government. ”

        https://arena.gov.au/renewable-energy/electric-vehicles/

    • TheLambKingMEMBER

      We get a wind drought every 5 years or so that would require something like 70,000 hornsdale batteries to get us through

      A few issues with that statement:
      1) The whole East Coast of Australia never goes into a wind drought – particularly off the coast – particularly at higher altitudes.

      2) The sun still shines every day. Solar PV cells are reducing in price every year.

      3) Battery installations are increasing by a factor of 4 every year (more if you add EV’s) – so that is only 8-9 years to get to 70,000 hornsdale batteries! Battery prices are reducing every year – we put in twice the capacity this year for the same price as last year. You guys don’t understand the speed of compounding – the speed of which batteries and solar PV are reducing in price will change everything.
      Read https://www.rethinkx.com/energy – we will have so much spare, (close to) free energy that we will make industries to consume that power (like producing proteins in factories to make milk and meat cheaper than via agriculture).

      • In the long run you’re (I hope at least) probably right. In the short to medium term continued massive cost reductions in batteries and solar will be tough – solar cost hasn’t really come down any further in the past few years and battery was only down 6% between 2020 and 2021 (just quoting BNEF).

        The transition to a renewable grid looks like it’s going to be volatile and expensive.

  7. Avg age of Aust vehicles is 10.1 yrs, and the way Aussies take to shiny new things when it hits desired price point, its maybe to be closer to 8 or 9 yrs to hit a critical mass.
    What’s the current cost trajectory of ev’s and home batteries ? They may start to be packaged together.
    My guess is smart charging systems will be required / mandated with instruction from the grid to cease charging (maybe at 50%) if the grid gets unstable or overloaded at a given time.
    There’s any number of ways it could be managed.

    Offsite / remote infrastructure will be key. Govt will have to implement / subsidise before flogging it off to mates once profitable.

  8. It wont be a problem for the grid because most people wont get an EV. Have a look at what people are driving. Just stand and watch the road. Don’t be distracted by the nice cars, look at the majority instead. Most people are driving sub $15k cars. If not you are in a fancy part of town and where you live doesnt reflect reality for the majority.

    So if average people are going to be in EVs, they will be old EVs. At least 10 years old. Old batteries wont hold charge like a new EV. The older they get and harder you charge them the worse they get. So the majority of EVs on the road will require frequent charging. Daily charging. Like your phone.

    I think most people need to get used to the idea that they wont be driving anymore. Which is fine. Ive said for a decade, most peoples (average not well off peoples) EV future is an E bike and a Gortex jacket.

    • Most people won’t get an EV.

      I’ll take that bet.
      Could you kindly define the parameters for the time and define most? Is this a simple majority of car owners?

      Thanks Forrest and Damien.

      Doesn’t V2G, smart inverters, and the inbuilt smarts in the cars themselves sort this out?

      Sounds like vested interests crying to me.

    • drsmithyMEMBER

      Cheaper EVs will be on the second hand market at similar to ICE prices in 7-10 years. Because cheaper new EVs are coming down the pipe, latent demand is huge, in probably five, ten at the most, the majority of new cars sold will be EVs – that’s all that will be available – and people will continue to buy them.

      Who knows, the next Government (or the one after that) might even relax import laws which will completely change the second hand car market.

      Electric bike and scooter takeup will be huge, but the big loser there will be public transport, not cars. And much like in countries with very high public transport utilisation, car ownership will still be commonplace.

    • Spot on. Someone else commented on the fact that batteries have not come down in price much recntly. In fact,with oil waning (yes, we’re over the energy cliff) mining all those minerals that go to make up the batteries (with oil/diesel) is going to make those EVs ever so much more expenseive, not less